183
INDOOR ENVIRONMENTAL QUALITY
This chapter deals the indoor environmental quality assessment in the following parts
with respective literature survey included:
Introduction to Indoor Environmental Quality(IEQ)
Methodology of IEQ optimization in Office Buildings with parameter and algorithm
wise results and discussions
Methodology of IEQ optimization in Resident Buildings with parameter and
algorithm wise results and discussions.
Bibliography
4.1. INTRODUCTION
There is a large untapped opportunity for economic benefits, resulting from
improvements in indoor environmental quality (IEQ) in non-industrial work places and
houses. (WilliamFisk & OlliSeppanen, 2007). The most clearly established sources of
economic benefits include improved work performance, e.g., work speed or quality, reduced
absence, and reduced health care costs. There is also evidence that providing better IEQ can
improve student learning which, in turn, should lead to more effective future workforces.
(WilliamFisk & OlliSeppanen, 2007). At the societal level, economic value can also be
assigned to the reduced suffering of ill health and to extended average lifetimes expected
when IEQ is improved.
The effects of indoor air quality on productivity became an issue only in the
last decade, as a result of extensive research and an understanding of the strong connections
between factors such as ventilation, air-conditioning, indoor pollutants and the adverse
effects on health and comfort. The complexity of a real environment makes it very difficult to
evaluate the impact of a single parameter on human performance, mostly because many of
them are present at the same time and as a consequence, act together on each individual. One
way of evaluating the performance is the use of self-reported performance.
184
It is clear that the indoor environment was evaluated to have the biggest
influence on performance, providing better job satisfaction and reducing job stress. A
common approach, to evaluate the influence of climatic factors on human performance could
be to measure the extent to which the Sick Building Syndrome (SBS) symptoms occur, as
these symptoms are known to cause distraction from work or even short-term absenteeism.
However, this link is not well established yet and must be better understood and recognized.
Possible reasons may be cited as follows: (1) inadequate ventilation or superfluous
emissions from different sources increase the concentration of pollutants, which negatively
affect perceived air quality; (2) reduced air quality negatively affects the central nervous
system, increasing SBS symptoms such as headache, difficulty in concentration, tiredness;
(3) these symptoms will cause distraction from work and decreased work ability, i.e.
productivity loss. (Olesen & Bjarne, 2009), indoor pollution may also exacerbate the
sensation of dryness and irritation of eyes. As a consequence, a higher blinking rate and
watery eyes will negatively affect visual skills and decrease the performance of visually
demanding work.
The indoor environmental quality (IEQ) in offices and residences are
examined from the prospect of an occupants‘ acceptance in four aspects: thermal comfort,
indoor air quality, noise level and illumination level. Based on the evaluations made by the
occupants of the IEQ, empirical expressions have been proposed to approximate an overall
IEQ acceptance of an office environment at certain operative temperature (To), carbon
dioxide concentration (CO2), equivalent noise level (Leq) and illumination level (lux). The
overall IEQ acceptance is calculated from a multivariate logistic regression model.
4.1.1. Literature survey
Physical environmental parameters such as air temperature, relative humidity,
acoustics, air quality, lighting, ventilation and air distribution are all interrelated, and the
feeling of comfort is a composite state of an occupant‘s mind, responding to the senses to
these factors (Goldman, 1999) (Haghighat & Donnini, 1999) (Mendell M. , 2003) (Naganoa
& Horikoshib, 2005). This state of mind is an intricate response to the indoor environmental
factor groups, including physical environment sustained by the building and its service
system, and individual physiological conditions such as health, social relations, financial
state, etc. Studies showed that an occupant‘s acceptance of an environment depended on a
185
number of environmental parameters. Four basic components, namely thermal comfort,
indoor air quality, aural and visual comfort were identified for determining an acceptable
IEQ. Conventional studies on indoor environment addressed each of them separately. They
are still addressed independently by designers for many office designs. More recently, the
equivalence of the discomfort caused by different physical qualities has been considered
(Clausen, Carrick, Fanger, KIm, & Rindel, 1993) ; (Fanger, Olf, & Decipol, 1988) (Pellerin
& Candas, 2004). Discomforts caused by indoor air pollution, thermal load and noise were
investigated. It was reported that at an operative temperature between 23 and 29 1C, each
degree Celsius change would associate the same effect on human comfort with a change in
perceived air quality of 2.4 decipol, or a change in noise level of 3.9 dB (Fanger, Olf, &
Decipol, 1988) . For levels of perceived air quality up to 10 decipol, a unit change had the
same effect on human comfort as a change in noise level of 1.2 dB (Clausen, Carrick, Fanger,
KIm, & Rindel, 1993).
The equivalence of acoustic sensation to the thermal one was proposed for
short-term exposure. Each degree Celsius change in temperature had the same effect of 2.6
dBA (Pellerin & Candas, 2004). Workplace variables inducing the largest number of health
symptoms, comfort or odour concerns were investigated by multivariate regression analysis
(Bulysssen & Cox, 2002); (Dan, 1993); (Toftem, 2002); (Sofuoglu & Moschandreas, 2003);
(Wallace, Nelson, & Dunteman, 1991). It was realised that a successful control of the indoor
environment required an understanding of the integral indoor environmental parameters. An
overall IEQ index would be derived to describe the state of mind of a user in balance with the
indoor environment. Unacceptable indoor environments are often manifest in some forms and
symptoms of sick building syndrome (SBS) prevail in many office buildings (Sofuoglu &
Moschandreas, 2003).
This study argues that the subjective evaluation of an indoor environment
being perceived by an occupant can be used to assess the acceptance of the IEQ. In particular,
occupants‘ acceptance of the four basic components of IEQ was evaluated and correlated
with the overall IEQ acceptance of an office environment. The occupants‘ attitudes towards
the operative temperature, CO2 concentration, equivalent noise level and illumination level
and the overall IEQ acceptance recorded by a dichotomous scale were studied (Houser &
Tiller, 2003); (InternationalStandardISO7730, 1994); (Wong & Leung, 2005).
186
Mathematical expressions were proposed for the overall IEQ acceptance using
a multivariate logistic regression model with the former four parameters recorded. The
proposed overall IEQ acceptance can be used as a quantitative assessment criterion for an
office environment and similar environment where an occupant‘s evaluation is expected. The
occupant‘s acceptance of the perceived indoor environment given by four aspects, namely
thermal environment, indoor air quality, equivalent noise level and illumination level, was
studied with a dichotomous assessment scale (Portney & Watkins, 2000).
Indoor environmental quality (IEQ) and occupant comfort are closely related.
Current indoor environmental assessment includes four aspects, namely thermal comfort
(TC), indoor air quality (IAQ), visual comfort (VC) and aural comfort (AC) (Clausen,
Carrick, Fanger, KIm, & Rindel, 1993); (Wong, Mui, & Hui, 2008). Maintaining satisfactory
thermal comfort conditions for the occupants by an adjustable indoor temperature set point of
the air conditioning system is one of the primary concerns in many air conditioned office
buildings. Thermal comfort relates human sensation and perception with a number of
environmental and physical parameters (Faanger, 1970). It is by definition the perception of
satisfaction, a subject experiences, in a given thermal environment (ANSI/ASHRAE55-2004,
2004). Extensive studies resulted in a number of thermal comfort equations as proposed in
some widely used design guides and standards (ANSI/ASHRAE55-2004, 2004);
(InternationalStandardISO7730, 1994). Three indices were derived based on Fanger‘s
comfort equation: Predicted Mean Vote (PMV), Predicted Percentage Dissatisfied (PPD) and
Lowest Possible Percentage Dissatisfied (LPPD). The PMV and PPD indices are relatively
common in practical applications (Fanger.P.O., 2002); (Han, Yang, Zhou, Zhang, Zhang, &
Moschandreas, 2009).
`The former predicts the mean thermal comfort votes among a large group of people;
the latter is a quantitative measure of the number of thermally dissatisfied persons in a group
under particular thermal conditions. Field studies on the thermal comfort of occupants
working in an air-conditioned environment can be used to examine the neutral temperature—
a temperature associated with a neutral thermal sensation (Oseland.N.A, 1995); (Fanger,
1995); (Mui,K.M & Wong,L.T, 2007). This temperature is a key factor for selecting an
appropriate air temperature set point for an indoor thermal environment (Wong,L.T,
Mui,K.W, Fong,N.K, & Hui,P.S, 2007). As climate, besides occupant factors, including
lifestyle, economic status and adaptive behaviour, plays an important role in affecting the
187
indoor thermal environment (Yoshino, et al., 2006); (Brager,G.S & Dedear,R.J, 1998),
neutral temperatures for different climatic zones have been studied (Clausen, Carrick, Fanger,
KIm, & Rindel, 1993); (Mui,K.M & Wong,L.T, 2007); (Wang, 2005). IAQ, as the nature of
air in an indoor environment in relation to occupant health and comfort, is neither a simple
nor an easily defined concept. In a broad context, it is the result of the complex interactions
among buildings, building systems and people. Comparative risk studies performed by the
United States Environmental Protection Agency (USEPA) ranked IAQ, as one of the top five
environmental risks to public health (Wang, 2005).
Over the past decades, exposure to indoor air pollutants is believed to have
increased due to a variety of factors, including the construction of more tightly sealed
buildings, the reduction of ventilation rates (for energy saving), and the use of synthetic
building materials and furnishings as well as chemically formulated personal care products,
pesticides and household cleaners. As investigating all types of indoor air pollutants for
general air quality monitoring and assessment is a complicated matter (Wong, Mui, & Hui,
2006); (Hui, Wong, & Mui, Feasibility study of an express assesment protocol for the indoor
air quality of air-conditioned offices, 2006); (Mui K. , Wong, HUI, & Law, 2008), it was
suggested that the measurement and analysis of indoor carbon dioxide (CO2) concentration
could be useful for understanding IAQ and ventilation effectiveness (Persily, 1997); (ASTM,
2003).
Although healthy people can tolerate a CO2 level up to 10,000 ppm without
serious health effects, an acceptable indoor CO2 level should be kept below 1000 ppm or 650
ppm above the ambient level in order to prevent any accumulation of associated human body
odour (ANSI/ASHRAEstandard62-2007, 2007); (Mui.K.W & Wong,L.T, 2007). In terms of
occupant satisfaction, acceptable IAQ means room air in which no known contaminants are at
harmful concentration levels and at least 80% of the people exposed to it do not express any
dissatisfaction (ANSI/ASHRAEstandard62-2007, 2007); (Mui.K.W & Wong,L.T, 2007);
(Hui, Wong, & Mui, 2008). Light enables humans to see. A number of available guides and
codes of practice provide recommendations on adequate indoor lighting designs. For
example, an illumination level of 2000 lx with a colour rendering index not less than 90 is
required for a fabric inspection factory while 500 lx with a colour rendering index between
60 and 80 should be maintained in a general office. Lighting quality attributed to the quantity
and colour spectrum of light can be expressed by a comprehensive comfort, satisfaction, and
performance (CSP) index (Bean & Bell, 1992).
188
Since indoor visual comfort is closely related to the horizontal and cylindrical
illumination levels, adjustment of the illumination level is essential to improve visual comfort
and occupant acceptance (Mui & Wong, 2006). All sounds that are distracting, annoying, or
harmful to everyday activities such as work, rest, study and entertainment are regarded as
noises. In fact, any sound judged undesirable by the recipient can be considered a noise.
Noise can be continuous or impulsive and both types can cause adverse effects on physical,
mental and social well-being. A number of measures were proposed for indoor aural comfort
evaluation, for instance, the equivalent sound pressure level (SPL), the noise criterion (NC)
curves, the balanced noise criterion (BNC), the noise rating (NR), the preferred noise
criterion (PNC), the room criterion (RC), and the loudness level. Given the nature of noises
generally encountered in offices, a previous research indicated the A-weighting equivalent
sound pressure level that has been widely adopted in the studies of noise level within
buildings would be the best and most convenient measure, although the loudness level would
also be a very good alternative (Ayr, Cirillo, Fato, & Martellotta, 2003).
In Hong Kong, surveys showed that the equivalent continuous noise level
correlated well with the occupant acceptance reported in air-conditioned offices (Mui &
Wong, 2006), and in construction site offices where the background noise was not dominated
by air-conditioning but by outdoor noise sources (Mui, Wong, & Wong, 2009).The neutral
sound pressure level found for aural comfort in some typical air-conditioned offices was
between 45 and 70 dBA, with a mean of 57.5 dBA. Studies showed that an occupant‘s
acceptance of an acceptable IEQ environment would be closely related to aforesaid
environmental parameters. More recently, the equivalence of the discomfort caused by
different physical qualities for indoor air quality, thermal sensation and noise has been
investigated (Fanger, Olf, & Decipol, 1988); (Clausen, Carrick, Fanger, KIm, & Rindel,
1993) ; (Pellerin & Candas, 2004); (Hikmat, Hind, & Muna, 2009). It was reported that every
1oC change of operative temperature in range between 23 and 29
oC would produce the same
feeling on human comfort due to a deviation 2.4 decipol regarding the perceived air quality
(Fanger, Olf, & Decipol, 1988), or due to a change in noise level of 3.9 dB. For levels of
perceived air quality up to 10 decipol, a 1 decipol change in perceived air quality had the
same effect on human comfort as a change in noise level of 1.2 dB (Clausen & Wyon, 2008);
(Clausen, Carrick, Fanger, Kim, Poulson, & Rindel, 1993).
189
The equivalence between acoustic and thermal sensation was proposed for
short-term exposure as a 1 8C change in temperature had the same effect on 2.6 dBA
(Pellerin & Candas, 2004). Workplace variables inducing the largest number of health
symptoms, comfort or odour concerns were investigated by multivariate regression analysis
(Hikmat, Hind, & Muna, 2009); (Mendell M. , 2003); (Bulysssen & Cox, 2002); (Dan, 1993);
(Toftem, 2002).. It was realised that successful control of the indoor environment required an
understanding of the integral indoor environmental parameters. The occupants‘ acceptance of
the four basic IEQ components was evaluated and correlated with the overall IEQ acceptance
of an office environment (Wong, Mui, & Hui, 2008).
Mathematical expressions were proposed for the overall IEQ acceptance,
using a multivariate logistic regression model and it can be used as a quantitative measure for
an office environment design. IEQ parameters are interdependent and must be considered
interactively. However, conflicts were reported in the above cases where these parameters
were treated discretely, e.g. maximization of openings for natural ventilation and daylight
resulted in poor acoustic performance and thermal discomfort in an indoor environment in
subtropical climates (Koenigsberger, Ingersoll, Mayhew, & Szokolay, 1974). Laboratory
experiments of controlled auditory and visual stimuli also showed that the visual parameter
was predominant in audio–visual interactions and the visual information would affect the
auditory judgment (Viollin, 2003). Reportedly, attention to a visual form would reduce the
conscious perception of sound, and vice versa when the sound was related to a scene (Yang
& Kang, 2005). In fact, interactions between visual and auditory perceptions gave people a
sense of involvement. This study investigated the occupant acceptance of residential IEQ
through physical measurements and subjective surveys. Mathematical expressions were
proposed for the overall IEQ acceptance using a multivariate logistic regression model with
the four environmental parameters discussed above. The study provides useful information
for developing quantitative assessments for residential environments where an occupant‘s
evaluation is expected.
4.2. Components of IEQ
IEQ depends on the following four components
1. Thermal comfort
2. Carbon dioxide
190
3. Sound
4. Illumination
4.2.1. Thermal comfort
Thermal comfort models that take into account human adaptability have been
developed over the years (Fanger.P.O, 1970); (ANSI/ASHRAE55-2004, 2004). The concept
of adaptive thermal comfort can be described as (Auliciems.A, 1983) ‗When a change occurs
causing thermal discomfort, people react in such a way that their thermal comfort is re-
established.‘ This description refers to behavioural adaptation that can be discerned in
personal, technical, environmental, cultural and organizational adaptation. Physiological
adaptation or acclimatization does not seem to affect peoples‘ neutralities, but there is some
evidence that the acceptability is altered (Nicol & Humphreys.A', 2001). Psychological
adaptation implies a changed perception of, or response to, sensory information. Thermal
sensations are influenced by an individual‘s experiences and expectations in a direct manner.
When applying models of adaptive thermal comfort, one should distinguish between different
types of buildings, usage and climatic circumstances. Occupants in naturally ventilated
buildings have possibilities for increasing the air velocity in the room by operating windows.
By doing so, they can still create a comfortable environment in higher indoor temperatures.
Additionally, it turns out that psychological adaptation plays an important part especially in
this type of buildings: because of the more direct contact to the weather outside, higher
temperatures are also expected for the indoor climate. Fanger‘s PMV-model can only take the
effects of behavioural adaptation into account: the adjustment of clothing and the level of
activity, and the increase of the air velocity.
4.2.2. Carbon di-oxide (CO2)
CO2
has long been used as a basis for ventilation (providing fresh outdoor air
to indoor spaces) design and control. CO2
is a natural product of human respiration whose
rate can be predicted based on an occupant‘s age and activity level. Beginning as early as
1916, CO2 of 800 to 1,000 ppm and 1,000 ppm respectively were recommended. However,
the key point is that CO2
levels are good predictors or surrogates for human emitted bio
effluents (i.e., odours) that are considered undesirable for the overall human comfort inside
conditioned spaces. Thus CO2
is a surrogate for levels of other bio effluents that cause odours
191
that are likely to be viewed as unacceptable by others in the space, not because of their
presence as a direct health hazard.
