DETECTION OF VIBRATIONS : EXPERIMENT WITH AN EPILATOR

Project : limitations of the e-puck accelerometer to detect earthquakes

supervisor : Bahar Haghighat Valérie Berclaz Karine Sarrasin

PLAN

Introduction

Principe of an accelerometer

Approach of the project

Experiment

Epilator

Cell phone

Conclusion

INTRODUCTION

Goal of the project : recognize a vibration of a certain frequency by turning the LED’s on and blinking

Type of vibrations

Shocks : pulsed acceleration

Longer vibration : can be determined by the displacement, the velocity and the acceleration of a given point on a structure

PRINCIPE OF AN ACCELEROMETER

What is an accelerometer ?

Transform mechanical motion into an electrical signal

Several types used : piezoresistive, piezoelectric, capacitive….

Applications of accelerometer

Aircraft

Digital camera, cell phone

Laptop

PRINCIPE OF AN ACCELEROMETER

On the e-puck

MMA7260Q : capacitive micromachined accelerometer

Different sensitivities : 1.5g, 2g, 4g, 6g

In our case : 1.5g

Figure 1 : Electrical schema of a capacitive accelerometer

APPROACH OF THE PROJECT

First try : with a cell phone (Xperia Sony Ericsson) but

Not always the same response (depending on the battery of the phone)

Low amplitudes

Second try : with an epilator

Always the same response

Bigger amplitudes

Tests

With another cell phone

By shaking the e-puck

EXPERIMENT

Source of vibration : an epilator

Setup

Figure 2 and 3 : experiment’s setup

EPILATOR

Signal response

Uninterrupted signal

az values : around 800 (due to the gravity)

Figure 4 : signal of the epilator Figure 5 : Fourier transform of the z coordinate

Fourier transform

Contains a lot of frequency

Retained frequency : 26 Hz

CHOICE OF THE FILTER

Figure 6 : Design of the filter

Figure 7 : Attenuation of the amplitude Figure 8 : Fourier transform of the filtered az

RANGE OF DETECTION

Method : power spectral density for each direction

where a is the acceleration value and k is the filter order.

Final range :

PSD=∑n=0

n=i∣an∣

2

k

xmin xmax ymin ymax zmin zmax

0 70 40 100 0 20

VIDEO

Response of the e-puck when it is on the epilator

CELL PHONE

Smaller range of amplitude and periodic signal

Accelerations in z coordinate are well filtered

Figure 9 : Signal of the cell phone Figure 10 : e-puck on a HTC Wildfire S A510e

CELL PHONE

Frequency of the cell phone : around 18 Hz

After filtering : reduction of the peak

Figure 11 : Fourier transform of az Figure 12 : Fourier transform of the filtered az

Conclusion

E-puck detects the desired frequency

Empirical approach : experiments with different devices

Improvements:

Better filtering

Improve efficiency of the threshold

Thank you for your attention!

Questions?

Good times :