Geometry Optimization of IPM Machines Using Orthogonal Experimental Design Method and Artificial Neural Network

Abstract

Interior permanent magnet (IPM) machines are widely used in various applications due to many advantages such as high power density, high efficiency, low manufacturing cost and wide constant power operating range. However, high torque pulsation of IPM machines is a major drawback especially when concentrated windings are adopted. Rotor geometry has significant effect on torque pulsations in IPM machines. In this paper, a novel technique for rotor geometry optimization is proposed. Interactive effects of different rotor parameters on machine performance are studied using orthogonal experimental design (OED) method. An artificial neural network (ANN) is modeled thereafter to find the optimum rotor geometry design. The validity of the results is verified by finite element analysis (FEA).