Since people exhale CO2 as a consequence of their normal metabolic
processes, the concentrations of carbon dioxide inside occupied spaces are higher than the
concentrations of CO2 in the outdoor air. In general, a larger peak difference between indoor
and outdoor CO2 concentration indicates a smaller ventilation rate per person. The ventilation
rate per person can be estimated with reasonable accuracy from the difference between the
maximum steady-state (equilibrium) indoor CO2 concentration and the outdoor CO2
concentration, if several critical assumptions are met, including: the occupied space has
nearly constant occupancy and physical activity level for several hours, the ventilation rate is
nearly constant, and the measured indoor CO2 concentration is representative of the average
indoor or exhaust airstream concentration in the space . For example, in an office space under
these conditions, if the equilibrium indoor CO2 concentration is 650 parts per million (ppm)
above the outdoor concentration, the ventilation rate is approximately 15 cubic feet per meter
(cfm) per person. In many real buildings, occupancy and ventilation rates are not stable for
sufficient periods and other critical assumptions may not be met to enable an accurate
determination of ventilation rate from CO2 data. The American Society for Testing and
Materials (ASTM) states that this technique has been misused, when the necessary
assumptions have not been verified and the results have been misinterpreted. Nevertheless,
CO2 concentrations remain a rough and easily measured surrogate for ventilation rate. In
addition, many studies have found that occupants of buildings with higher indoor CO2
concentrations have an increased prevalence of sick building syndrome symptoms. However,
indoor CO2 concentrations may be poor indicators of health risks in buildings and spaces with
strong pollutant emissions from the building or building furnishings, particularly when
occupant densities are low.
4.2.3. Sound
Noise has several adverse effects on human beings. On the physiological side,
these effects include hearing damage and hearing loss. On the psychological side, they
include interference with speech communication, impairment of performance, and
annoyance. Noise can be very distracting and prevent concentrated mental work. In extreme
192
cases, it can also result in physical disorders. Noise can be characterized in two ways, - direct
and indirect. A direct noise is determined by the intensity of the source and the distance from
the ears. Reflected noise is dependent on the reflection factors of the floor, walls, ceiling, etc.,
and on the position of these surfaces. Direct noise should be suppressed by placing covers
over or by isolating sources of noise from the rest of the work area. A distinction should also
be made between meaningful noise and general background noise. Most working
environments will have some background noise. However, this noise can become
uncomfortable if an irregularity, such as a malfunctioning machine, develops. Reflected noise
can be reduced by introducing sound absorbing materials into the environment. Acoustical
noise is considered a human factor because it affects such factors as a workers‘ comfort, job
satisfaction and performance. Fortunately, however, the noise levels of modern workstations
designed for office or laboratory environments are relatively low, much lower than those of
typical data processing equipment found in computer room installations.
The noise is generated primarily by the single small fan in the system unit,
used to cool the electronics or by the spinning hard disk drive. Displays are usually cooled by
convection and are very quiet. Thus, the primary concern from a human factor point of view
is that, the noise from a workstation may be disrupting and annoying. The noise is not very
loud, but it may be reported as objectionable by the user simply because the office
environment itself is very quiet. Annoyance is a subjective response and difficult to quantify,
but it should not be treated lightly. From an employer's point of view, an annoyed employee
can present morale problems which may affect performance and reliability. Noise control
engineers are striving to lower noise levels of workstations, while at the same time studying
and identifying the psycho acoustical aspects of particular noises that most contribute to
annoyance.
4.2.4. Illumination
Today there is great value in the task/ambient approach to lighting. This
method at first provides general room illumination and then specific, brighter illumination -
only where needed. In this respect, ambient lighting levels may be reduced to save energy
and task area lighting may be increased for optimum human performance.
Poor lighting can be a safety hazard misjudgement of the position, shape or
speed of an object can lead to accidents and injury. Poor lighting can affect the quality of
193
work, specifically in situation where precision is required, and overall productivity. Poor
lighting can be a health hazard too much or too little light strains eyes and may cause eye
discomfort (burning, etc.) and headaches. The amount of light we need varies and depends
on:
The type of task being done (such as demands for speed and accuracy),
Type of surfaces (does it reflect or absorb light),
The general work area, and
The individual's vision.
The amount of light falling on a surface is measured in units called lux.
Lux = Lumens (quantity of light) per square metre.
Illuminance is the amount of light falling on a surface. The unit of
measurement is lux (or lumens per square metre = 10.76 foot candles (fc)). A light meter is
used to measure it. Readings are taken from several angles and positions.
4.2.4.1. IES - RECOMMENDATIONS
Since 1958 the Illuminating Engineering Society has published illuminance
recommendations in table form. These tables cover both generic tasks (reading, writing etc),
and 100's of very specific tasks and activities (such as drafting, parking, milking cows,
blowing glass and baking bread). All tasks fall into 1 of 9 illuminance categories, covering
from 20 to 20,000 lux.
To reach proper light levels, many light fixtures are designed to reflect light
off walls, ceilings and objects. The amount of light reflected off a surface can be measured.
Canadian centre for occupational health and safety suggest the percent of light reflected off
surfaces in a typical office include:
Window blinds (40-50%),
Walls (50% maximum),
Business machines (50% maximum),
Ceiling (70-80%),
Floor (20-40%), and
194
Furniture (25-45%).
The percent value refers to the amount of light that a surface reflects, relative
to the amount that falls on the surface. In addition, light fixtures that are too widely spaced or
wrongly positioned can create shadows. Objects between the light fixture and work being
done can block the light and cast shadows. Likewise, workers sitting with their backs to
windows, with light fixtures directly overhead or to the rear, cast shadows on their own work
surfaces. The immediate work area should be brighter than surrounding areas. If the
surrounding area is brighter than the work area, your attention is distracted away from the
work area. The contrast between colours of objects, such as between the print itself and paper
or text and background on computer screens, can also cause problems. Too little contrast
between print and the paper or little contrast between characters on a video display terminal
screen and the background makes reading tasks difficult. In an industrial setting, moving and
stationary machine parts are hard to distinguish if they are of the same colour.
4.2.4.2. Luminance
Luminance is the amount of light reflected from a surface. The unit of
measurement is candela per square metre (equals 0.29 foot-lamberts). An Illuminance meter
is used to measure it. Several measurements are made and averaged. Luminance tables are
consulted for reference values. Illuminance is a measure of the amount of light falling on a
surface. It is defined as: 'the density of the luminous flux incident on a surface'.
One footcandle is the Illuminance at a point on a surface which is one foot
from, and perpendicular to, a uniform point source of one candela. One Lux is the
Illuminance at the same point at a distance of 1 meter from the source. One lumen uniformly
distributed over one square foot of surface provides an illumination of 1 footcandle.
If you work in feet, your results will be in footcandles - (1 footcandle = 1 lumen/square ft.)
If you work in meters, your results will be in Lux - (1 Lux = 1 lumen/square meter)
Formerly the term 'ILLUMINATION', was used for Illuminance.
195
4.2.4.3 Illumination Levels
The Illuminating Engineering Society (IES) measures light in foot candles, or
"lux," which translates in scientific terms to one lumen per square foot. Full daylight is
characterized as approximately 10,752 lux, while an overcast day measures only 1,075. Lux
is measured in terms of the amount of illumination, or light, covered per square foot.
Recommended lux for the workplace varies by field. A classroom environment has a
recommended lux standard of 250. Workers who do detailed drawing work should have
lighting illumination at 1,500 to 2,000 lux. Specialized visual tasks can require upward of
10,000 lux depending on the profession. It is important today that the lighting designer
provide appropriate lighting levels for the required task(s). It is also equally important not to
diminish light in a task. There is generally little value in reducing lighting in a task where
human performance is concerned. The electrical energy saved is often offset by a far greater
loss in human performance or productivity.
As the eye ages, it requires more light to see the same detail with the same
speed and accuracy. For this reason, lighting systems must be designed with specific human
needs in mind.
Energy restrictions and building codes often tend to limit lighting to 'x'
number of watts per square feet (or meter) in new constructions. It must be remembered that
these are usually 'average' figures in that, a storage room might require lower lighting levels
and an office area might require higher lighting levels than average. These average levels can
and should be increased with the object of providing sufficient lighting for effective human
performance.
4.2.4.4 . Leveraging Daylight
The Illuminating Engineering Society (IES) recommends employers to
leverage available daylight in order to cut down on energy costs while improving lighting for
workers. Sunlight provides a higher lux level than indirect lighting and is more cost effective.
Glazing windows is one approach the IES recommends to prevent glare while still allowing
daylight into the office. Placement of offices, such as east-facing or west-facing, should also
be considered carefully to make the most of the available sun light.
196
4.2.4.5 . Indirect Lighting
Providing the proper levels of illumination to workers prevents safety hazards
and increases overall productivity. The IES recommends use of indirect lighting at
consecutive intervals. Measurements should be obtained to ensure workers have the
recommended light levels. The use of task lights is recommended for certain areas.
Employers must consider the nature of the work an employee is doing while establishing an
illumination target. Greater density of lighting fixtures should be used in offices where detail-
oriented work is performed. In this problem, to find the optimum indoor environmental
quality, ten solvers are used. Each solver has its own characteristics. The characteristics lead
to different solutions and run times. The results are examined based on various criteria.
4.3. IEQ OFFICE
4.3.1. Introduction
Today, the concept of an acceptable indoor environmental quality (IEQ) as an
integral part of the total building performance approach is still not fully appreciated. This
state of mind is an intricate response to the indoor environmental factor groups, including
physical environment sustained by the building and its service system, and individual
physiological conditions such as health, social relations, financial state, etc.
Four basic components, namely thermal comfort, indoor air quality, aural and
visual comfort were identified for determining an acceptable IEQ. Conventional studies on
indoor environment address each of them separately. They are still addressed independently
by designers for many office designs. It was realised that successful control of the indoor
environment required an understanding of the indoor environmental parameters. An overall
IEQ index would be derived to describe the state of mind of a user in balance with the indoor
environment.
Subjective evaluation of an indoor environment being perceived by an
occupant can be used to assess the acceptance of the IEQ. In particular, occupants‘
acceptance of the four basic components of IEQ was evaluated and correlated with the overall
IEQ acceptance of an office environment. The occupants‘ attitudes towards the operative
temperature, CO2 concentration, equivalent noise level and illumination level and the overall
IEQ acceptance recorded by a dichotomous scale were studied .Mathematical expressions
were proposed for the overall IEQ acceptance, using a multivariate logistic regression model
197
with the former four parameters recorded. The proposed overall IEQ acceptance can be used
as a quantitative assessment criterion for an office environment and similar environment
where an occupant‘s evaluation is expected.
Note: We could not find any standards for India, and so we have used ASHRAE,
since they seemed reasonable and applicable in Indian context also. (living-smartly.com)
4.3.2. Methodology
Subjective evaluations made by 220 occupants of indoor environmental
conditions in natural ventilated Faculty rooms in Karunya University were studied. The
sample offices had floor areas ranging from 233.3 m2 to 937.77 m
2. The offices were
spacious with ordinary design and with good quality finishes; flexible layout, average-sized
floor plates, adequate lobbies; good lift services zone and parking facilities were available.
The occupant‘s acceptance of the perceived indoor environment given by four
aspects, namely thermal environment, indoor air quality, equivalent noise level and
illumination level, was studied with a dichotomous assessment scale . This scale was used for
a direct feedback of acceptability with the question ‗Is the thermal environment/indoor air
quality/noise level/illumination level being perceived in the office environment acceptable to
you?‘ being asked. The ranks ‗(1) Yes, acceptable‘ and ‗(0) No, not acceptable‘ were self-
explanatory. In order to confirm the validity of their responses, each respondent had to use a
semantic differential evaluation scale for the subjective assessment of the first two aspects,
and a visual analogue assessment scale for the evaluation of the aural and visual comfort . At
the end of the survey, an overall acceptance of the IEQ was determined.
A total of 220 occupants were interviewed and their evaluations of the IEQ and the
four parameters were obtained. The results are summarized in Table 4.1. The correlation
between subjective response to each parameter and the overall IEQ acceptance was evaluated
by a statistic χ2-test. Results showed that all the four parameters contributed to the overall
IEQ; and the significance between the acceptance votes on the latter and those on the former,
ranking from the most important to the least, was as follows: p = for thermal
environment, for air quality, for noise level and
for illumination level.
198
Table 4.1 Occupant’s votes on acceptance of a perceiving indoor environmental quality
U - Unacceptable; A -Acceptable.
The overall IEQ acceptance θ for an office environment perceived by an occupant
expressed by a multivariate logistic regression model is proposed
Where the regression constants determined from the 220 occupant evaluations are
k0 = -14.98; k1 = 6.04; k2 = 4.92; k3 = -4.70; k4 = 3.74;
Values of k1, k2, k3, k4 confirm the relative importance of the four contributors to θ, the larger
the value, the greater the importance and it is seen that the occupants were very sensitive to
the operative temperature as compared with the other three parameters.
Various combinations of contributors i=1, 2, 3, 4 and the corresponding
overall IEQ acceptance were considered. A total of 24 possibilities were found. Taking the
binary notation for the acceptance i.e., 0 for ‗unacceptable‘ and 1 for ‗acceptable‘ the
predicted acceptance of IEQ (θ) is calculated.
φ1=1 - (PPD/100).
Where, PPD = 100 – 95 x e-(0.03353 (PMV^4) + 0.2179 (PMV^2));
-2 PMV
φ2=1-
(
Overall
acceptance
Θ
Votes Thermal
environment
φ1
Air quality
φ 2
Noise level
φ 3
Illumination
level
φ 4
U
A
98
122
U A U A U A U A
47
3
51
119
52
12
46
110
30
7
68
115
26
9
72
113
TOTAL 220 50 170 64 156 37 183 35 185
199
φ3=1 -
; 67 ≤ ≤ 78,
φ4=1 -
187 1522.
Table 4.2. Overall IEQ acceptance
4.3.3. Algorithms
4.3.3.1. Genetic algorithm
4.3.3.1.1. Options Set for the Algorithm:
Initial population: 20.
Elite count: 2.
Cross over fraction as 0.8.
Max Time Limit: ∞.
Max Generations: 100.
Fitness Limit: -∞.
Selection: Stochastic.
4.3.3.1.2. Stopping Criteria:
If the maximum generations is reached (100).
If maximum time is reached (∞).
Case
No.
Survey
Sample
Contributors Predicted
acceptance
of IEQ θ
Φ1 Φ2 Φ3 Φ4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0
5
1
14
3
3
5
20
1
8
9
27
3
16
13
92
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
1
1
1
1
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
3.1208 10-7
1.313 10-5
2.84 10-9
1.1949 10-7
4.275 10-5
1.7967 10-3
3.88 10-7
1.637 10-5
5.486 10-3
1.31 10-4
1.1918 10-6
5.017 10-5
0.017636
0.43045
1.6326 10-4
6.827 10-3
200
If average change in function value < 10¯⁶.
TABLE.4.3. Results of GA in 20 trails for office IEQ
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 2 1799.9 59.6 1553.7 1 1.208193 51
2 2 1740.6 62.8 1370.9 1 1.133192 51
3 2 1799.7 71.7 1569.3 1 0.775846 51
4 2 1070.5 64.3 1196.7 1 1.089181 51
5 1.9 1662.3 57 1513.8 1 1.059298 51
6 2 1713.5 55.7 1520.9 1 1.140209 51
7 2 1217.2 60.2 1524.2 1 0.768988 51
8 2 1775.7 69.7 1395.2 1 0.91929 51
9 2 1724.6 57.4 1503.8 1 1.172938 51
10 2 1771 63.8 1018.5 1 0.917832 51
11 2 1295.4 71.3 1440.9 1 0.913941 51
12 2 1668.8 60.2 1434 1 1.169327 51
13 2 1549.2 71.4 1232.4 1 1.059507 51
14 2 1510.7 52.7 792.1 1 0.887249 51
15 1.9 1799.9 69 1538.8 1 1.115581 51
16 2 1722.7 65.6 1495 1 1.195485 51
17 2 1611.4 63.8 1265.3 1 1.032374 51
18 1.9 1756.2 54.5 1234.4 1 0.932092 51
19 -2 1789.5 71.9 1580.4 1 1.162234 51
20 2 1295.4 71.3 1440.9 1 0.913941 51
Avg 1.785 1613.71 63.695 1381.06 1 1.028335 51
201
Fig.4.1.Convergence of GA
4.3.3.2. Simulated annealing
4.3.3.2.1. Options Set:
Initial Temperature: 100.
Annealing Function: Fast Annealing.
Reannealing interval: 100.
Time Limit: ∞.
Max.function evaluation: 3000* No. of variables.
Max. Iterations: ∞.
Function Tolerance: 10¯⁶.
Objective Limit: 10¯⁶
0 10 20 30 40 50 60 70 80 90 100-1
-0.5
0
0.5
1
1.5
Generation
Fitness v
alu
e
Best: -0.99999 Mean: -0.9526
202
4.3.3.2.2. Stopping Criteria:
Max. Time reached.
The average change in value of the objective function is < 10¯⁶.
Max. Iterations are reached.
If the number of function evaluations reached.
If the best objective function value is less than or equal to the value of Objective
Limit.
TABLE.4.4. Results of SA in 20 trails for office IEQ
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 2 1379.6 68.4 1014.9 1 2.582472 2004
2 -2 1213.4 70.6 1125.6 1 2.227751 2003
3 2 1513.7 50.5 970.4 1 2.322216 2001
4 2 1160.6 64.2 846.1 1 2.482702 2006
5 -2 1341.8 72 746.3 1 1.90792 2005
6 2 1467.3 53.6 1100.5 1 3.029657 2017
7 -2 1318.7 62.4 1029.3 1 2.866003 2031
8 -2 1219.8 58.4 878.4 1 2.279764 2005
9 2 1127 71.3 959.8 1 2.089766 2007
10 -2 1486.7 56.2 1042.4 1 2.134044 2006
11 -2 982.3 64.3 1116.7 1 2.91151 2011
12 2 1330.5 45.3 789.4 1 3.014331 2002
13 2 1041.9 61.3 854.1 1 2.283927 2013
14 -2 1284.2 71.5 964.4 1 2.42402 2007
15 -2 1303.7 45.5 791.1 1 3.054213 2007
16 2 1174.6 46.5 961.3 1 2.085067 2015
17 2 1410.5 52.1 1145.8 1 2.568922 2016
18 -2 1313 68.8 1081.6 1 2.904454 2004
19 2 1228.1 71.2 828.1 1 2.865376 2016
20 2 1157.9 60.2 994.8 1 1.904684 2025
Avg 0.2 1272.765 60.715 962.05 1 2.49694 2010.05
203
Fig.4.2.Convergence of SA
4.3.3.3. Pattern search
4.3.3.3.1. Options Set:
Poll Method: GPS positive Basis 2N.
