The Application Of Genetic Algorithm In The Steady-state Model Of Induction Motor Parameter Identification

This paper presents a nonlinear estimation algorithm to determine the steady-state equivalent circuit parameters of a three-phase induction machine. The estimation procedure is based on the steady-state characteristics versus slip curves. The sum of squares of differences between calculated and experimental characteristics is employed as the fitting criterion. The machine parameters are obtained by minimizing the least-squares cost function using genetic algorithms.The first chapter quotes three methods for determining the steady-state equivalent circuit parameters of an induction machine. They are the classical locked-rotor and no-load tests, the direct modification method and the recursive least-squares method. Then a new parameter estimation method based on genetic algorithms is presented.In Chapter 2, the characteristics functions are firstly derived. Two equivalent circuits are employed as machine models to describe the steady-state operation of a squirrel-cage induction machine and a double-cage or deep-bar induction machine. Then a digital simulation program developed with object-oriented analysis and design techniques is introduced.In Chapter 3, a brief introduction is presented about biological foundations and characteristics of genetic algorithms. Then basic techniques employed in programming genetic algorithms are determined.Chapter 4 is the main part of this paper. Freedom degrees of the parameters of the characteristics are firstly discussed to ensure that there is one and only solution to the problem. Then local searching method and pseudo-parallelism are both adopted to improve the simple genetic algorithm for avoidance of the premature convergence and enhancement of the estimation precision. Finally, the influences of the measurement noise and the range of slip on the estimation precision are evaluated .In Chapter 5, the proposed algorithm is applied to estimate the parameters of a squirrel-cage induction machine. Firstly, the experimental phase current and inputpower data obtained from load test are employed to estimate the parameters inIIrated condition. Then, the phase current and input power data in locked rotor condition are used to track the variation of the parameters of the secondary coil with slip frequency. The classical method and the direct modification method are also used to estimate the parameters of the induction machine in comparison with this technique.