Multiobjective Deterministic and Robust Optimization Design of a New Spoke-Type Permanent Magnet Machine for the Improvement of Torque Performance
Guohai Liu, Yanyang Wang, Qian Chen, Gaohong Xu, Chengyan Song
Abstract
This article proposes a comprehensive multiobjective optimization design method for a new spoke-type permanent magnet machine with auxiliary salient poles for the improvement of torque performance. The key of this method is to consider the deterministic and robust performances simultaneously. First, the sensitivity analysis is adopted to classify design parameters into different levels and the corresponding optimization techniques are skillfully employed to realize the tradeoff design based on the appropriate model constructed by the response surface methodology. However, inevitable perturbations especially during the mass production will result in big fluctuations of torque performances and affect the reliability and quality of the machine, which are often neglected in deterministic design. Therefore, a robust design method, i.e., the design for six sigma, is presented to address the parametric uncertainties. Afterward, the Monte Carlo analysis is used to obtain the statistic samples and the barebones multiobjective particle swarm optimization algorithm is employed to find the robust optimal solution. The corresponding electromagnetic performances and reliability of the products are compared and analyzed. Finally, a prototype is manufactured and tested to verify the validity of the proposed optimization design method.