Initial Mesh size: 1.
Expansion Factor: 2.
Contraction Factor: 0.5.
Mesh Tolerance: 10¯⁶.
Max. Iteration: 100* No. of Variables.
Max. Function Evaluation: 2000* No. of Variables.
Max. Time Limit: Inf.
Function Tolerance: 10¯⁶
4.3.3.3.2. Stopping Criteria:
Mesh Tolerance: 10¯⁶.
Max. Iteration: 100* No. of Variables.
Max. Function Evaluation: 2000* No. of Variables.
Max. Time Limit: ∞.
Function Tolerance: 10¯⁶.
0 500 1000 1500 2000 2500-1
-0.5
0
0.5
1
Iteration
Function v
alu
eBest Function Value: -1
1 2 3 4-200
0
200
400
600
800
1000
1200
1400Best point
Number of variables (4)
Best
poin
t
0 10 20 30 40 50 60 70 80 90 100
Time
Iteration
f-count
% of criteria met
Stopping Criteria
0 500 1000 1500 2000 2500-1
-0.5
0
0.5
1
Iteration
Function v
alu
e
Current Function Value: -0.99999
204
TABLE.4.5. Results of PS in 20 trails for office IEQ.
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 2 1800 70.5 1600 1 0.156075 60
2 2 1800 70.5 1600 1 0.148124 60
3 2 1800 70.5 1600 1 0.148314 60
4 2 1800 70.5 1600 1 0.14655 60
5 2 1800 70.5 1600 1 0.146723 60
6 2 1800 70.5 1600 1 0.150499 60
7 2 1800 70.5 1600 1 0.148128 60
8 2 1800 70.5 1600 1 0.147931 60
9 2 1800 70.5 1600 1 0.148575 60
10 2 1800 70.5 1600 1 0.147136 60
11 2 1800 70.5 1600 1 0.149625 60
12 2 1800 70.5 1600 1 0.146856 60
13 2 1800 70.5 1600 1 0.146429 60
14 2 1800 70.5 1600 1 0.146925 60
15 2 1800 70.5 1600 1 0.147402 60
16 2 1800 70.5 1600 1 0.147624 60
17 2 1800 70.5 1600 1 0.146341 60
18 2 1800 70.5 1600 1 0.14618 60
19 2 1800 70.5 1600 1 0.149296 60
20 2 1800 70.5 1600 1 0.149342 60
Avg 2 1800 70.5 1600 1 0.148204 60
Fig.4.3.Convergence of PS
0 10 20 30 40 50 60-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Iteration
Func
tion
valu
e
Best Function Value: -1
205
4.3.3.4. Particle swarm optimization
4.3.3.4.1. Options Set:
Max.Generation = 200.
Max. Time Limit=∞.
Average change in fitness value= 10-6
.
Time Limit = ∞.
Function Tolerance= 10-6.
Cognitive Attraction = 0.5.
Population Size = 40.
Social Attraction = 1.25.
4.3.3.4.2. Stopping Criteria:
Max.Generation = 200.
Max. Time Limit=∞.
Average change in fitness value= 10-6
Time Limit = ∞.
Function Tolerance= 10-6
0 10 20 30 40 50 600
200
400
600
800
1000
1200
Iteration
Mes
h si
ze
Current Mesh Size: 9.5367e-007
206
TABLE.4.6. Results of PSO in 20 trails for office IEQ
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 2 1528.9 61.2 1302.9 1 0.14490
4 57
2 2 1436.2 53.6 991.6 1 0.11841
2 73
3 -2 1769.3 45.2 945.7 1 0.08162
9 53
4 -2 1687.9 54.7 1191.7 1 0.09000
9 56
5 2 1335.3 60.5 1222.3 1 0.08634
6 53
6 2 1758.5 50.7 1192.7 1 0.08790
1 55
7 -1.9997 774.4371 49.8432 966.685
2 1 0.09591
1 59
8 2 1073.6 49.9 1359.2 1 0.08510
6 51
9 -2 1785.2 60.3 1181.9 1 0.08805
8 54
10 -2 1718.7 60.1 1148.9 1 0.07966
8 52
11 -2 1770.7 55.8 1139.2 1 0.10200
6 66
12 2 1753.2 65.1 1049.7 1 0.09481
4 62
13 -2 1777.2 61 1218.2 1 0.09750
2 63
14 -2 1585.4 66.4 1365.9 1 0.09300
3 53
15 2 1712.2 49.8 1128.5 1 0.09859
6 59
16 -2 1675.8 50 1152.2 1 0.08977 57
17 -2 1641.3 67.9 1459.9 1 0.08328 53
18 -2 1455.5 70.2 1466.1 1 0.08996
2 51
19 2 1706.3 58 1308 1 0.09390
9 58
20 2 1678 71.5 981.6 1 0.07920
3 51
avg -0.19999 1581.18185
5
58.0871
6
1188.64
4 1 0.09399
9 56.8
4.3.3.5. GODLIKE
4.3.3.5.1. Options Set & Stopping Criteria:
Max.FunEvals = 10-5
.
Max. Iterations= 20.
Min. Iterations = 2.
Total. Iterations = 15.
Function Tolerance = 10-4
207
TABLE.4.7. Results of GL in 20 trails for office IEQ
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 -2 1791.1 69.7 1565.3 1 1.546532 4
2 2 1799.9 58.7 1548.5 1 1.885316 4
3 -2 1799.9 70.6 1565.7 1 1.012803 4
4 2 1799.6 54 1560.7 1 1.505692 4
5 -2 1756.2 56.8 1380.8 1 1.191908 4
6 -2 1799.8 71.7 1580.9 1 1.099644 4
7 2 1799.5 58.7 1557.4 1 1.044684 4
8 2 1798.7 66.2 1592.6 1 1.15372 4
9 -2 1799.5 61.4 1491.2 1 1.076416 4
10 -2 1795.6 63 1541 1 1.219463 4
11 2 1799.5 65.8 1589 1 1.07033 4
12 -2 1799.1 53.2 1435.2 1 0.935505 4
13 -2 1799.1 55.6 1599 1 1.032483 4
14 2 1799.9 66.7 1599.2 1 1.285674 4
15 -2 1776.1 58.6 1473.2 1 0.933089 4
16 -2 1794.9 59.7 1164.1 1 0.953743 4
17 -2 1800 71.1 1556.6 1 1.379827 4
18 2 1769.3 48.3 1451.6 1 1.021947 4
19 2 1787.9 62.7 1446.4 1 1.07574 4
20 2 1780.6 70.9 1483.2 1 1.512711 4
avg -0.2 1792.31 62.17 1509.08 1 1.196861 4
4.3.3.6. Fmincon.
4.3.3.6.1. Options Set for ‘Fmincon’:
Max.Iterations:400.
Max.function Evaluations: 100* No. of Variables.
Max.Time: ∞.
Max. Function Tolerance: 10-6
.
4.3.3.6.2. Stopping Criteria for Global Search:
Max.Time: ∞.
Max Wait cycle: 20
208
4.3.3.6.3. Stopping Criteria for Fmincon:
Max.Iterations > 400.
Function Tolerance: 10-6
TABLE.4.8. Results of Fmincon in 20 trails for office IEQ.
Trials
no.
Therma
l CO2 Soun
d
Illu
m IEQ Time Func.Coun No.Local.Solver
s 1 2 180
0 72 1600 1 8.86601
9 5002 198
2 2 180
0 72 1600 1 13.4064
1 7277 334
3 2 180
0 72 1600 1 14.7796 7042 362
4 2 180
0 72 1600 1 12.6790
6 6897 324
5 2 180
0 72 1600 1 7.11649
2 4067 144
6 2 180
0 72 1600 1 11.6128
6 5507 266
7 2 180
0 72 1600 1 7.61832 5312 160
8 2 180
0 72 1600 1 6.78551
4 4672 184
9 2 180
0 72 1600 1 8.38266
8 4847 204
10 2 180
0 72 1600 1 13.3524
9 6477 373
11 2 180
0 72 1600 1 9.97782
8 4662 270
12 2 180
0 72 1600 1 11.5853
1 6032 253
13 2 180
0 72 1600 1 9.27154
5 5712 199
14 2 180
0 72 1600 1 8.60228
6 4612 199
15 2 180
0 72 1600 1 12.3196
6 6592 344
16 2 180
0 72 1600 1 14.4853
1 5987 344
17 2 180
0 72 1600 1 10.1845 4567 213
18 2 180
0 72 1600 1 7.90361
3 4912 167
19 2 180
0 72 1600 1 13.6878
9 6077 396
20 2 180
0 72 1600 1 12.0780
2 6227 236
Avg 2 180
0 72 1600 1 10.7347
7 5624 258.5
4.3.3.7. Direct evolution
4.3.3.7.1. Options Set:
Min. Value to Reach = 10-6
.
Population Size = 10*D.
Max. Iterations = 200.
Step Size F = 0.8.
Cross Over Probability = 0.5.
Strategy= 7 (DE/rand/1/bin)
209
DE/x/y/z, where DE stands for DE, x represents a string denoting the vector to be
perturbed, y is the number of difference vectors considered for perturbation of x, and
z stands for the type of crossover being used (exp: exponential; bin: binomial).
4.3.3.7.2. Stopping Criteria:
Max.Value of function reached= 10-6
.
Max.Iterations=200
TABLE.4.9. Results of DE in 20 trails for office IEQ
Trails
no. Thermal CO2 Sound Illum IEQ Time Iterations
1 1.8902 858.2017 45.1112 555.2472 1 0.00303 40
2 2 1060.1 60.6 865.3 1 0.003735 40
3 -2 1729.5 48.6 1302.6 1 0.002558 40
4 -1.9906 897.3013 53.8414 790.6261 1 0.01054 40
5 -2 1583 58.2 877.1 1 0.003255 40
6 1.9 1234.7 54.9 550.2 1 0.004085 40
7 -2 1305.5 66.3 1110.2 1 0.00335 40
8 1.9833 690.9403 48.5607 478.5444 1 0.001541 40
9 2 1652.5 59.7 1587.9 1 0.001507 40
10 1.9283 725.1315 46.8846 491.9793 1 0.003296 40
11 -1.9913 649.2797 63.5242 584.1255 1 0.002578 40
12 2 1251.8 64.7 1380.1 1 0.002499 40
13 1.9 1020.9 71.7 1450.5 1 0.00378 40
14 1.9933 970.858 47.7356 803.1949 1 0.003476 40
15 -2 1697.5 52.6 1197.2 1 0.00366 40
16 -1.9362 708.0032 65.3522 799.9138 1 0.00321 40
17 -2 1610.8 60.2 1566.2 1 0.002003 40
18 2 1770.6 59.1 957.6 1 0.003846 40
19 2 1225.4 45.9 1537 1 0.003841 40
20 2 1152.7 65.8 1581.2 1 0.001516 40
Avg 0.38385 1189.735785 56.9655 1023.337 1 0.003365 40
4.3.3.8. LGO
4.3.3.8.1. Stopping Criteria:
If the current best solution did not improve for
Program execution time limits > 600 seconds.
210
4.3.3.8.2. Local search termination criterion parameter:
First local search phase ends, if the function difference is less than
If max. constrain violation exceeds
TABLE.4.10. Results of LGO in 20 trails for office IEQ
Trials
no.
Therma
l CO2 Sound Illum IEQ Time Func.Coun
1 2 1759.80654
8
68.7651
6
1561.97
1 1 0.91222
6 2699
2 2 1759.80654
8
68.7651
6
1561.97
1 1 0.73261
5 2699
3 2 1759.80654
8
68.7651
6
1561.97
1 1 0.73121
9 2699
4 2 1759.80654
8
68.7651
6
1561.97
1 1 0.72443
2 2699
5 2 1759.80654
8
68.7651
6
1561.97
1 1 0.71583
8 2699
6 2 1759.80654
8
68.7651
6
1561.97
1 1 0.75345
1 2699
7 2 1759.80654
8
68.7651
6
1561.97
1 1 0.73637
8 2699
8 2 1759.80654
8
68.7651
6
1561.97
1 1 0.74391
7 2699
9 2 1759.80654
8
68.7651
6
1561.97
1 1 0.74018
2 2699
10 2 1759.80654
8
68.7651
6
1561.97
1 1 0.96158
5 2699
11 2 1759.80654
8
68.7651
6
1561.97
1 1 0.71396
3 2699
12 2 1759.80654
8
68.7651
6
1561.97
1 1 0.94339
1 2699
13 2 1759.80654
8
68.7651
6
1561.97
1 1 0.97151
4 2699
14 2 1759.80654
8
68.7651
6
1561.97
1 1 0.96947
1 2699
15 2 1759.80654
8
68.7651
6
1561.97
1 1 1.00635 2699
16 2 1759.80654
8
68.7651
6
1561.97
1 1 0.96290
8 2699
17 2 1759.80654
8
68.7651
6
1561.97
1 1 0.74592 2699
18 2 1759.80654
8
68.7651
6
1561.97
1 1 0.983 2699
19 2 1759.80654
8
68.7651
6
1561.97
1 1 0.811 2699
20 2 1759.80654
8
68.7651
6
1561.97
1 1 0.70414
6 2699
avg 2 1759.80654
8
68.7651
6
1561.97
1 1 0.82817
5 2699
4.3.3.9. glcCluster
4.3.3.9.1. Stopping Criteria:
Maximum Iterations = 10000;
Maximum Function count = 10000;
Tolerance of Variables = 10-5
Function Tolerance =10-7
211
TABLE.4.11. Results of glcCluster in 20 trails for office IEQ
Trials
No. Thermal CO2 Soun
d Illum IEQ Time Iteration
s
Func.Coun
t 1 1.987562 1150.00001
8 49.5 433.33
36 1 0.59777
8 1 1515
2 1.987562 1150.00001
8 49.5 433.33
36 1 0.71545
5 1 1511
3 1.987562 1150.00001
8 49.5 433.33
36 1 0.61269
5 1 1514
4 1.987562 1150.00001
8 49.5 433.33
36 1 0.55204
4 1 1512
5 1.987562 1150.00001
8 49.5 433.33
36 1 0.72806
2 1 1513
6 1.987562 1150.00001
8 49.5 433.33
36 1 0.63516
5 1 1511
7 1.987562 1150.00001
8 49.5 433.33
36 1 0.62625 1 1513
8 1.987562 1150.00001
8 49.5 433.33
36 1 0.56273
4 1 1511
9 1.987562 1150.00001
8 49.5 433.33
36 1 0.59834
3 1 1513
10 1.987562 1150.00001
8 49.5 433.33
36 1 0.73717
1 1 1511
11 1.987562 1150.00001
8 49.5 433.33
36 1 0.58418
5 1 1516
12 1.987562 1150.00001
8 49.5 433.33
36 1 0.61178
3 1 1513
13 1.987562 1150.00001
8 49.5 433.33
36 1 0.56361
1 1 1513
14 1.987562 1150.00001
8 49.5 433.33
36 1 0.58698
7 1 1513
15 1.987562 1150.00001
8 49.5 433.33
36 1 0.57058
9 1 1511
16 1.987562 1150.00001
8 49.5 433.33
36 1 0.66160
9 1 1513
17 1.987562 1150.00001
8 49.5 433.33
36 1 0.60496
7 1 1514
18 1.987562 1150.00001
8 49.5 433.33
36 1 0.63104 1 1515
19 1.987562 1150.00001
8 49.5 433.33
36 1 0.60053
5 1 1515
20 1.987562 1150.00001
8 49.5 433.33
36 1 0.56090
3 1 1513
Avg 1.987562 1150.00001
8 49.5 433.33
36 1 0.61709
5 1 1513
4.3.3.10. glcSolve
4.3.3.10.1. Stopping Criteria:
Max.Iterations is exceeded > No. of variables*1000.
Max.function evaluations > No. of variables*2000.
If the difference of objective function is < 10-6
212
Table.4.12 Results of glcSolve in 20 trails for office IEQ
Trail
s
No.
Thermal Co2 Soun
d Illum IEQ Time Iter Func.Coun
t 1 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.93305
3 222 1503
2 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.29115
4 222 1503
3 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.87620
1 222 1503
4 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.23000
4 222 1503
5 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.99029
7 222 1503
6 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.90531
8 222 1503
7 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.94392
1 222 1503
8 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.95278
3 222 1503
9 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.09855 222 1503
10 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.95906
5 222 1503
11 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.10017
7 222 1503
12 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.01511
2 222 1503
13 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.90944
4 222 1503
14 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.96646
6 222 1503
15 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.01381
2 222 1503
16 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.01043
5 222 1503
17 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.09151
1 222 1503
18 -1.92593 1727.77777
8 70.5 1522.22
2 1 1.19118
3 222 1503
19 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.86412 222 1503
20 -1.92593 1727.77777
8 70.5 1522.22
2 1 0.92683 222 1503
Avg -1.92593 1727.77777
8 70.5 1522.22
2 1 1.01347
2 222 1503
Table.4.13. Comparative results of optimization methods for office IEQ
Method Thermal
sensation
Carbon
dioxide
Sound
level
Horizontal
illumination
IEQ TIME iterations
GA 1.785 1613.71 63.695 1381.06 1 1.02834 51
SA 0.2 1272.77 60.715 962.05 1 2.49694 2010.05
PS 2 1800 70.5 1600 1 0.1482 60
PSO -0.2 1581.18 58.0872 1188.64 1 0.094 56.8
GL -0.2 1792.31 62.17 1509.08 1 1.19686 4
Fmincon 2 1800 72 1600 1 10.73477 5624
DE 0.38385 1189.74 56.9655 1023.34 1 0.00337 40
LGO 2 1759.81 68.7652 1561.97 1 0.82818 2699
glcCluster 1.98756 1150 49.5 433.334 1 0.6171 1513
glcSolve -1.9259 1727.78 70.5 1522.22 1 1.01347 1503
213
Fig.4.4. Comparative graph for office Indoor Environmental Quality
4.3.4. Discussion of the comparative results
All the ten methods yield the IEQ value as 1 which is the acceptable optimum value.
