IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0...

4
lnternat. J. Math. & Math. Sci. Vol. i0, No.3 (1987) 609-612 RESEARCH NOTES AN EXISTENCE THEOREM FOR OPTIMAL CONTROL WITH NONSTANDARO COST FUNCTIONALS 609 R.N. MUKHERJEE and SHIV PRASAD athematics Section School of Applied Sciences Institute of Technology Banaras Hindu University Varanasi-221 005, INDIA (Received September 21, 1983) ABSTRACT. An existence theorem for optimal control is obtained for a general nonstan- dard cost functional of fractional type in this work. As an application of our result we can derive an existence theorem for optimal control given by M. B. Subrahamanyam for a cost functional, which is a ratio of two given integral cost functionals. KEY WORDS AND PHRASES. Nonstandard cost functional, optimal control. 1980 AMS SUBJECT CLASSIFICATION CODE. 49B27, 65KI0. I. INTRODUCTION. Consider the n-dimensional system x A(t)x + B(t)u, x(t o) 0 (I.I) for t [t t l] t < ; where A(t) and B(t) are matrices of size nxn and o n x r respectively. We lay some further restrictions on x(t) and u(t) later. We minimize a functional of the following type F (u) F(x,u) (1.2) [F 2 (x) ]a where a > 0 and u is a measurable control. We make the following assumptions: (a) A(t) and B(t) are continuous n xn and n x r matrix functions respectively. (b) (i) F (.) is convex in u and F F 2 both are continuous and (ii) IF1 (u) < for all admissable u. (c) F l(u) , a lull a > 0 p > and F2(x) . 0 along any x(t) which is response P to some admissable u(t). (d) For each K < such that if lull < K then p IF 2(x) < (1.3) for any admissable u whose trajectory obeys any constraints imposed. (e) There exists k 0 such that for every c > 0

Transcript of IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0...

Page 1: IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0 along any x(t) which is response P to some admissable u(t). (d) For each K

lnternat. J. Math. & Math. Sci.Vol. i0, No.3 (1987) 609-612 RESEARCH NOTES

AN EXISTENCE THEOREM FOR OPTIMAL CONTROL WITHNONSTANDARO COST FUNCTIONALS

609

R.N. MUKHERJEE and SHIV PRASAD

athematics SectionSchool of Applied Sciences

Institute of TechnologyBanaras Hindu UniversityVaranasi-221 005, INDIA

(Received September 21, 1983)

ABSTRACT. An existence theorem for optimal control is obtained for a general nonstan-

dard cost functional of fractional type in this work. As an application of our result

we can derive an existence theorem for optimal control given by M. B. Subrahamanyam

for a cost functional, which is a ratio of two given integral cost functionals.

KEY WORDS AND PHRASES. Nonstandard cost functional, optimal control.

1980 AMS SUBJECT CLASSIFICATION CODE. 49B27, 65KI0.

I. INTRODUCTION.

Consider the n-dimensional system

x A(t)x + B(t)u, x(to) 0 (I.I)

for t [t tl] t < ; where A(t) and B(t) are matrices of size n x n ando

n x r respectively. We lay some further restrictions on x(t) and u(t) later.

We minimize a functional of the following type

F (u)F(x,u) (1.2)

[F2 (x) ]a

where a > 0 and u is a measurable control. We make the following assumptions:

(a) A(t) and B(t) are continuous n x n and n x r matrix functions respectively.

(b) (i) F (.) is convex in u and F F2 both are continuous and

(ii) IF1 (u) < for all admissable u.

(c) Fl(u) , a lull a > 0 p > and F2(x) . 0 along any x(t) which is responseP

to some admissable u(t).

(d) For each K < such that if lull < K thenp

IF2(x) < (1.3)

for any admissable u whose trajectory obeys any constraints imposed.

(e) There exists k 0 such that for every c > 0

Page 2: IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0 along any x(t) which is response P to some admissable u(t). (d) For each K

610 R.N. MUKHERJEE AND S. PRASAD

Fl(cu) Ck

F1(u)

F2(cx) ck/aF2(x) (1.4)

By (1.3) this assumption implies that for every c 0 F(cx, cu) F(x, u)

(g) There exists an admissable control, the trajectory of which satisfies the imposed

constraints and is such that

F2(x) > 0 (1.5)

i(h) If {u is a sequence of functions in L [t ,t I] going weakly to u inp o o

L [t ,tl thenp o

Fl(u) < lim inf Fl(uI) (1.6)

We call a constraint regular if the following two conditions hold:

(I) (x,u) satisfies the constraint => (cx,cu)

satisfies the constraint for every c > 0

i o(2) Let (xl,uI), (x2,u2), be admissable pairs such that u u weakly in

L (to

tl). Suppose (xn un) satisfies the constraint for each n >.. Then (x,u)

obeys the constraint.

