Numerical multi-objective optimization of a squirrel cage induction motor for industrial application

Abstract

This paper reports the optimization process of a MW-size, medium voltage, squirrel cage induction motor for industrial applications. Starting from an existing construction, the aim of the optimization is to maximize the induction machine power density, while fulfilling high demanding, specific international Oil&Gas industry regulations. For every considered design, the motor performance is estimated by means of an algorithm based on T-type equivalent circuit. The circuit parameters are determined with a combination of magnetostatic and time harmonic Finite Elements Analysis (FEA), together with analytical formulas. This approach allows the evaluation of the precise parameters dependency on punctual current and frequency values, and allows good performance prediction ability. The multi-objective optimization process input are geometrical variables related to the motor electromagnetic configuration having direct impact on the output performance, is based on Genetic Algorithms (GA), and has been developed with the commercial software ModeFRONTIER®.