A Hybrid Algorithm Using Shape and Topology Optimization for the Design of Electric Machines

Abstract

The conventional topology optimization (TO) algorithm for an electric machine has a critical problem as the derivation of the impractical model to manufacture. In the case of a conventional step optimization, such as shape optimization (SO) after TO, the optimized model by TO is seriously distorted. To solve these problems, a hybrid optimization algorithm that conducts SO and TO simultaneously with the pre-processing and the post-processing is proposed in this paper.