Multi-objective Optimization of Electric Motors with a Kriging Surrogate Model

Since the electric drive systems are already used in numerous cars and other vehicles as well, the widely varying fields of applications require custom motor design. The most efficient tool for different specified motor designs is the multi-objective optimization tool based on validated simulations. An electrical motor optimization system with a kriging surrogate model based on FEM simulations is developed. The application of this system is presented in this paper. The models and the simulations were created in ANSYS Maxwell and MATLAB. The optimization was performed with the multi-objective genetic optimization algorithm by MATLAB which can be controlled by a simple input-output MATLAB interface.