The time taken by Fmincon is maximum. Number of iterations was maximum for DE and
minimum for GL. Carbon dioxide, sound and illumination level are more or less the same for
all 10 methods. The thermal sensation alone was different for different methods
0.001
0.01
0.1
1
10
100
1000
10000
GA
SA
PS
PSO
GL
Fmincon
DE
LGO
glcCluster
glcSolve
214
4.3.5. PARAMETERS.
4.3.5.1. THERMAL SENSATION
Thermal comfort is that condition of mind which expresses satisfaction with the
thermal environment. Thermal environment encompasses characteristics which affect a
person's heat loss. In terms of bodily sensations, thermal comfort is a sensation of hot, warm,
slightly warmer, neutral, slightly cooler, cool and cold. From the physiological point of view,
thermal comfort occurs when there is a thermal equilibrium in the absence of regulatory
sweating between the heat exchange between the human body and the environment.
TABLE.4.14. Thermal sensation results in all 10 methods
Trial no. GA SA PS PSO GL DE NL LGO glcCluster glcSolve
1 2 2 2 2 -2 1.8902 2 2 -1.92592593 -1.92592593
2 2 -2 2 2 2 2 2 2 -1.92592593 -1.92592593
3 2 2 2 -2 -2 -2 2 2 -1.92592593 -1.92592593
4 2 2 2 -2 2 -1.9906 2 2 -1.92592593 -1.92592593
5 1.9 -2 2 2 -2 -2 2 2 -1.92592593 -1.92592593
6 2 2 2 2 -2 1.9 2 2 -1.92592593 -1.92592593
7 2 -2 2 -1.9997 2 -2 2 2 -1.92592593 -1.92592593
8 2 -2 2 2 2 1.9833 2 2 -1.92592593 -1.92592593
9 2 2 2 -2 -2 2 2 2 -1.92592593 -1.92592593
10 2 -2 2 -2 -2 1.9283 2 2 -1.92592593 -1.92592593
11 2 -2 2 -2 2 -1.9913 2 2 -1.92592593 -1.92592593
12 2 2 2 2 -2 2 2 2 -1.92592593 -1.92592593
13 2 2 2 -2 -2 1.9 2 2 -1.92592593 -1.92592593
14 2 -2 2 -2 2 1.9933 2 2 -1.92592593 -1.92592593
15 1.9 -2 2 2 -2 -2 2 2 -1.92592593 -1.92592593
16 2 2 2 -2 -2 -1.9362 2 2 -1.92592593 -1.92592593
17 2 2 2 -2 -2 -2 2 2 -1.92592593 -1.92592593
18 1.9 -2 2 -2 2 2 2 2 -1.92592593 -1.92592593
19 -2 2 2 2 2 2 2 2 -1.92592593 -1.92592593
20 2 2 2 2 2 2 2 2 -1.92592593 -1.92592593
AVG 0.2 2 -0.19999 -0.2 0.38385 2 2 -1.92592593 -1.92592593
215
FIG.4.5. Graph of Thermal sensation results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
-202
TRAILS
GA
-202
SA
123
PS
-202
PS
O
-202
GL
-202
DE
123
NL
123
LG
O
-2
glc
Clu
-2
glc
So
l
From the graph we can see that Pattern search, Fmincon, LGO, glcCluster and
glcSolve are consistent in all 20 trials.
216
4.3.5.2. CARBON DI-OXIDE
Carbon dioxide (CO2) is the chief greenhouse gas that results from human activities
and causes global warming and climate change. Though carbon dioxide is not toxic in itself,
the amount found in the indoor environment is used as an indicator for human comfort.
Elevated levels of carbon dioxide indicate that an insufficient amount of fresh, outdoor air is
being delivered to the occupied areas of the building. This also indicates that other pollutants
in the building may exist at elevated levels since there is not enough fresh air to dilute them.
Since carbon dioxide is an unavoidable, predictable, and easily measured product of human
occupancy, it is used as a marker for other pollutants emanating from humans or other
sources in the building. However, carbon dioxide is mostly a threat to health, when the
concentration is high enough to displace the oxygen, which can lead to suffocation in a
confined space.
Table.4.15. Carbon Dioxide results in all 10 methods
Tria
no. GA SA PS PSO GL DE NL LGO GLCc GLCs
1 1799.9 1379.6 1800 1528.
9 1791.1 858.20
17 1800 1759.806
548
1150.000
018
1727.77
778 2 1740.6 1213.4 1800 1436.
2 1799.9 1060.1 1800 1759.806
548 1150.000
018 1727.77
778 3 1799.7 1513.7 1800 1769.
3 1799.9 1729.5 1800 1759.806
548 1150.000
018 1727.77
778 4 1070.5 1160.6 1800 1687.
9 1799.6 897.30
13 1800 1759.806
548
1150.000
018
1727.77
778 5 1662.3 1341.8 1800 1335.
3 1756.2 1583 1800 1759.806
548
1150.000
018
1727.77
778 6 1713.5 1467.3 1800 1758.
5 1799.8 1234.7 1800 1759.806
548
1150.000
018
1727.77
778 7 1217.2 1318.7 1800 774.4
371 1799.5 1305.5 1800 1759.806
548
1150.000
018
1727.77
778 8 1775.7 1219.8 1800 1073.
6 1798.7 690.94
03 1800 1759.806
548 1150.000
018 1727.77
778 9 1724.6 1127 1800 1785.
2 1799.5 1652.5 1800 1759.806
548 1150.000
018 1727.77
778 10 1771 1486.7 1800 1718.
7 1795.6 725.13
15 1800 1759.806
548
1150.000
018
1727.77
778 11 1295.4 982.3 1800 1770.
7 1799.5 649.27
97 1800 1759.806
548
1150.000
018
1727.77
778 12 1668.8 1330.5 1800 1753.
2 1799.1 1251.8 1800 1759.806
548
1150.000
018
1727.77
778 13 1549.2 1041.9 1800 1777.
2 1799.1 1020.9 1800 1759.806
548
1150.000
018
1727.77
778 14 1510.7 1284.2 1800 1585.
4 1799.9 970.85
8 1800 1759.806
548 1150.000
018 1727.77
778 15 1799.9 1303.7 1800 1712.
2 1776.1 1697.5 1800 1759.806
548 1150.000
018 1727.77
778 16 1722.7 1174.6 1800 1675.
8 1794.9 708.00
32 1800 1759.806
548
1150.000
018
1727.77
778 17 1611.4 1410.5 1800 1641.
3 1800 1610.8 1800 1759.806
548
1150.000
018
1727.77
778 18 1756.2 1313 1800 1455.
5 1769.3 1770.6 1800 1759.806
548
1150.000
018
1727.77
778 19 1789.5 1228.1 1800 1706.
3 1787.9 1225.4 1800 1759.806
548
1150.000
018
1727.77
778 20 1295.4 1157.9 1800 1678 1780.6 1152.7 1800 1759.806
548 1150.000
018 1727.77
778 AVG
1613.7
1272.7 1800 1581.182
1792.31
1189.736
1800 1759.806548
1150.000018
1727.77778
217
FIG.4.6. Graph for Carbon Dioxide concentration results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
1000
1500
TRAILS
GA
1000
1500
SA
1800PS
10001500
PSO
176017801800
GL
50010001500
DE
1800NL
1600
1800
LGO
11001200
glcC
lu
1600
1800
glcS
ol
From the graph we can see that Pattern search, Fmincon, LGO, glcCluster and
glcSolve are consistent in all 20 trials.
218
4.3.5.3. SOUND PRESSURE LEVEL
Acoustics is the interdisciplinary science that deals with the study of all mechanical
waves in gases, liquids, and solids including vibration, sound, ultrasound and infrasound. The
perception of sound in any organism is limited to a certain range of frequencies. Hearing loss
due to prolonged exposure to noise is well documented. Excessive noise also has an adverse
effect on personal health and wellbeing, ability to perform quiet tasks, and productivity in
general. Because land is becoming scarcer, buildings are being constructed closer together
and closer to noise sources such as highways, railways, and airports. As a result, sound or
acoustic control is becoming increasingly important. The reduction of airborne sound through
a wall is called sound transmission loss (STL).
TABLE.4.16. Sound Pressure level results in all 10 methods
Trial
no.
GA SA PS PSO GL DE NL LGO GLCc GLCs
1 59.6 68.4 70.5 61.2 69.7 45.1112 72 68.76515
98
49.500000
9 70.5
2 62.8 70.6 70.5 53.6 58.7 60.6 72 68.76515
98
49.500000
9 70.5
3 71.7 50.5 70.5 45.2 70.6 48.6 72 68.76515
98
49.500000
9 70.5
4 64.3 64.2 70.5 54.7 54 53.8414 72 68.76515
98
49.500000
9 70.5
5 57 72 70.5 60.5 56.8 58.2 72 68.7651598
49.5000009
70.5
6 55.7 53.6 70.5 50.7 71.7 54.9 72 68.7651598
49.5000009
70.5
7 60.2 62.4 70.5 49.843 58.7 66.3 72 68.76515
98
49.500000
9 70.5
8 69.7 58.4 70.5 49.9 66.2 48.5607 72 68.76515
98
49.500000
9 70.5
9 57.4 71.3 70.5 60.3 61.4 59.7 72 68.76515
98
49.500000
9 70.5
10 63.8 56.2 70.5 60.1 63 46.8846 72 68.76515
98
49.500000
9 70.5
11 71.3 64.3 70.5 55.8 65.8 63.5242 72 68.7651598
49.5000009
70.5
12 60.2 45.3 70.5 65.1 53.2 64.7 72 68.7651598
49.5000009
70.5
13 71.4 61.3 70.5 61 55.6 71.7 72 68.76515
98
49.500000
9 70.5
14 52.7 71.5 70.5 66.4 66.7 47.7356 72 68.76515
98
49.500000
9 70.5
15 69 45.5 70.5 49.8 58.6 52.6 72 68.76515
98
49.500000
9 70.5
16 65.6 46.5 70.5 50 59.7 65.3522 72 68.76515
98
49.500000
9 70.5
17 63.8 52.1 70.5 67.9 71.1 60.2 72 68.7651598
49.5000009
70.5
18 54.5 68.8 70.5 70.2 48.3 59.1 72 68.7651598
49.5000009
70.5
19 71.9 71.2 70.5 58 62.7 45.9 72 68.76515
98
49.500000
9 70.5
20 71.3 60.2 70.5 71.5 70.9 65.8 72 68.76515
98
49.500000
9 70.5
AV
G 63.695 60.7
15 70.5 58.087
16 62.17 56.9655 72 68.76515
98
49.500000
9 70.5
219
FIG.4.7. Graph for Sound Pressure level results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
6070
TRAILS
GA
506070
SA
657075
PS
506070
PS
O
506070
506070
DE
GL
65707580
NL
657075
LG
O
455055
glc
Clu
657075
glc
So
l
From the graph we can see that Pattern search, Fmincon, LGO, glcCluster and
glcSolve are consistent in all 20 trials.
220
4.3.5.4. HORIZONTAL ILLUMINATION
Lighting or illumination is the deliberate application of light to achieve some aesthetic
or practical effect. In some design instances, materials used on walls and furniture play a key
role in the lighting effect. Surfaces or floors that are too reflective create unwanted glare.
Specification of illumination requirements is the basic concept of deciding how
much illumination is required for a given task. Clearly, much less light is required to
illuminate a hallway or bathroom compared to that needed for a word processing work
station. Generally speaking, the energy expended is proportional to the design illumination
level. Beyond the energy factors being considered, it is important not to over-design
illumination, lest adverse health effects such as headache frequency, stress, and
increased blood pressure be induced by the higher lighting levels. In addition, glare or excess
light can decrease worker efficiency.
TABLE.4.17.Horizontal illumination results in all 10 methods
Trial
no. GA SA PS PSO GL DE NL LGO GLCc GLCs
1 1553.
7
1014
.9 1600 1302.
9
1565.
3
555.2
472 1600 1561.970
895
433.3335
754
1522.22
222 2 1370.
9 1125
.6 1600 991.6 1548.
5 865.3 1600 1561.970
895 433.3335
754 1522.22
222 3 1569.
3 970.
4 1600 945.7 1565.
7 1302.
6 1600 1561.970
895 433.3335
754 1522.22
222 4 1196.
7
846.
1 1600 1191.
7
1560.
7
790.6
261 1600 1561.970
895
433.3335
754
1522.22
222 5 1513.
8
746.
3 1600 1222.
3
1380.
8 877.1 1600 1561.970
895
433.3335
754
1522.22
222 6 1520.
9
1100
.5 1600 1192.
7
1580.
9 550.2 1600 1561.970
895
433.3335
754
1522.22
222 7 1524.
2
1029
.3 1600 966.6
852
1557.
4
1110.
2 1600 1561.970
895
433.3335
754
1522.22
222 8 1395.
2 878.
4 1600 1359.
2 1592.
6 478.5444
1600 1561.970895
433.3335754
1522.22222
9 1503.8
959.8
1600 1181.9
1491.2
1587.9
1600 1561.970895
433.3335754
1522.22222
10 1018.
5
1042
.4 1600 1148.
9 1541 491.9
793 1600 1561.970
895
433.3335
754
1522.22
222 11 1440.
9
1116
.7 1600 1139.
2 1589 584.1
255 1600 1561.970
895
433.3335
754
1522.22
222 12 1434 789.
4 1600 1049.
7
1435.
2
1380.
1 1600 1561.970
895
433.3335
754
1522.22
222 13 1232.
4
854.
1 1600 1218.
2 1599 1450.
5 1600 1561.970
895
433.3335
754
1522.22
222 14 792.1 964.
4 1600 1365.
9 1599.
2 803.1949
1600 1561.970895
433.3335754
1522.22222
15 1538.8
791.1
1600 1128.5
1473.2
1197.2
1600 1561.970895
433.3335754
1522.22222
16 1495 961.
3 1600 1152.
2
1164.
1
799.9
138 1600 1561.970
895
433.3335
754
1522.22
222 17 1265.
3
1145
.8 1600 1459.
9
1556.
6
1566.
2 1600 1561.970
895
433.3335
754
1522.22
222 18 1234.
4
1081
.6 1600 1466.
1
1451.
6 957.6 1600 1561.970
895
433.3335
754
1522.22
222 19 1580.
4
828.
1 1600 1308 1446.
4 1537 1600 1561.970
895
433.3335
754
1522.22
222 20 1440.
9 994.
8 1600 981.6 1483.
2 1581.
2 1600 1561.970
895 433.3335
754 1522.22
222 AVG 1381.
06 962.05
1600 1188.644
1509.08
1023.337
1600 1561.970895
433.3335754
1522.22222
221
FIG.4.8.Graph for horizontal illumination results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
1000
1500
TRAILS
GA
80010001200
SA
1600
PS
100012001400
PS
O
120014001600
GL
50010001500
DE
1600NL
1600
LG
O
400
450
glc
Clu
1400
1600g
lcS
ol
From the graph we can see that Pattern search, Fmincon, LGO, glcCluster and
glcSolve are consistent in all 20 trials.
222
4.3.5.5. IEQ.
TABLE.4.18. IEQ results in all 10 methods
Trial no. GA SA PS PSO GL DE NL LGO GLCc GLCs
1 1 1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
3 1 1 1 1 1 1 1 1 1 1
4 1 1 1 1 1 1 1 1 1 1
5 1 1 1 1 1 1 1 1 1 1
6 1 1 1 1 1 1 1 1 1 1
7 1 1 1 1 1 1 1 1 1 1
8 1 1 1 1 1 1 1 1 1 1
9 1 1 1 1 1 1 1 1 1 1
10 1 1 1 1 1 1 1 1 1 1
11 1 1 1 1 1 1 1 1 1 1
12 1 1 1 1 1 1 1 1 1 1
13 1 1 1 1 1 1 1 1 1 1
14 1 1 1 1 1 1 1 1 1 1
15 1 1 1 1 1 1 1 1 1 1
16 1 1 1 1 1 1 1 1 1 1
17 1 1 1 1 1 1 1 1 1 1
18 1 1 1 1 1 1 1 1 1 1
19 1 1 1 1 1 1 1 1 1 1
20 1 1 1 1 1 1 1 1 1 1
AVG 1 1 1 1 1 1 1 1 1 1
223
FIG.4.9. Graph for IEQ results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
0
2
GA
TRAILS
0
2
SA
0
2
PS
0
2
PS
O
0
2
GL
0
2
DE
0
2
NL
0
2
LG
O
0
2
glc
Clu
0
2g
lcS
ol
From the graph we can see that all ten methods are consistent in all 20 trials with
value 1 which is the optimum acceptable value.
224
4.3.5.6. ELAPSED TIME
CPU time is the time for which the CPU was busy executing the task. It does not take
into account the time spent in waiting for I/O (disk IO or network IO). Since I/O operations,
such as reading files from disk, are performed by the OS, these operations may involve
noticeable amount of time in waiting for I/O subsystems to complete their operations. This
waiting time will be included in the elapsed time, but not CPU time. Hence CPU time is
usually less than the elapsed time.
