Electromagnetic Vibration Reduction of Permanent Magnet Synchronous Motor Using Particle Swarm Optimization Based on Response Surface

In recent years, there is a growing demand for permanent magnet synchronous motors (PMSMs) with reduced electromagnetic (EM) vibration, so as to achieve better NVH performance. EM vibration is induced by a set of spatial and temporal EM force harmonics, and can be evaluated through a 2-step finite element (FE) simulation. However, with a large scale of input parameters, it is impractical to optimize the EM vibration based on direct FE simulation due to the time consumption. In this paper, a response surface is generated based on the historical FE simulation data, replacing the FE simulation procedure. Based on the response surface, the EM vibration amplitude is optimized using the Particle Swarm Optimization algorithm. Since the optimization process no longer requires FE simulation, the optimization speed is greatly accelerated, and the optimal results can be obtained in a few seconds. As a result, the EM vibration of the motor is significantly reduced, while other performances (e.g. torque and efficiency) remain unchanged.