Vol.6, Special Issue, 2007 pp. 169-184
UDC 681.586:519.863(045)=111

OPTIMISED SENSOR CONFIGURATIONS FOR A MAGLEV SUSPENSION SYSTEM
Konstantinos Michail, Argyrios C. Zolotas, Roger M. Goodall
Department of Electronic and Electrical Engineering, Loughborough University, Loughborough, England, United Kingdom
e-mail: k.mihail; a.c.zolotas; g.m.goodall @lboro.ac.uk

Abstract. This paper discusses a systematic approach for selecting the minimum number of sensors for an Electromagnetic suspension system that satisfies both optimised deterministic and stochastic performance objectives. The performance is optimised by tuning the controller using evolutionary algorithms. Two controller strategies are discussed, an inner loop classical solution for illustrating the efficacy of the evolutionary algorithm and a Linear Quadratic Gaussian (LQG) structure particularly on sensor optimisation.
Key words: Sensor optimisation, MAGLEV suspensions, EMS optimisation, genetic algorithms, Kalman filter, evolutionary algorithms

OPTIMALNA SENZORSKA KONFIGURACIJA ZA MAGLEV LEVITACIONI SISTEM SUSPENZIJE
Ovaj rad se bavi sistematskim pristupom u odabiranju minimalnog broja senzora za sistem elektromagnetne levitacione suspenziju koji zadovoljava obe ciljeve determinističke i stohastičke optimalne performanse. Performansa je optimalna podešavanjem upravljanja uz pomoć razvojnih algoritama. Proučavaju se dve strategije upravljanja, unutrašnja petlja klasičnog rešenja za ilustraciju efikasnosti razvojnih algoritama i Linearne Kvadratne Gausijeve (LQG) strukture posebno za optimizaciju senzora.
Ključne reči: Optimizacija senzora, MagLev suspenzija, EMS optimizacija, genetski algoritmi, Kalmanov filter, razvojni algoritmi