TABLE.4.19. Elapsed time results in all 10 methods
Trail GA SA PS PSO GL DE NL LGO GLCc GLCs
1 1.208
193
2.5824
72
0.156
075
0.144
904
1.546
532
0.003
03
8.86601
9
0.912
226
0.597
778
0.933
053 2 1.133
192
2.2277
51
0.148
124
0.118
412
1.885
316
0.003
735
13.4064
07
0.732
615
0.715
455
1.291
154 3 0.775
846
2.3222
16
0.148
314
0.081
629
1.012
803
0.002
558 14.7796 0.731
219
0.612
695
0.876
201 4 1.089
181
2.4827
02
0.146
55
0.090
009
1.505
692
0.010
54
12.6790
64
0.724
432
0.552
044
1.230
004 5 1.059
298
1.9079
2
0.146
723
0.086
346
1.191
908
0.003
255
7.11649
2
0.715
838
0.728
062
0.990
297 6 1.140
209
3.0296
57
0.150
499
0.087
901
1.099
644
0.004
085
11.6128
6
0.753
451
0.635
165
0.905
318 7 0.768
988
2.8660
03
0.148
128
0.095
911
1.044
684
0.003
35 7.61832 0.736
378
0.626
25
0.943
921 8 0.919
29
2.2797
64
0.147
931
0.085
106
1.153
72
0.001
541
6.78551
4
0.743
917
0.562
734
0.952
783 9 1.172
938
2.0897
66
0.148
575
0.088
058
1.076
416
0.001
507
8.38266
8
0.740
182
0.598
343
1.098
55 10 0.917
832
2.1340
44
0.147
136
0.079
668
1.219
463
0.003
296
13.3524
94
0.961
585
0.737
171
0.959
065 11 0.913
941
2.9115
1
0.149
625
0.102
006
1.070
33
0.002
578
9.97782
8
0.713
963
0.584
185
1.100
177 12 1.169
327
3.0143
31
0.146
856
0.094
814
0.935
505
0.002
499
11.5853
06
0.943
391
0.611
783
1.015
112 13 1.059
507
2.2839
27
0.146
429
0.097
502
1.032
483
0.003
78
9.27154
5
0.971
514
0.563
611
0.909
444 14 0.887
249
2.4240
2
0.146
925
0.093
003
1.285
674
0.003
476
8.60228
6
0.969
471
0.586
987
0.966
466 15 1.115
581
3.0542
13
0.147
402
0.098
596
0.933
089
0.003
66
12.3196
61
1.006
35
0.570
589
1.013
812 16 1.195
485
2.0850
67
0.147
624
0.089
77
0.953
743
0.003
21
14.4853
07
0.962
908
0.661
609
1.010
435 17 1.032
374
2.5689
22
0.146
341
0.083
28
1.379
827
0.002
003
10.1844
98
0.745
92
0.604
967
1.091
511 18 0.932
092
2.9044
54
0.146
18
0.089
962
1.021
947
0.003
846
7.90361
3 0.983 0.631
04
1.191
183 19 1.162
234
2.8653
76
0.149
296
0.093
909
1.075
74
0.003
841
13.6878
91 0.811 0.600
535
0.864
12 20 0.913
941
1.9046
84
0.149
342
0.079
203
1.512
711
0.001
516
12.0780
18
0.704
146
0.560
903
0.926
83 AVG
VG
1.028
335
2.4969
3995
0.148
204
0.093
999
1.196
861
0.003
365
10.7347
6955
0.828
1753
0.617
0953
1.013
4718
225
FIG.4.10.Graph for Elapsed time results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
0.81.01.2
TRAILS
GA
2.02.53.0
SA
0.1450.1500.155
PS
0.10
0.15
PS
O
1.01.52.0
GL
0.0000.0050.010
DE
10
15
NL
0.8
1.0
LG
O
0.60.7
glc
Clu
0.81.01.2
glc
Sol
From the graph, we can see that none of the methods was consistent in all 20 trials but
Pattern search method alone got a more or less same value – average of 0.148 seconds.
Which is second best in terms of time. Therefore it is the best method.
226
4.3.5.7. ITERATIONS
In a computational procedure, a cycle of operations is repeated, often to approximate
the desired result more closely. Iteration means the act of repeating a process usually with the
aim of approaching a desired goal or target or result. Iteration in computing is the repetition
of a process within a computer program. It may also refer to the process of iterating a
function i.e. applying a function repeatedly, using the output from one iteration as the input to
the next. Another use of iteration in mathematics is in iterative methods which are used to
produce approximate numerical solutions to certain mathematical problems.
TABLE.4.20. Iterations results in all 10 methods
Trial no. GA SA PS PSO GL DE NL LGO GLCc GLCs
1 51 2004 60 57 4 40 5002 2699 1515 1503
2 51 2003 60 73 4 40 7277 2699 1511 1503
3 51 2001 60 53 4 40 7042 2699 1514 1503
4 51 2006 60 56 4 40 6897 2699 1512 1503
5 51 2005 60 53 4 40 4067 2699 1513 1503
6 51 2017 60 55 4 40 5507 2699 1511 1503
7 51 2031 60 59 4 40 5312 2699 1513 1503
8 51 2005 60 51 4 40 4672 2699 1511 1503
9 51 2007 60 54 4 40 4847 2699 1513 1503
10 51 2006 60 52 4 40 6477 2699 1511 1503
11 51 2011 60 66 4 40 4662 2699 1516 1503
12 51 2002 60 62 4 40 6032 2699 1513 1503
13 51 2013 60 63 4 40 5712 2699 1513 1503
14 51 2007 60 53 4 40 4612 2699 1513 1503
15 51 2007 60 59 4 40 6592 2699 1511 1503
16 51 2015 60 57 4 40 5987 2699 1513 1503
17 51 2016 60 53 4 40 4567 2699 1514 1503
18 51 2004 60 51 4 40 4912 2699 1515 1503
19 51 2016 60 58 4 40 6077 2699 1515 1503
20 51 2025 60 51 4 40 6227 2699 1513 1503
AVG 51 2010.05 60 56.8 4 40 5624 2699 1513 1503
227
FIG.4.11. Graph for Iterations results in all 10 methods
0 2 4 6 8 10 12 14 16 18 20 22
455055
TRAILS
GA
2000
2020
SA
556065
PS
506070
PS
O
345
GL
35
40
45
DE
4000
6000
NL
26002800
LGO
1510
1515
glcC
lu1400
1600gl
cSol
From the graph we can see that GA, Pattern search, GL, DE, LGO and glcSolve are
consistent in all 20 trials.
228
Table 4.21- Comparative table for parameters in all 10 methods
Variables GA SA PS PSO GL fmincon DE LGO Glc
Cluster
Glc
Solve
PMV X X 2
X X 2
X 2
-1.9
-1.9
CO2 X X 1800
X X 1800
X 1759
1150
1727.7
Sound X X 70.5
X X 72
X 68.76
49.5
70
Illumina tion
X X 1600
X X 1600
X 1561.9
433.3
1522.2
IEQ 1
1
1
1
1
1
1
1
1
1
TIME 0.148 0.09
3
0.003
ITERS X 60
X 4
40
X X
- Represents the parameters which are consistent for all the 20 trials and the
corresponding parameter values are given in the respective cell.
X - Represents the parameters which are not consistent for all the 20 trials
In case of iterations and elapsed time only the two or three minimum values alone are given.
4.3.6. Results and Discussion
With the two extreme values of parameters from survey, the optimization is carried
out with different solvers. As they are of stochastic type, their results may vary from trial to
trial and so the problem is made to run for 20 trials (Elbeltagi, Tarek Hegazy, & & Grierson,
2005) and an average of all trials is taken as the final value of the parameter, by the solver.
The solvers are compared with three different criteria
1. Consistency
The consistency table gives the parameters that remain constant for all the
trails. All the solvers give the same value of IEQ for all the runs, which in turn
indicates that the quality requirements are acceptable.
Thermal - P.S (2), NL(2), LGO (2),glcSolve (-1.9), glcCluster (-1.9)
CO2 - P.S (1800), NL(1800), LGO (1759),glcSolve(1727.7), glcCluster
229
(1150)
Sound - P.S(70.5), NL(72), LGO (68.76),glcSolve (70), glcCluster (49.5)
Illumination -P.S (1600), NL(1600), LGO (1561.91),glcSolve (1522.2),
glcCluster (433.3)
So we see that the solvers Pattern Search,Fmincon, glcSolve, glcCluster&
LGO remain constant throughout their runs.
2. Minimum Run Time
For a minimum run time of the problem we got PS (0.093 seconds), Pattern
Search (0.148 seconds), DE (0.003 seconds).
3. Minimum Evaluation
This criterion will determine the effectiveness of the algorithm. From the
result table, we see that the Pattern Search, GODLIKE and DE algorithms have
minimum evaluation of 60 , 4 and 40 respectively
4. Simplicity of Algorithm
Of all the algorithms we have taken the Pattern Search algorithm is the
simplest followed by GA, PSO, DE, Simulated Annealing, GODLIKE, Non-
Linear, and Direct algorithm.
5. Results according to Standards
This is the most important criterion that determines whether the solver is
practical or not. We got the standard values from ASHRAE, IES, Guidance for
employers on the Control of Noise at Work Regulations 2005 as:
Thermal comfort: -3 to 3
Carbon dioxide: less than 1000ppm
Sound level: 40 dBA to 70dBA
Illumination level: 1000 lux to 2000 lux
With the above standards the solvers which adhere to the standard are:
Thermal comfort: GA, SA, PS, PSO, FMINCON, DE, GL,LGO,
glcCluster, glcSolve
230
Carbon dioxide: GA, SA, PS, PSO, FMINCON,GL, LGO, glcCluster,
glcSolve
Sound level: GA, SA, PS, PSO, FMINCON, DE, GL, LGO,glcCluster,
glcSolve
Illumination level: DE, glcCluster, glcSolve
The following table gives a summary of all the criteria for the solvers:
Table.4.22. Summary of all the criteria for the solvers
Criteria GA SA PS PSO Fmincon DE GL LGO glcClus glcSolve
Result
according
to
ASHRAE
¾
=75%
¾
=75%
¾
=75%
¾
=75%
¾
=75%
¾
=75%
¾
=75%
¾
=75%
4/4
=100%
4/4
=100%
Consistency - - - - - -
Min-Run
Time - - - - - - - -
Min-
Evaluation - - - - - - - -
Simple
Algorithm - - - - - - - - -
Thus it is seen that the Pattern Search solver satisfies all the criteria and scores 75%
for its practicality in giving result according to ASHRAE, IES and Guidance for
employers on the Control of Noise at Work Regulations 2005, So the appropriate
algorithm, for optimization of thermal comfort is suggested as Direct search
algorithm & the solver is PATTERN SEARCH
4.3.7. Conclusion
Office environment is generally designed to the design guides and practises for the
occupant‘s comfort. In this work, the overall IEQ of offices of Karunya University in
Coimbatore was evaluated by 220 occupants in four aspects, namely thermal comfort, indoor
air quality, equivalent noise level and illumination level. All the offices considered are
naturally ventilated buildings. The results showed that the operative temperature, carbon
231
dioxide concentration, equivalent noise level and illumination level had important effects on
the overall IEQ acceptance. Empirical expressions have been proposed to approximate the
occupant‘s acceptance of the IEQ in office. The non traditional algorithms are used to find
the optimum value of the IEQ.
Here, ten non-traditional optimization algorithms were presented. These
include: GA, SA, PS, PSO, GL, FMINCON, EA, LGO, glcCluster, glcSolve. A brief
description of each method is presented along with a Pseudo code to facilitate their
implementation. MATLAB programs were written to implement each algorithm. The IEQ
problem for the offices of the Karunya University was solved using all algorithms, and the
comparative results were presented.
4.4. IEQ Residence
4.4.1. Introduction
Indoor environmental quality (IEQ) and occupant comfort are closely related. IEQ
parameters are interdependent and must be considered interactively. Other than the office
people spend most of the times in their residences. If the residences are also thermally
comfortable then their health and other related things like production and so on will not be
affected. In the case of residences also, we have considered only residences where natural
ventilation is preferred and closing and opening of the windows and using of fans is
maximum.
4.4.2. Field measurements
Subjective as well as objective evaluations of indoor environmental conditions
made by 102 occupants from 11 typical residential apartments in Karunya University were
collected through individual interviews. The interviewees were mainly those occupants
staying at the quarters of Karunya University the longest time (as compared with other
activities), and the housing samples covered 11 residential flats. The inclusion of various
apartment types could cover a wide range of probable indoor environmental conditions. The
apartments varied in size from 330 m2 to 1336.17 m
2 and were equipped with window.
232
The indoor physical parameters describing the indoor environmental quality of a
space were CO2 concentration (ppm), horizontal illuminance level (lux) and sound pressure
level (dBA). Therefore, the measured IEQ data could sufficiently reflect the real-time
exposed indoor environment to the respondent. The number of measurements was determined
based on the arrangement and partitioning of the specific apartment and the distribution of
occupants.
Unlike the measurement approach in office, each assessment sample in residential
housing did not necessarily require a separate 15 min physical measurement due to the small
living area. In this case, two physical measurements carried out in both dining/living room
and bedrooms were considered representative for the assessment of the example apartment
case. The IEQ data logged in the dining/living room were considered to be applicable to the
individual occupant within the group. Such condition was judged, depending on the space
between each occupant.
Being the base for evaluating the energy benchmarking models, effective
measurements were essential and thus accurate and reliable data could be obtained through a
dichotomous assessment scale, the occupant acceptance of the perceived indoor environment
was recorded in the form of direct feedback using the question ‗Is the thermal environment/
indoor air quality/noise level/illumination level being perceived in the residential
environment acceptable to you?‘ [42] The ranks ‗(1) Yes, acceptable‘ and ‗(0) No, not
acceptable‘ were self-explanatory. In order to confirm the validity of a response, each
respondent had to use a semantic differential evaluation scale (ref) for the subjective
assessment of thermal environment and IAQ, and a visual analogue assessment scale for the
evaluation of aural and visual comforts. At the end of the survey, an overall IEQ acceptance
was determined.
233
Table 4.23.Occupant’s votes on acceptance of a perceiving indoor environmental quality
U – Unacceptable; A - Acceptable
The overall IEQ acceptance θ for a resident environment perceived by an occupant
expressed by a multivariate logistic regression model is proposed
1
θ = 1 - -----------------------------------------
1 + exp (C0,0 + )
Where the regression constants determined from the 102 occupant evaluations; Values of the
constants confirm the relative importance of the four contributors to θ, the larger the value,
the greater the importance. Occupants were very sensitive to the operative temperature when
compared to the other three parameters. Regression coefficients can be evaluated with
surveyed occupant responses from residential environment for the overall IEQ acceptance.
C0,0 and Ci,o are the regression constants which can be determined from filed measurements,
φ1 is the occupant acceptance correlated with the thermal sensation vote 1, CO2
concentration 2 (ppm). The equivalent sound pressure level 3 (dBA) and the horizontal
illumination level 4 (lux)
The thermal environment acceptance φ1, with the maximum acceptance = 0.95, is
given below, where C0,1 and C1.1 are the regression coefficients,
The acceptances φ2 , φ3 and φ4 are expressed by logistic regression models with regression
coefficients C0,j and C1,j
Overall
acceptance
Θ
Votes Thermal
environment
φ1
Air quality
φ2
Noise level
φ3
Illumination
level
φ4
U
A
8
95
U A U A U A U A
6
5
2
89
3
3
4
92
2
6
5
89
5
5
2
90
TOTAL 102 11 91 6 96 8 94 10 92
234
Table 4.24. Regression coefficients of logistic regression model.
No variable C0,i C1,i C2,i C3,i C4,i
0 Φ0 -33.24 21.95 1.614 11.779 21.90
1 Φ1 0.03353 0.2179 --- --- ---
2 Φ2 45.21 -0.0257 --- --- ---
3 Φ3 23.82 -0.2981 --- --- ---
5 Φ4 -14.08 0.9043 --- --- ---
Various combinations of contributors i=1, 2, 3, 4 and the corresponding overall IEQ
acceptance were considered. A total of 24 possibilities were found. Taking the binary notation
for the acceptance i.e., 0 for ‗unacceptable‘ and 1 for ‗acceptable‘ the predicted acceptance of
IEQ (θ) is calculated.
Table 4.25.Overall IEQ acceptance
Case No. Survey Sample Contributors Predicted
acceptance of IEQ θ Φ1 Φ2 Φ3 Φ4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
0
1
2
0
1
2
5
1
0
0
2
0
6
6
75
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
1
1
1
1
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1.188 10-5
4.8 10-10
0.608
1.18876 10-5
1.2497 10-5
2.47 10-9
0.8889
1.2497 10-5
0.9999
0.61987
0.9999
6.2768 10-5
0.9999
0.891194
1.00000
235
4.4.3. Algorithms
4.4.3.1. Genetic algorithm
4.4.3.1.1. Stopping Criteria reached:
The options and the stopping criteria which are set are same as that for the IEQ Office
Buildings problem. This case also reaches the final solution by the stopping condition,
‖ the change in the final value of the system is less than 10-6‖. Hence we say that the global
optimum solution is obtained naturally.
