An Effective Identification of the Induction Machine Parameters using a Classic Genetic Algorithm, NSGA II and θ-NSGA III

To remain competitive, the manufacturing industry is using computer processing power to innovate, develop and optimize new cost-efficient production strategies. This is the reason why optimization of automation systems is deployed to improve productivity, quality and robustness of the production. The different existing goals of optimization as the control machine, management of the power consumption, design of electrical installation and prediction of motor faults lead to the necessity of estimating the induction machine parameters (the stator and rotor resistances, the stator and rotor inductances and the magnetizing inductance). To these ends, researchers and companies are investigating efficient methods to identify these parameters. In this paper, we propose an effective method for the induction machine parameters identification based on the new theta-NSGA III genetic algorithm. A comparison between a classic single objective genetic algorithm (GA) and two well-known multiobjectives GAs (NSGA II and theta-NSGA III) is performed. Our results show that the multi-objective GA theta-NSGA III provides a better estimation of parameters than the classic single objective GA and the multi-objective GA NSGA II.