An improved method of induction motor parameter identification based on genetic algorithm

Xiao Xi,Wang Yating,Xu Qingsong,Shi Yuchao

2013-01-01

Abstract:To aim at the defects in the traditional parameter identification method based on genetic algorithms of induction motor, an improved method is proposed. By detecting motor stator voltage, current and rotor speed signals, a motor model without flux component is obtained in this method combining with the mathematical model of the induction motor. And a different fitness function is employed from which in traditional genetic algorithm, thus, the complicated procedures of flux observation is avoided, while the disturbance for flux observer need motor parameters is reduced and the accuracy of parameter identification is improved. The simulation and experimental results show that four main parameters of the induction motor can be identified simultaneously in this method, moreover, the four identified parameters can converge to the true value in short time under different operating conditions, and high identification accuracy is also guaranteed. The effectiveness of the method is demonstrated.