TABLE.4.26. Results of GA in 20 trails for resident IEQ
Trails Thermal
sensation
vote
CO2 Sound
pressure
level
Illumination IEQ TIME ITERATIONS
1 0 744.2 67 1264.3 1 0.381663 51
2 -0.0005 803.7882 67.0003 591.4519 1 0.390658 51
3 0 1045.3 67 672.3 1 0.41928 51
4 0.0303 648.083 67.0001 508.4575 1 0.383461 51
5 0.0007 382.3028 67 988.4221 1 0.403158 51
6 -0.0004 951.0608 67.0219 261.205 1 0.377546 51
7 0.0066 354.1662 67 893.4416 1 0.391053 51
8 0.0016 623.0983 67.2189 768.8616 1 0.385192 51
9 0 1153 67 187.1 1 0.405951 51
10 0.0335 863.3774 67.0058 707.8476 1 2.318421 51
11 0 1108.4 68.5 1221.7 1 4.457991 51
12 0 609.3 67.1 1217.3 1 0.400942 51
13 0 369.5 67.3 1059.3 1 0.395878 51
14 0.0026 330 67 187.0001 1 0.41974 51
15 0 330 67 1475.6 1 0.396296 51
16 0.0007 684.0009 67.1748 187.001 1 0.415877 51
17 0 347.5 67.2 1381.9 1 0.405463 51
18 0.0033 485.9581 67.0087 89.5593 1 0.3879 51
19 0 684.3 67 1458.9 1 0.382807 51
20 0 1062.3 67 349 1 0.380528 51
Avg 0.00392 678.9818 67.12653 773.532385 1 0.69499 51
236
Fig.4.12.Convergence of GA
4.4.3.2. Simulated annealing
4.4.3.2.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for SA in the IEQ
Office Buildings problem. Though the iterations are of large number, this case also reaches
the final solution by the stopping condition,‖ the change in the final value of the system is
less than 10-6‖. The large number of iterations is because that the SA algorithm is
Metaheuristics type . Hence we say that the global optimum solution is obtained naturally.
0 10 20 30 40 50 60 70 80 90 100-1
-1
-1
-1
-1
-1
-1
Generation
Fitness v
alu
eBest: -1 Mean: -1
Best f itness
Mean fitness
237
TABLE4.27. Results of SA in 20 trails for resident IEQ
Trails Thermal
sensation
vote
CO2 Sound
pressure
level
Illumination IEQ Time Iterations
1 0 915 67 854.5 1 2.741429 2000
2 0 915 67 854.5 1 1.024044 2000
3 0 915 67 854.5 1 1.05276 2000
4 0 915 67 854.5 1 1.031611 2000
5 0 915 67 854.5 1 1.035856 2000
6 0 915 67 854.5 1 1.034596 2000
7 0 915 67 854.5 1 1.029433 2000
8 0 915 67 854.5 1 1.067096 2000
9 0 915 67 854.5 1 1.044101 2000
10 0 915 67 854.5 1 1.042642 2000
11 0 915 67 854.5 1 1.026046 2000
12 0 915 67 854.5 1 1.021253 2000
13 0 915 67 854.5 1 1.018455 2000
14 0 915 67 854.5 1 1.02123 2000
15 0 915 67 854.5 1 1.018589 2000
16 0 915 67 854.5 1 1.007763 2000
17 0 915 67 854.5 1 1.027883 2000
18 0 915 67 854.5 1 1.018682 2000
19 0 915 67 854.5 1 1.010296 2000
20 0 915 67 854.5 1 1.021411 2000
Avg 0 915 67 854.5 1 1.114759 2000
Fig.4.13.Convergence of SA
0 200 400 600 800 1000 1200 1400 1600 1800 2000-2
-1.5
-1
-0.5
0
0.5
Iteration
Function v
alu
e
Best Function Value: -1
1 2 3 40
200
400
600
800
1000Best point
Number of variables (4)
Best
poin
t
0 10 20 30 40 50 60 70 80 90 100
Time
Iteration
f-count
% of criteria met
Stopping Criteria
0 200 400 600 800 1000 1200 1400 1600 1800 2000-1
-0.8
-0.6
-0.4
-0.2
0
Iteration
Function v
alu
e
Current Function Value: -0.99134
238
4.4.3.3. Pattern search
4.4.3.3.1. Stopping Criteria Reached:
The solution is reached by the stopping condition, ―difference in function value less
than 10-6‖ and also comparatively the iterations are of less in number, this indicates quick
convergence. The final value of the solution is naturally obtained.
TABLE.4.28. Results of PS in 20 trails for resident IEQ
Trails Thermal CO2 Sound Illumination IEQ Time Iterations
1 0 915 67 854.5 1 0.979432 20
2 0 915 67 854.5 1 0.053695 20
3 0 915 67 854.5 1 0.062384 20
4 0 915 67 854.5 1 0.054444 20
5 0 915 67 854.5 1 0.060776 20
6 0 915 67 854.5 1 0.06348 20
7 0 915 67 854.5 1 0.054188 20
8 0 915 67 854.5 1 0.050838 20
9 0 915 67 854.5 1 0.062327 20
10 0 915 67 854.5 1 0.043569 20
11 0 915 67 854.5 1 0.07328 20
12 0 915 67 854.5 1 0.053983 20
13 0 915 67 854.5 1 0.62883 20
14 0 915 67 854.5 1 0.054073 20
15 0 915 67 854.5 1 0.062371 20
16 0 915 67 854.5 1 0.052222 20
17 0 915 67 854.5 1 0.04451 20
18 0 915 67 854.5 1 0.063024 20
19 0 915 67 854.5 1 0.070971 20
20 0 915 67 854.5 1 0.075022 20
Avg 0 915 67 854.5 1 0.133171 20
239
Fig.4.14.Convergence of PS
0 2 4 6 8 10 12 14 16 18 20-2
-1.5
-1
-0.5
0
0.5
Iteration
Function v
alu
e
Best Function Value: -1
0 2 4 6 8 10 12 14 16 18 200
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Iteration
Mesh s
ize
Current Mesh Size: 9.5367e-007
240
4.4.3.4. Particle swarm optimization
4.4.3.4.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for PSO in the
IEQ Office Buildings problem. This case also the final solution reaches by the stopping
condition,‖ the change in the final value of the system is less than 10-6‖ but the specialty is
the elapsed time which is less than other solvers. The global optimum solution is obtained
without any other stopping conditions.
TABLE.4.29. Results of PSO in 20 trails for resident IEQ
Trails Thermal CO2 Sound level Illumination IEQ Time Iterations
1 0 1076.8 67 437.1 1 0.120159 70
2 0 1221.6 67 425.9 1 0.121556 56
3 0 849 67 1288.8 1 0.130095 52
4 0 849 67 1288.8 1 0.130094 52
5 -0.0282 920.0475 67.0023 246.3722 1 0.121258 51
6 0 1204.9 67 998.7 1 0.10458 51
7 0 1111.5 67 489.9 1 0.140347 54
8 0.0069 534.5203 67.0005 344.6397 1 0.147119 64
9 0 1279.8 67 1276.6 1 0.145017 68
10 0.0043 827.5092 67 544.1189 1 0.120778 52
11 0 1021.7 67 1253.8 1 0.104316 51
12 0 1085.6 67 1367.2 1 0.123623 57
13 0 1305 67 871.2 1 0.126779 60
14 0 843.1 67 1220.5 1 0.138437 52
15 0.002 672.6782 67.0011 372.4912 1 0.114034 51
16 0 1384 67 780 1 0.114465 52
17 0.0064 650.2896 67.0001 459.1902 1 0.157973 58
18 -0.0115 645.3098 67.0017 763.8825 1 0.094197 54
19 0 1414.4 67 511 1 0.112067 51
20 0.152 453.1459 67.0029 458.3543 1 0.152493 51
Avg 0.006595 967.495 67.00043 769.92745 1 0.125969 55.35
241
4.4.3.5. GODLIKE
4.4.3.5.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for GODLIKE in
the IEQ Office Buildings problem. This case also the final solution reaches by the stopping
condition,‖ the change in the final value of the system is less than 10-6‖.The solver exchanges
the population among the solvers hence the iteration indicates number of times the population
is exchanged. The global optimum solution is obtained without any other stopping conditions.
TABLE.4.30. Results of GL in 20 trails for resident IEQ
Trails Thermal vote CO2 Sound Illumination IEQ Time Iterations
1 -0.0005 880.7416 67 510.4552 1 1.421388 4
2 0 707.1 67 1040.7 1 1.483061 4
3 0 885 67 1093.7 1 1.011846 4
4 -0.0002 945.5216 67 356.0738 1 1.652818 4
5 0 922.4489 67 561.348 1 1.396136 4
6 0 1194.2 67 927.2 1 1.606247 4
7 0 1273.4 67 1101.4 1 1.888101 4
8 0.0001 833.6192 67 640.5575 1 1.345902 4
9 0 1088.7 67 1021.5 1 3.324052 4
10 -0.0006 728.5694 67 800.8187 1 1.21074 4
11 0 768.5 67 1095.1 1 2.176719 4
12 0.0013 868.9358 67 575.0751 1 1.179232 4
13 0 451.4 67 1049 1 1.23686 4
14 -0.001 859.5101 67.00006 948.2773 1 1.025074 4
15 0 946.1 67 1248 1 2.032979 4
16 0 800.9 67 1250.9 1 1.237546 4
17 0 1127.9 67 1159.2 1 1.085968 4
18 0.0006 753.0979 67 557.074 1 1.954029 4
19 0 1053.7 67 616.5 1 1.552567 4
20 0 525.6 67 1382 1 1.325332 4
Avg -0.000015 880.7472 67 896.74398 1 1.55733 4
242
4.4.3.6. Fmincon.
4.4.3.6.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for
Fmincon in the IEQ Office Buildings problem. This case also the final solution reaches by the
stopping condition,‖ the change in the final value of the system is less than 10-6‖. The global
optimum solution is obtained without any other stopping conditions. The exception is that the
elapsed time is high comparatively; this is due to the traditional technique modified version
of using lagranges multipliers.
TABLE.4.31. Results of Fmincon in 20 trails for resident IEQ
Trails Thermal vote CO2 Sound Illumination IEQ Time Iterations
1 0 915 67.0001 854.5 1 4.758093 2972
2 0 915 67.0001 854.5 1 3.753865 2694
3 0 915 67.0001 854.5 1 6.801059 4607
4 0 915 67.0001 854.5 1 4.662614 3188
5 0 626.5 67 1199.6 1 4.542072 2728
6 0 915 67.0001 854.5 1 5.275708 3777
7 0 915 67.0001 854.5 1 5.449035 3034
8 0 915 67.0001 854.5 1 4.073841 2881
9 0 915 67.0001 854.5 1 4.747727 3214
10 0 915 67.0001 854.5 1 5.406024 4080
11 0 915 67.0001 854.5 1 5.538329 3780
12 0 915 67.0001 854.5 1 3.909716 2722
13 0 915 67.0001 854.5 1 3.911251 2953
14 0 915 67.0001 854.5 1 4.54207 2728
15 0 915 67.0001 854.5 1 8.01847 3505
16 0 915 67.0001 854.5 1 4.79385 2929
17 0 928.2 67 1111.2 1 4.99689 3108
18 0 915 67.0001 854.5 1 3.32011 2562
19 0 915 67.0001 854.5 1 9.42467 3923
20 0 915 67.0001 854.5 1 4.21304 2783
AVG 0 901.235 67.0000 884.59 1 5.10692 3208
243
4.4.3.7. Direct evolution
4.4.3.7.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for DE in the IEQ
Office Buildings problem. This case also the final solution reaches by the stopping
condition,‖ the change in the final value of the system is less than 10-6‖. It is seen from the
results that the final vectors (parameter values) is not consistent, this is because DE uses
different type of cross over method. The global optimum solution is obtained without any
other stopping conditions.
TABLE.4.32. Results of DE in 20 trails for resident IEQ
Trails Thermal CO2 Sound Illumination IEQ Time Iterations
1 -0.4 1295.6 67.9 364.8 1 0.015565 40
2 0.2 710.4 68.2 1002.6 1 0.001693 40
3 0.3 1173.6 72.7 1513.6 1 0.002347 40
4 -0.2 1088.4 67.3 1311.3 1 0.004192 40
5 0.3 696.7 68.8 1018 1 0.001568 40
6 0.3 1012.8 72.6 297.3 1 0.00224 40
7 0.3 1457.3 67.5 1485.9 1 0.002525 40
8 0.1019 981.3327 68.7214 937.3448 1 0.004635 40
9 0.1427 639.9009 67.7519 769.4966 1 0.0045 40
10 -0.4 1206.5 67.4 1450.3 1 0.004322 40
11 -0.2 802.6 69 1480.5 1 0.004393 40
12 0 940.3 68 1394.7 1 0.004464 40
13 0.1 1446 67.8 463.4 1 0.002196 40
14 -0.6 1427.8 69.8 898.8 1 0.004563 40
15 -0.2 1441.2 67.9 560.6 1 0.004486 40
16 -0.1 1396.9 67.2 396.2 1 0.004362 40
17 0.209 668.6973 67.7521 300.4359 1 0.004467 40
18 0.1 1255.1 69.2 1092.3 1 0.004208 40
19 0 718.4 68.9 1022.2 1 0.002391 40
20 0.1 1235.9 68.3 1022 1 0.002312 40
Avg 0.00268 1079.772 68.63627 939.088865 1 0.004071 40
244
4.4.3.8. LGO
4.4.3.8.1. Stopping Criteria Reached:
The options and the stopping criteria which are set are same as that for LGO in the
IEQ Office Buildings problem. The global solution reaches by the stopping condition,‖ the
change in the final value of the system did not improve‖. The elapsed time is close to that of
other Direct algorithm solvers but it does not use Lipchitz constant
TABLE.4.33 .Results of LGO in 20 trails for resident IEQ
Trails Thermal CO2 Sound Illumination IEQ Time Iterations
1 0.0027 497.0895 67 931.8641 1 1.166753 3223
2 0.0027 497.0895 67 931.8641 1 0.42993 3223
3 0.0027 497.0895 67 931.8641 1 0.42055 3223
4 0.0027 497.0895 67 931.8641 1 0.43747 3223
5 0.0027 497.0895 67 931.8641 1 0.442481 3223
6 0.0027 497.0895 67 931.8641 1 0.414941 3223
7 0.0027 497.0895 67 931.8641 1 0.421402 3223
8 0.0027 497.0895 67 931.8641 1 0.438488 3223
9 0.0027 497.0895 67 931.8641 1 0.422659 3223
10 0.0027 497.0895 67 931.8641 1 0.431362 3223
11 0.0027 497.0895 67 931.8641 1 0.427368 3223
12 0.0027 497.0895 67 931.8641 1 0.435591 3223
13 0.0027 497.0895 67 931.8641 1 0.440821 3223
14 0.0027 497.0895 67 931.8641 1 0.431039 3223
15 0.0027 497.0895 67 931.8641 1 0.426922 3223
16 0.0027 497.0895 67 931.8641 1 0.435226 3223
17 0.0027 497.0895 67 931.8641 1 0.431876 3223
18 0.0027 497.0895 67 931.8641 1 0.428309 3223
19 0.0027 497.0895 67 931.8641 1 0.437069 3223
20 0.0027 497.0895 67 931.8641 1 0.429926 3223
Avg 0.0027 497.0895 67 931.8641 1 0.467509 3223
245
4.4.3.9. glcCluster
4.4.3.9.1. Stopping Criteria Reached:
The default options are taken from the solver from the previous run of the IEQ Office
Buildings problem. The global solution reaches by the stopping condition,‖ the change in the
final value of the system is less than 10-7‖.Though glcCluster uses Clustering algorithm in
addition it has very less elapsed time.
TABLE4.34. Results of glcCluster in 20 trails for resident IEQ
Trails Thermal CO2 Sound Illumination IEQ Time Iterations
1 0 915 67.6111 1299.5 1 0.027167 1516
2 0 525 67.6111 1299.5 1 0.027167 1516
3 0 655 67.6111 1299.5 1 0.027167 1516
4 0 1045 67.6111 1299.5 1 0.027167 1516
5 0 1045 67.6111 1299.5 1 0.027167 1516
6 0 655 67.6111 1299.5 1 0.027167 1516
7 0 1175 67.6111 1299.5 1 0.027167 1516
8 0 525 67.6111 1299.5 1 0.027167 1516
9 0 655 67.6111 1299.5 1 0.027167 1516
10 0 655 67.6111 1299.5 1 0.027167 1516
11 0 915 67.6111 1299.5 1 0.027194 1516
12 0 525 67.6111 409.5 1 0.027194 1516
13 0 655 67.6111 854.5 1 0.027194 1516
14 0 1045 67.6111 1299.5 1 0.027194 1516
15 0 1045 67.6111 1299.5 1 0.027194 1516
16 0 655 67.6111 1299.5 1 0.027194 1516
17 0 1175 67.6111 1299.5 1 0.027194 1516
18 0 525 67.6111 1299.5 1 0.027194 1516
19 0 655 67.6111 409.5 1 0.027194 1516
20 0 655 67.6111 1299.5 1 0.027194 1516
Avg 0 785 67.6111 1188.25 1 0.02718 1516
246
4.4.3.10. glcSolve
4.4.3.10.1. Stopping Criteria Reached:
The options and the stopping criteria are taken from the previous run of IEQ Office
Building problem. The final solution reaches by the stopping condition,‖ the change in the
final value of the system is less than 10-6‖. glcSolve uses one of the complex algorithm and
even after giving long range values for parameters (which is not recommended) it takes little
time to complete optimization.
