Shape Optimization of Rotor in Interior Permanent Magnet Motor Based on Topology OptimizationMethod Using Normalized Gaussian Network

Abstract

This paper presents a new topology optimization method based on normalized Gaussian network (NGnet). In this method, the machine region is subdivided into small elements whose material states are determined from the output of NGnet so that the objective function is extremized under the given constraints. The present method is applied to shape optimization of the rotor in an interior permanent magnet motor. The rotor shape is optimized to minimize the torque ripple while keeping the average torque. It is shown that the optimization halves the torque ripple while the average torque remains unchanged. This result was validated through a comparison between the computed and measured torques.