Topology Optimization Empowers the Design of Interior Permanent Magnet (IPM) Motors

摘要

Interior permanent magnet (IPM) motor optimization has been a challenging part of machine design process due to the different materials and shapes that are involved. Design experiences are not always available for the similar topology and operating conditions. This paper explored optimized designs for I-shape, single V-shape, double V-shape, and Delta-shape rotor topologies by utilizing genetic algorithm (GA) with the On/Off method based on the normalized Gaussian network (NGnet), which does not depend much on experiences. To evaluate the effectiveness of this approach, this optimization process is also applied to the well-known Prius 2010 motor and improved the torque performance.