TABLE.4.35. Results of glcSolve in 20 trails for resident IEQ
Trails Thermal CO2
concen
tration
Sound Illumination IEQ Time Fun eval Iterations
1 0 915 67.611 1299.5 1 0.31502 1529 110
2 0 915 67.611 1299.5 1 0.360233 1529 110
3 0 915 67.611 1299.5 1 0.332232 1529 110
4 0 915 67.611 1299.5 1 0.334579 1529 110
5 0 915 67.611 1299.5 1 0.355238 1529 110
6 0 915 67.611 1299.5 1 0.388838 1529 110
7 0 915 67.611 1299.5 1 0.342736 1529 110
8 0 915 67.611 1299.5 1 0.332214 1529 110
9 0 915 67.611 1299.5 1 0.333896 1529 110
10 0 915 67.611 1299.5 1 0.336873 1529 110
11 0 915 67.611 1299.5 1 0.336153 1529 110
12 0 915 67.611 1299.5 1 0.323308 1529 110
13 0 915 67.611 1299.5 1 0.333478 1529 110
14 0 915 67.611 1299.5 1 0.320211 1529 110
15 0 915 67.611 1299.5 1 0.340051 1529 110
16 0 915 67.611 1299.5 1 0.337327 1529 110
17 0 915 67.611 1299.5 1 0.333373 1529 110
18 0 915 67.611 1299.5 1 0.343034 1529 110
19 0 915 67.611 1299.5 1 0.325855 1529 110
20 0 915 67.611 1299.5 1 0.334273 1529 110
Avg 0 915 67.611 1299.5 1 0.337946 1529 110
247
TABLE.4.36 Comparative results of optimization methods for Resident IEQ.
IEQ-RESIDENCE
Methods
Thermal
sensation
vote
Co2
concentration
Sound
pressure
level
Horizontal
Illumination
IEQ Time Iterations
Genetic
Algorithm 0.00392 678.9818 67.1265 773.53238 1 0.6949 51
Simulated
annealing
ANNEALING
0 915 67 854.5 1 1.1147 2000
PATTERN
SEARCH 0 915 67 854.5 1 0.1331 20
PSO 0.00659 967.495 67.0004 769.92745 1 0.1259 55.35
GODLIKE 0.00001 880.7472 67 896.74398 1 1.5573 4
fmincon 0 901.235 67.0000 884.59 1 5.1069 3208.4
DE
SOLUTION 0.00268 1079.772 68.6362 1605.31387 1 0.004 40
LGO 0.0027 497.0895 67 931.8641 1 0.4675 3223
glcCluster 0.00105 882.688 67.4026 1111.1647 1 0.770 1286.98
glcSolve 0 915 67.611 1299.5 1 0.3379 1529
Fig .4.15.Comparative graph for residence thermal comfort
4.4.4. Comparison of results
The IEQ values for all the ten optimization is 1 and that is the optimum value. The
elapsed time is maximum for DE and minimum for PS. The carbon dioxide concentration,
0.001
0.01
0.1
1
10
100
1000
10000 GA
SA
PS
PSO
GL
Fmincon
DE
LGO
glcCluster
glcSolve
248
sound pressure level and illumination level are more or less the same for all methods. The
thermal sensation is neutral for few methods.
4.4.5. PARAMETERS.
4.4.5.1. THERMAL SENSATION
Thermal comfort is that condition of mind which expresses satisfaction with the
thermal environment. Thermal environment encompasses characteristics of the environment
which affects a person's heat loss. In terms of bodily sensations, thermal comfort is a
sensation of hot, warm, slightly warmer, neutral, slightly cooler, cool and cold
TABLE.4.37. Thermal sensation results in all 10 methods
Trials GA SA PS PSO GODLIKE fmincon DE LGO glcClu glcSol
1 0 0 0 0 -0.0005 0 -0.4 0.0027 0 0
2 -0.0005 0 0 0 0 0 0.2 0.0027 0 0
3 0 0 0 0 0 0 0.3 0.0027 0 0
4 0.0303 0 0 0 -0.0002 0 -0.2 0.0027 0 0
5 0.0007 0 0 -0.0282 0 0 0.3 0.0027 0 0
6 -0.0004 0 0 0 0 0 0.3 0.0027 0 0
7 0.0066 0 0 0 0 0 0.3 0.0027 0 0
8 0.0016 0 0 0.0069 0.0001 0 0.1019 0.0027 0 0
9 0 0 0 0 0 0 0.1427 0.0027 0 0
10 0.0335 0 0 0.0043 -0.0006 0 -0.4 0.0027 0 0
11 0 0 0 0 0 0 -0.2 0.0027 0 0
12 0 0 0 0 0.0013 0 0 0.0027 0 0
13 0 0 0 0 0 0 0.1 0.0027 0 0
14 0.0026 0 0 0 -0.001 0 -0.6 0.0027 0 0
15 0 0 0 0.002 0 0 -0.2 0.0027 0 0
16 0.0007 0 0 0 0 0 -0.1 0.0027 0 0
17 0 0 0 0.0064 0 0 0.209 0.0027 0 0
18 0.0033 0 0 -0.0115 0.0006 0 0.1 0.0027 0 0
19 0 0 0 0 0 0 0 0.0027 0 0
20 0 0 0 0.152 0 0 0.1 0.0027 0 0
avg 0.00392 0 0 0.00659
5
-0.000015 0 0.0026
8
0.0027 0 0
249
FIG.4.16. Graph for Thermal sensation results in all 10 methods
250
4.4.5.2. CARBON DI-OXIDE
Carbon dioxide (CO2) is the chief greenhouse gas that results from human activities
and causes global warming and climate change. Though carbon dioxide is not toxic in itself,
the amount found in the indoor environment is used as an indicator for human comfort.
Elevated levels of carbon dioxide indicate that an insufficient amount of fresh, outdoor air is
being delivered to the occupied areas of the building. This also indicates that other pollutants
in the building may exist at elevated levels since there is not enough fresh air to dilute them.
Since carbon dioxide is an unavoidable, predictable, and easily measured product of human
occupancy, it is used as a marker for other pollutants emanating from humans or other
sources in the building. Carbon dioxide is mostly a threat to health, when the concentration is
high enough to displace the oxygen, which can lead to suffocation in a confined space.
TABLE.4.38. Carbon Dioxide results in all 10 methods
Trials GA SA PS PSO GODLIKE fmincon DE LGO glcClu glcSol
1 744.2 915 915 1076.8 880.7416 915 1295.6 497.0895 915 915
2 803.7882 915 915 1221.6 707.1 915 710.4 497.0895 525 915
3 1045.3 915 915 849 885 915 1173.6 497.0895 655 915
4 648.083 915 915 849 945.5216 915 1088.4 497.0895 1045 915
5 382.3028 915 915 920.0475 922.4489 626.5 696.7 497.0895 1045 915
6 951.0608 915 915 1204.9 1194.2 915 1012.8 497.0895 655 915
7 354.1662 915 915 1111.5 1273.4 915 1457.3 497.0895 1175 915
8 623.0983 915 915 534.5203 833.6192 915 981.332
7
497.0895 525 915
9 1153 915 915 1279.8 1088.7 915 639.900
9
497.0895 655 915
10 863.3774 915 915 827.5092 728.5694 915 1206.5 497.0895 655 915
11 1108.4 915 915 1021.7 768.5 915 802.6 497.0895 915 915
12 609.3 915 915 1085.6 868.9358 915 940.3 497.0895 525 915
13 369.5 915 915 1305 451.4 915 1446 497.0895 655 915
14 330 915 915 843.1 859.5101 915 1427.8 497.0895 1045 915
15 330 915 915 672.6782 946.1 915 1441.2 497.0895 1045 915
16 684.0009 915 915 1384 800.9 915 1396.9 497.0895 655 915
17 347.5 915 915 650.2896 1127.9 928.2 668.697
3
497.0895 1175 915
18 485.9581 915 915 645.3098 753.0979 915 1255.1 497.0895 525 915
19 684.3 915 915 1414.4 1053.7 915 718.4 497.0895 655 915
20 1062.3 915 915 453.1459 525.6 915 1235.9 497.0895 655 915
avg 678.9818 915 915 967.4950
3
880.747225 901.235 1079.77
1545
497.0895 785 915
251
FIG.4.17. Graph for Carbon Dioxide results in all 10 methods
252
4.4.5.3. SOUND PRESSURE LEVEL
Acoustics is the interdisciplinary science that deals with the study of all mechanical
waves in gases, liquids, and solids including vibration, sound, ultrasound and infrasound. The
perception of sound in any organism is limited to a certain range of frequencies. Hearing loss
due to prolonged exposure to noise is well documented. Excessive noise also has an adverse
effect on personal health and wellbeing, ability to perform quiet tasks, and productivity in
general. Because land is becoming scarcer, buildings are being constructed closer together
and closer to noise sources such as highways, railways, and airports. As a result, sound or
acoustic control is becoming increasingly important. The reduction of airborne sound through
a wall is called sound transmission loss (STL).
TABLE.4.39. Sound Pressure level results in all 10 methods
Trials GA SA PS PSO GODLIKE fmincon DE LGO glcClu glcSol
1 67 67 67 67 67 67.0001 67.9 67 67.6111 67.611
2 67.0003 67 67 67 67 67.0001 68.2 67 67.6111 67.611
3 67 67 67 67 67 67.0001 72.7 67 67.6111 67.611
4 67.0001 67 67 67 67 67.0001 67.3 67 67.6111 67.611
5 67 67 67 67.0023 67 67 68.8 67 67.6111 67.611
6 67.0219 67 67 67 67 67.0001 72.6 67 67.6111 67.611
7 67 67 67 67 67 67.0001 67.5 67 67.6111 67.611
8 67.2189 67 67 67.0005 67 67.0001 68.721
4
67 67.6111 67.611
9 67 67 67 67 67 67.0001 67.751
9
67 67.6111 67.611
10 67.0058 67 67 67 67 67.0001 67.4 67 67.6111 67.611
11 68.5 67 67 67 67 67.0001 69 67 67.6111 67.611
12 67.1 67 67 67 67 67.0001 68 67 67.6111 67.611
13 67.3 67 67 67 67 67.0001 67.8 67 67.6111 67.611
14 67 67 67 67 67.00006 67.0001 69.8 67 67.6111 67.611
15 67 67 67 67.0011 67 67.0001 67.9 67 67.6111 67.611
16 67.1748 67 67 67 67 67.0001 67.2 67 67.6111 67.611
17 67.2 67 67 67.0001 67 67 67.752
1
67 67.6111 67.611
18 67.0087 67 67 67.0017 67 67.0001 69.2 67 67.6111 67.611
19 67 67 67 67 67 67.0001 68.9 67 67.6111 67.611
20 67 67 67 67.0029 67 67.0001 68.3 67 67.6111 67.611
avg 67.12653 67 67 67.0004
3
67.000003 67.00009 68.636
27
67 67.6111 67.611
253
FIG.4.18. Graph for Sound Pressure level results in all 10 methods
254
4.4.5.4. HORIZONTAL ILLUMINATION.
Lighting or illumination is the deliberate application of light to achieve some aesthetic
or practical effect. In some design instances, materials used on walls and furniture play a key
role in the lighting effect. Surfaces or floors that are too reflective create unwanted glare.
Specification of illumination requirements is the basic concept of deciding how
much illumination is required for a given task. Clearly, much less light is required to
illuminate a hallway or a bathroom compared to that needed for a word processing work
station. Generally speaking, the energy expended is proportional to the design illumination
level. Beyond the energy factors being considered, it is important not to over-design
illumination, lest adverse health effects such as headache frequency, stress, and
increased blood pressure be induced by the higher lighting levels. In addition, glare or excess
light can decrease worker efficiency.
TABLE.4.40.Horizontal illumination results in all 10 methods.
Trials GA SA PS PSO GODLIKE fmincon DE LGO glcClu glcSol
1 1264.3 854.5 854.5 437.1 510.4552 854.5 364.8 931.86 1299.5 1299.5
2 591.4519 854.5 854.5 425.9 1040.7 854.5 1002.6 931.86 1299.5 1299.5
3 672.3 854.5 854.5 1288.8 1093.7 854.5 1513.6 931.86 1299.5 1299.5
4 508.4575 854.5 854.5 1288.8 356.0738 854.5 1311.3 931.86 1299.5 1299.5
5 988.4221 854.5 854.5 246.3722 561.348 1199.6 1018 931.86 1299.5 1299.5
6 261.205 854.5 854.5 998.7 927.2 854.5 297.3 931.86 1299.5 1299.5
7 893.4416 854.5 854.5 489.9 1101.4 854.5 1485.9 931.86 1299.5 1299.5
8 768.8616 854.5 854.5 344.6397 640.5575 854.5 937.344
8
931.86 1299.5 1299.5
9 187.1 854.5 854.5 1276.6 1021.5 854.5 769.496
6
931.86 1299.5 1299.5
10 707.8476 854.5 854.5 544.1189 800.8187 854.5 1450.3 931.86 1299.5 1299.5
11 1221.7 854.5 854.5 1253.8 1095.1 854.5 1480.5 931.86 1299.5 1299.5
12 1217.3 854.5 854.5 1367.2 575.0751 854.5 1394.7 931.86 409.5 1299.5
13 1059.3 854.5 854.5 871.2 1049 854.5 463.4 931.86 854.5 1299.5
14 187.0001 854.5 854.5 1220.5 948.2773 854.5 898.8 931.86 1299.5 1299.5
15 1475.6 854.5 854.5 372.4912 1248 854.5 560.6 931.86 1299.5 1299.5
16 187.001 854.5 854.5 780 1250.9 854.5 396.2 931.86 1299.5 1299.5
17 1381.9 854.5 854.5 459.1902 1159.2 1111.2 300.435
9
931.86 1299.5 1299.5
18 89.5593 854.5 854.5 763.8825 557.074 854.5 1092.3 931.86 1299.5 1299.5
19 1458.9 854.5 854.5 511 616.5 854.5 1022.2 931.86 409.5 1299.5
20 349 854.5 854.5 458.3543 1382 854.5 1022 931.86 1299.5 1299.5
avg 773.5324 854.5 854.5 769.92745 896.74398 884.59 939.088
865
931.86 1188.2
5
1299.5
255
FIG.4.19. Graph for Horizontal Illumination results in all 10 methods
256
4.4.5.5. IEQ
TABLE.4.41. IEQ results in all 10 methods
Trails GA SA PS PSO GOD
LIKE
fmincon DE LGO glcCluster glcSolve
1 1 1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
3 1 1 1 1 1 1 1 1 1 1
4 1 1 1 1 1 1 1 1 1 1
5 1 1 1 1 1 1 1 1 1 1
6 1 1 1 1 1 1 1 1 1 1
7 1 1 1 1 1 1 1 1 1 1
8 1 1 1 1 1 1 1 1 1 1
9 1 1 1 1 1 1 1 1 1 1
10 1 1 1 1 1 1 1 1 1 1
11 1 1 1 1 1 1 1 1 1 1
12 1 1 1 1 1 1 1 1 1 1
13 1 1 1 1 1 1 1 1 1 1
14 1 1 1 1 1 1 1 1 1 1
15 1 1 1 1 1 1 1 1 1 1
16 1 1 1 1 1 1 1 1 1 1
17 1 1 1 1 1 1 1 1 1 1
18 1 1 1 1 1 1 1 1 1 1
19 1 1 1 1 1 1 1 1 1 1
20 1 1 1 1 1 1 1 1 1 1
Avg 1 1 1 1 1 1 1 1 1 1
257
FIG.4.20 Graph for IEQ results in all 10 methods
258
4.4.5.6. ELAPSED TIME
CPU time is the time for which the CPU was busy executing the task. It does not take
into account the time spent in waiting for I/O (disk IO or network IO). Since I/O operations,
such as reading files from disk, are performed by the OS, these operations may involve
noticeable amount of time in waiting for I/O subsystems to complete their operations. This
waiting time will be included in the elapsed time, but not CPU time. Hence CPU time is
usually less than the elapsed time.
TABLE.4.42. Elapsed time results in all 10 methods
Trials GA SA PS PSO G-L fminco
n
DE LGO glcClu
ster
glcSol
ve 1 0.3816
63
2.74 0.97 0.12 1.42 4.75 0.0155
65
1.16 0.0271
67
0.31
2 0.3906
58
1.02 0.05 0.12 1.48 3.75 0.0016
93
0.42 0.0271
67
0.36
3 0.4192
8
1.05 0.06 0.13 1.01 6.80 0.0023
47
0.42 0.0271
67
0.33
4 0.3834
61
1.03 0.05 0.13 1.65 4.66 0.0041
92
0.43 0.0271
67
0.33
5 0.4031
58
1.03 0.06 0.12 1.39 4.54 0.0015
68
0.44 0.0271
67
0.35
6 0.3775
46
1.03 0.06 0.10 1.60 5.27 0.0022
4
0.41 0.0271
67
0.38
7 0.3910
53
1.02 0.05 0.14 1.88 5.44 0.0025
25
0.42 0.0271
67
0.34
8 0.3851
92
1.06 0.05 0.14 1.34 4.07 0.0046
35
0.43 0.0271
67
0.33
9 0.4059
51
1.04 0.06 0.14 3.32 4.74 0.0045 0.42 0.0271
67
0.33
10 2.3184
21
1.04 0.04 0.12 1.21 5.40 0.0043
22
0.43 0.0271
67
0.33
11 4.4579
91
1.02 0.07 0.10 2.17 5.53 0.0043
93
0.42 0.0271
94
0.33
12 0.4009
42
1.02 0.05 0.12 1.17 3.90 0.0044
64
0.43 0.0271
94
0.32
13 0.3958
78
1.01 0.09 0.12 1.23 3.91 0.0021
96
0.44 0.0271
94
0.33
14 0.4197
4
1.02 0.05 0.13 1.02 4.54 0.0045
63
0.43 0.0271
94
0.32
15 0.3962
96
1.01 0.06 0.11 2.03 8.01 0.0044
86
0.42 0.0271
94
0.34
16 0.4158
77
1.00 0.05 0.11 1.23 4.79 0.0043
62
0.43 0.0271
94
0.33
17 0.4054
63
1.02 0.04 0.15 1.08 4.99 0.0044
67
0.43 0.0271
94
0.33
18 0.3879 1.01 0.06 0.09 1.95 3.32 0.0042
08
0.42 0.0271
94
0.34
19 0.3828
07
1.01 0.07 0.11 1.55 9.42 0.0023
91
0.43 0.0271
94
0.32
20 0.3805
28
1.02 0.07 0.15 1.32 4.21 0.0023
12
0.42 0.0271
94
0.33
Avg 0.6949
9
1.11 0.13 0.12 1.55 5.10 0.0040
7145
0.46 0.0271
8
0.33
259
FIG.4.21. graph for Elapsed time results in all 10 methods
260
4.4.5.7. ITERATIONS
Iteration is a computational procedure in which a cycle of operations is repeated, often
to approximate the desired result more closely. Iteration means the act of repeating a process
usually with the aim of approaching a desired goal or target or result. Iteration in computing
is the repetition of a process within a computer program. It may also refer to the process of
iterating a function i.e. applying a function repeatedly, using the output from one iteration as
the input to the next. Another use of iteration in mathematics is in iterative methods which
are used to produce approximate numerical solutions to certain mathematical problems.