REMARK. Observe that in the proof of theorem I. below it has been shown that u is

necessarily admissable.

First of all we prove the following lemma which is the modificaiton of Proposition

I.I of [I] in the general setting.

LEMMA I.I.

Consider all pairs that obey (1.1) and the constraints. Assume that all the con-

straints are regular, and let

F (u)I if F(x,u) inf -/----- (1.7)

(x,u) (x,u) IF2 (x) ]e

(% is well defined by assumption (b) part (ii) and (g)). Also let

inf F1(u) J subject to [Fl(x)] M o. (1.8)U

Then % J/M.

PROOF. One can easily see that J/M > % To prove the reverse inequality, let

be such that F(,) < + for some > o. Let [F2()] (< by assumption

(b) part (ii), (c) and (d)) and (M/) I/kThen ( ,B ) obeys all the con-

straints by regularity of the constraints and by assumption (e), [F2(u)]e M and

F(, B) < + e By (1.8), J/M < + since is arbitrary, the conclusion of

the lemma follows.

THEOREM 1.1.

Consider the system (I.I) and (1.2) along with assumptions (a) to (h). Also

assume that the constraints on x and u are regular. Then there exists a control

among all admissable controls that minimizes (1.2).

PROOF. By lemma i.i it suffices to exhibit a minimizing control over all admlssable

controls for which [F2(x)]e M > o and trajectories of which satisfy (I.I) and all

Page 3: IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0 along any x(t) which is response P to some admissable u(t). (d) For each K

EXISTENCE THEORF#I FOR OPTIMAL CONTROL 611

the constraints. Let J inf Fl(u) subject to [F2(x)]a

M Choose {(xi,ui)}u

isuch that lira Fl(ul) J with [F2(xl)]a M for each i. By assumption (c) u

i+

form a bounded sequence in L (to

t and hence a subsequence still denoted by {ui}p

o oconverges weakly to some u in L (t

ot Let x be the response of (I I) to u

pi

xo

By assumption (a) and by the convergence, x (t) (t) for all t[to,tl (see ref.

x[23). By regularity of constraints, (t) obeys all the constraints. Assumption (b)

i oF2(xI) F (x) as i (Since x x in L (t

ot )) Sinceimplies p

for some K < By assumption (d),

By assumption (h)

[F2(x)]a lira [F2(xI)] Mi+

F (u) < lira inf F1(ui) J

From which it follows that u is the minimizing control.o

APPLICATION. If we specialize the functional F(x,u) as given by F(x,u)

otl l(x(t),t) dt

tl 2 (u(t), t)dt]e

Where I and 2 satisfy the condition as given in ([I], theorem I.I) we get the

existence theorem of [I] as a particular case of our thoerem.

DISCUSSIONS. The general method, what we have discussed, can be extended to cover

other variants of the functional given by Subrahmanyam [I]. For example, we can con-

sider functional of the form

F(x,u) . 1 (x(t),t) dt

l(U(t),t)]el [{o n(U(t)’t)dt]n

where I, {i}.=l all have to satisfy certain restrictions analogous to the restric-

tions given as in [I].

REFERENCES

i. SUBRAHMANYAM, M.B. On Applications of ContrQl Theory to Integral Inequalities II,Siam J. Con__t_r__o_an_d_._O_t.im_ization, 19(1981), 479-489.

2. LEE, E.B. and MARKUS, L. Foundations of Optimal Control Theory, John Wiley, NewYork, 1967.

3. SUBRAHMANYAM, M.B. Necessary Conditions for Minimum in Problems with NonstandardCost Functionals, J. Math. Anal. Appl. 60(1977), 601-616.

4. SUBRAHMANYAM, M.B. On Applications of Control Theory to Integral in Equalities,J. Math. Anal. Appl. 77(1980), 47-59.

Page 4: IF1 - core.ac.uk · (ii) IF1 (u) < for all admissable u. (c) Fl(u), a lull a >0 p and F2(x). 0 along any x(t) which is response P to some admissable u(t). (d) For each K

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

MathematicsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Mathematical Problems in Engineering

Hindawi Publishing Corporationhttp://www.hindawi.com

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Mathematical PhysicsAdvances in

Complex AnalysisJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

OptimizationJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Operations ResearchAdvances in

Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of Mathematics and Mathematical Sciences

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Algebra

Discrete Dynamics in Nature and Society

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Decision SciencesAdvances in

Discrete MathematicsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com

Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Stochastic AnalysisInternational Journal of