Multi-Objective Robust Optimization of a Low-Cost and Efficient IPMSM for Battery Electric Vehicle

This paper proposes a surrogate-assisted robust multi-objective optimization (MOO) approach for interior permanent magnet synchronous machines (IPMSMs). Traditional MOO methods often overlook mechanical constraints and do not focus on preserving solution diversity within the optimization process, resulting in impractical designs. By integrating robustness and diversity preservation into an effective MOO process, dimensional uncertainties in manufacturing may be managed and diversity among solutions can be maintained. This study optimizes a 165 kW peak power IPMSM found in a battery electric vehicle (BEV), balancing permanent magnet (PM) material cost and average efficiency using a robust non-dominated sorting genetic algorithm III (NSGA-III) for the preservation of diversity among Pareto-optimal solutions. Results demonstrate that incorporating robust optimization metrics into the MOO process can reduce manufacturing costs while preserving the intended machine performance, offering a promising approach for future traction motor design.