Robust Optimization Design for Permanent Magnet Machine Considering Magnet Material Uncertainties

In this paper, a robust optimization design method considering magnet material uncertainties is proposed for permanent magnet (PM) machines. The key of this method is to improve the machine design robustness to resist the uncertainties influence from PM material, including material diversity, manufacture error, and assembly accuracy. To verify the proposed optimization method, a double permanent magnet machine is selected as an example. During the implementation of optimization design, firstly, the robust optimization model is constructed based on the analysis of PM uncertainties. Then, the Kriging surrogate model and non-dominated sorting genetic algorithm II (NSGA-II) are adopted to solve the optimization model. Finally, to verify the effectiveness of the method, the machine performance and corresponding reliability are evaluated, and a prototype is manufactured and tested. Both simulation and experimental results indicate that the proposed robust optimization method can effectively reduce the influence of magnet material uncertainties on PM machine performance.