Multi-Objective Optimization of the Rotor Design to Improve the Acoustic Behavior of High Power Density Interior Permanent Magnet Synchronous Machines

摘要

The shape of the flux barriers in IPMSM has a significant influence on the acoustic behavior as well as on equivalent stress and torque. The magnetic field is the main source of noise excitation of electric machines. In this paper a method is proposed which constantly fits an approximation model with simulation results to improve its accuracy of predicting torque as well as NVH. Due to significantly large simulation time a non-dominated sorting genetic algorithm searches for new Pareto optimal designs based on approximation model predictions. By optimizing the shape of flux barriers in IPMSM this method minimizes NVH and maximizes torque as a trade-off between electromagnetism and structural strength in a weakly coupled simulation. In this case the maximum equivalent stress is used as a constraint.