TABLE.4.43 . Iterations results in all 10 methods
Trials GA SA PS PSO GODLIKE fmincon DE LGO glcCluster glcSolve
1 51 2000 20 70 4 2972 40 3223 1516 1529
2 51 2000 20 56 4 2694 40 3223 1516 1529
3 51 2000 20 52 4 4607 40 3223 1516 1529
4 51 2000 20 52 4 3188 40 3223 1516 1529
5 51 2000 20 51 4 2728 40 3223 1516 1529
6 51 2000 20 51 4 3777 40 3223 1516 1529
7 51 2000 20 54 4 3034 40 3223 1516 1529
8 51 2000 20 64 4 2881 40 3223 1516 1529
9 51 2000 20 68 4 3214 40 3223 1516 1529
10 51 2000 20 52 4 4080 40 3223 1516 1529
11 51 2000 20 51 4 3780 40 3223 1516 1529
12 51 2000 20 57 4 2722 40 3223 1516 1529
13 51 2000 20 60 4 2953 40 3223 1516 1529
14 51 2000 20 52 4 2728 40 3223 1516 1529
15 51 2000 20 51 4 3505 40 3223 1516 1529
16 51 2000 20 52 4 2929 40 3223 1516 1529
17 51 2000 20 58 4 3108 40 3223 1516 1529
18 51 2000 20 54 4 2562 40 3223 1516 1529
19 51 2000 20 51 4 3923 40 3223 1516 1529
20 51 2000 20 51 4 2783 40 3223 1516 1529
avg 51 2000 20 55.35 4 3208 40 3223 1516 1529
261
FIG.4.22. Graph for Iterations results in all 10 methods
262
TABLE.4.44 Comparative table for parameters in all 10 methods
Variables GA SA PS PSO GL fminc
on
DE LGO Glc
Cluster
Glc
Solve
PMV X
0
0 X X
0 X X
0
0
CO2 X
915
915 X X X X
497
785
915
Sound X
67
67 X
67 X X
67
67.6
67.6
Illumina
tion
X
854.5
854.5 X X X X
931.8 X
1299.5
IEQ
1
1
1
1
1
1
1
1
1
1
TIME 0.13 0.12 0.027
ITERS X 20
X 4
X
- Represents the parameters are consistent for all the 20 trials and the corresponding
parameter values are given in the respective cell.
X - Represents the parameters are not consistent for all the 20 trials
In case of iterations and elapsed time only the two or three minimum values alone are given.
4.4.6. Result and Discussion
With the two extreme values of parameters from survey, the optimization is carried
out with different solvers. As they are of stochastic type their results may vary from trial to
trial so and the problem is made to run for 20 trials (Elbeltagi, Tarek Hegazy, & & Grierson,
2005) and an average of all trials is taken as the final value of the parameter, by the solver.
The solvers are compared with different criteria
263
1. Consistency
The consistency table gives the parameters that remain constant for all the
trails. All the solvers give the same value of IEQ for all the runs. Which in
turn indicate that the quality requirements are in the acceptable range.
Thermal – PS (0),NL (0),SA(0), glcSolve (0), glcCluster (0), LGO (0.0027)
CO 2 - PS(915), SA(915), glcSolve (915), LGO (497)
Sound - PS(67),SA(67), glcSolve (67.611), glcCluster (67.611), LGO (67)
Illumination - PS(854.5),SA(854.5), glcSolve (1299.5), LGO (931.86)
So we see that the solvers SA, Pattern Search, glcSolve, glcCluster& LGO
remain constant throughout their runs.
2. Minimum Run Time
For minimum run time of the problem we got PSO (0.12 seconds), Pattern
Search (0.13 seconds).
3. Minimum Evaluation
This criterion will determine the effectiveness of the algorithm. From the
result table we see that the Pattern Search and GODLIKE algorithms have
minimum evaluation of 20 and 4 respectively
4. Simplicity of Algorithm
Of all the algorithms we have taken the Pattern Search algorithm is the most
simplest followed by GA, PSO, DE, Simulated Annealing, GODLIKE, Non-
Linear, Direct algorithm.
5. Results according to Standards
This is the most important criterion that determines whether the solver is
practical or not. We got the standard values from ASHRAE, IES, Guidance for
employers on the Control of Noise at Work Regulations 2005 as:
Thermal comfort: -3 to 3
Carbon dioxide: less than 1000ppm
264
Sound level: 40 dBA to 70dBA
Illumination level: 800 lux to 1200 lux
With the above standards the solvers which adhere to the standard are:
Thermal comfort: GA, SA, PS, PSO, FMINCON, DE, GL,LGO,
glcCluster, glcSolve
Carbon dioxide: GA, SA, PS, PSO, FMINCON,GL, LGO, glcCluster,
glcSolve
Sound level: GA, SA, PS, PSO, FMINCON, DE, GL, LGO,
glcCluster, glcSolve
Illumination level: SA, PS, FMINCON, GL, LGO, glcCluster,
The following table gives a summary of all the criteria for the solvers:
Table.4.45. Summary of all the criteria for the solvers
Criteria GA SA PS PSO Fmincon DE GL LGO glcClus glcSolve
Result
according
to
ASHRAE
¾
=75%
4/4
=100%
4/4
=100%
¾
=75%
4/4
=100%
2/4
=50%
4/4
=100%
4/4
=100%
4/4
=100%
3/4
=75%
Consistency - - - - - -
Min-Run
Time - - - - - - - -
Min-
Evaluation - - - - - - - -
Simple
Algorithm - - - - - - - - -
Thus it is seen that the Pattern Search solver satisfies all the criteria and scores 100%
for its practicality in giving result according to ASHRAE, IES, Guidance for
employers on the Control of Noise at Work Regulations 2005, So the appropriate
algorithm, for optimization of thermal comfort is suggested as Direct search
algorithm & the solver is PATTERN SEARCH
265
4.4.7. CONCLUSION
Overall indoor environmental quality (IEQ) in terms of an occupant‘s acceptance has
not been considered in many residential buildings neither it is optimized. In this study, the
overall IEQ of residential apartments which are naturally ventilated at the Karunya University
quarters in Coimbatore was evaluated by 102 occupants in four aspects, namely thermal
comfort, indoor air quality , equivalent noise level and illumination level. All the offices
considered are naturally ventilated buildings. The results showed that the operative
temperature, carbon dioxide concentration, equivalent noise level and illumination level had
important effects on the overall IEQ acceptance. Empirical expressions were proposed to
approximate the occupant acceptance. The values are optimized using ten different non-
traditional optimization techniques.
Here, ten non-traditional optimization algorithms were presented. These include: GA,
SA, PS, PSO, GL, FMINCON, EA, LGO, glcCluster, glcSolve. A brief description of each
method is presented along with a pseudo code to facilitate their implementation. MATLab
programs were written to implement each algorithm. The IEQ problem for the Residential
buildings of the Karunya University was solved using all algorithms, and the comparative
results were presented.
Bibliography ANSI/ASHRAE55-2004. (2004). Thermal Environmental conditions for Human occupancy. Atlanda,
USA: American Society of Heating, Refrigerating and Air-Conditioning Engineers.
ANSI/ASHRAEstandard62-2007. (2007). Design for acceptable indoor air quality,. Atlanta: American
Society of Heating, Refrigerating and AIr-conditioning Engineers, Inc.,.
ASHRAE. (1989). Handbook-fundamentals, chapter 8,, Physiological Principles, Comfort and Health.
ASTM. (2003). litStandard guide for using indoor carbon dioxide concentrations to evaluate indoor air
quality and ventilation. D6245-98.
Auliciems.A. (1983). Psychophysical criteria for global thermal zones of building design. International
Journal of Biometerology , 69 -83.
Auliciems.A. (1984). Thermobile controls for human comfort. Heating and ventlating Engineeers ,
April/May 31-33.
Ayr, u., Cirillo, e., Fato, I., & Martellotta, F. (2003). A new approach to assessing the performance of
noise indices in buildings. Applied Acoustics , 64(2), 129-145.
266
Bean, A., & Bell, R. (1992). the CSP index : A practical measure of office jighting quality as perceived
by the office workers. Lighting Research and Technolgy , 24(2), 214-225.
Brager,G.S, & Dedear,R.J. (1998). Thermal adaptation in the building environment: a literature
review. Energy and Buildings , 27(1) 83 -96.
Bulysssen, P., & Cox, C. (2002). Indoor environmental quality and upgrading of European office
buildings. Energy and Buildings , 34(2),155-162.
Clausen, G., & Wyon, D. (2008). The combined effects of many different indoor environmental
factors on acceptability and office work performance. HVAC&R Research , 14(1), 103-113.
Clausen, G., Carrick, L., Fanger, P., Kim, S., Poulson, T., & Rindel, J. (1993). A comparative study of
discomft caused by indoor air pollution, thermal load and noise. Indoor Air , 3, 255-262.
Dan, J. (1993). Total Environmental Quality. In ASHRAE Transaction (pp. 960-967).
Elbeltagi, E., Tarek Hegazy, 1., & & Grierson, D. (2005). . Comparison among five evolutionary-based
optimization algorithms. . Advanced Engineering Informatics , 19, 43-53.
Fanger, P. (1970). Thermal comfort: Analysis and applications in Environmental engineering,. New
York: McGraw-Hill.
Fanger, P. (1995). Comments on 'a comparison of the predicted and reported thermal sensation vote
in homes during winter and summer'. Energy and Buildings , 22(1) 89.
Fanger, P., Olf, & Decipol. (1988). OLf and decipol: new units for perceived air qualtiy,. Building
Services engineering Research and Technolgy , 9, 155-157.
Fanger.P.O., T. (2002). Extension of the PMV model to non-air-conditioned buildings in warm
climates. Energy and Buildings , 34(6), 533-536.
Federspiel, C. (1882). Used-adaptable and minimum-power thermal comfort Control. Ph.D
Thesiis,MIT,Department of Mechanicla Engineering.
Fountain, m., Brager, G., Arens, E., Bauman, F., & Benton, C. (1994). Comfort Control for short-term
occupancy. Energy and Buildings , 211-213.
Gan, G., & Croome, D. (1994). Thermal comfort models based on Field measurements. ASHRAE
Transactions , 782-704.
Goldman, R. (1999). Extrapolating ASHRAE's comfort model. HVAC&R Research , 5(3),189-194.
Haghighat, F., & Donnini, G. (1999). Impace of psycho-social factors on perception of the indoor air
environment studies in 12 office buildings. Buildinh and Environment , 34, 479-503.
Han, J., Yang, W., Zhou, J., Zhang, G., Zhang, Q., & Moschandreas, D. (2009). A comparative analysis
of Urban and Rural residential thermla comfort under natural ventilation environment. Energy and
Buildings , 41(2), 139-145.
267
Hikmat, H., Hind, M., & Muna, H. (2009). Evaluating indoor environmental quality of public school
buildings in Jordan. Indoor and Built Environment , 18(1),66-76.
Houser, K., & Tiller, D. (2003). Measuring and Subjective response to interior lighting; paired
comparisons and semantic differential scaling. Lighting Research Technology , 35(3), 183-198.
Hui, P., Wong, L., & Mui, K. (2008). Using Carbon dioxide concentration to assess indoor air quality in
offices. Indoor Built Environment , 17(3), 213-219.
Hui, P., Wong, L., & Mui, L. (2006). Feasibility study of an express assesment protocol for the indoor
air quality of air-conditioned offices. Indoor and Built Environment , 15(4), 373 -378.
InternationalStandardISO7730. (1994). Moderate thermal environments- Determinatiob of the PMV
and PPD indices and specification of the conditions for thermal comfort.
ISO7730, B. (1995). Moderate thermla environments, Determination of PMV and PPF indices and
specification of the conditions for thermal comfort.
living-smartly.com. (n.d.). Retrieved September 09, 2009, from http://living-
smartly.com/2010/01/ac-and-ventilation-for-india-%E2%80%93-white-paper/
MacArthur, J. (1986). Humidity and predicted-mean-vote-based PMV-Based Comfort control.
ASHRAE Transactions , 15-17.
Mendell, M. (2003). Indices for IEQ and building related symposiums. Indoor Air , 13(4), 364-368.
Mui, K., & Wong, L. (2006). A method of assessing the acceptability of noise levels in air conditioned
offices. Building Serviecs Engineering Research and Technology , 27(3), 249-243.
Mui, K., & Wong, L. (2006). Acceptable illumination levels for office occupants. Architectural Science
review , 49(2), 116-119.
Mui, K., & Wong, L. (2006). Minimum Acceptable noise level for office occupants. Building Servies
Engineering Research and Technology , 27(3);249-254.
Mui, K., Wong, L., & Wong, Y. (2009). Acceptable noise levels for construction site offices. Building
Servies Engineering Research and Technology , 30, 87-94.
Mui, K., Wong, L., HUI, P., & Law, K. (2008). Epistemic evalution of policy influence on workplace
indoor air qualityof Hong Kong in 1996-2005. Building services Engineering Research and Technology
, 29(2),157-164.
Mui,K.M, & Wong,L.T. (2007). Neutral Temperature in subtropical climates- a filed survey in air-
conditioned offices. Building and environment , 42(2) 699 -706.
Mui.K.W, & Wong,L.T. (2007). Evaluation of neutral criterion of indoor air quality for iar conditioned
offices in subtropical climates. Building Services Engineering Research and Technology , 28(1), 23-33.
Naganoa, K., & Horikoshib, T. (2005). New comfort index during combined conditions of moderate
low ambient temperature and traffic noise . Energy and Buildings , 37, 287- 294.
268
Nicol, J., & Humphreys.A'. (2001). Adaptive thermal comfort and sustainable thermal standards for
buildings. conference Moving thermal comfort Standards into the 21st century, Windor,UK , 5 -8.
Olesen, & Bjarne. (2009). Productivity and Indoor Air qualtiy. Lyngby, Denmark: International centre
for Indoor Environment and Energy.
Oseland.N.A. (1995). Predicted and reported thermal sensation in climate chambers, offices and
homes. Energy and Buildings , 23(2) 105-115.
Pellerin, N., & Candas, V. (2004). Effects of steady-state noise and temperature conditions on
environmental perception and acceptability. Indoor Air , 14(2), 129-136.
Persily, A. (1997). Evaluating Building IAQ and Ventilation wiht indoor carbon dioxide. ASHRAE
Transactions , 103(2), 193=203.
Portney, L., & Watkins, M. (2000). Foundations of Clinical Research - Applications to Practise, second
ed.,. NJ: Prentice Hall Health.
Sherman, M. (1985). A Simplified Model of Thermal Comfort. Energy and Buildings , 8, 37-50.
Sofuoglu, S., & Moschandreas, D. (2003). The link between symptoms of office building occupants
and in-office air pollution:the indoor air pollution index. Indoor Air , 13(4), 332-343.
Toftem, J. (2002). Human response to combined indoor environment exposures. Energy and
Buildings , 34(6),601-606.
Viollin, S. (2003). Two examples of audio-visual interactions in an urban context. Proceedings of the
5th European Conference on noise control, (p. 73). Euronoise.
Wallace, L., Nelson, C., & Dunteman, G. (1991). Workplace characteristics associated wiht health and
comfort concerns in three office buildings in Washington,DC,Healthy Buildings 1991. American
Society of Heating, Refrigerating and Air-conditioning Engineers , 36-39.
Wang, Z. (2005). A field study of the thermal comfort in residential buildings in Harbin. Building and
Environment , 41(8(, 1034 - 1039.
WilliamFisk, & OlliSeppanen. (2007). Providing better indoor Environmental qualtiy brings economic
benefits. Proceedings of Clima 2007 well being indoors. Helsinki University of Technology, Espoo,
Finland.
Wong, L., & Leung, L. (2005). Minimum fire alarm sound pressure level for elder centers. Building
and Environment , 40(1), 125-133.
Wong, L., Mui, K., & Hui, P. (2008). A multivariate-logistic model of acceptance for indoor
environmental quality(IEQ) in offices,. Building and Environment , 43(1),1-6.
Wong, L., Mui, K., & Hui, P. (2006). A statistical model for characterizing common air Pollutants in
air-conditioned offices. Atmospheric Environment , 40(23),4246-4257.
269
Wong,L.T, Mui,K.W, Fong,N.K, & Hui,P.S. (2007). Bayesian Adaptive Comfort Temperature of air-
condioning system in subtropical climate. Building and Environment , 42(5) 1983-1988.
Yang, W., & Kang, J. (2005). Acoustic comfort evaluation in urban open public spaces. Applied
Acoustics , 66(2), 211-229.
Yoshino, H., Yoshino,Y, Zhang,Q, Mochida,A, Li,N, LI,Z, et al. (2006). Indoor thermal environmental
and energy saving for urban residential buildings in China. Energy and Buildings , 38 (11) 1308 -1319.
Top Related