Litcius/Paper detail

A Robust Optimization Design Approach for Hybrid PM Machine Considering Asymmetric Uncertainties of PMs

Jiqi Wu, Xiaoyong Zhu, Deyang Fan, Zixuan Xiang, Lei Xu, Li Quan

2022IEEE Transactions on Magnetics14 citationsDOI

Abstract

This article proposes a robust optimization method for hybrid permanent magnet (HPM) machines, aiming to eliminate the influence of permanent magnets (PMs) asymmetric uncertainties. Exemplified by a 12-slots/10-poles (12s10p) HPM machine, the optimization method is specifically clarified. First, uncertainties of two types of PMs are comprehensively analyzed and the worst combination of these uncertainties are estimated. Then, the design of experiments (DOEs) technique is utilized to take samples, and the corresponding performance of samples with and without consideration about uncertainties are simulated by finite element analysis (FEA), respectively. After that, a dual-level kriging surrogate model is constructed. And the difference of the two levels, as a quantitative index of robustness, is added as an additional optimization objective. Finally, a multi-objective optimization algorithm non-dominated sorting genetic algorithm (NSGA-II) is carried out. The optimization results indicate that the proposed method can effectively reduce the additional torque tipple brought from asymmetric uncertainties of PMs. In addition, a prototype is manufactured and tested to further verify the method.

Topics & Concepts

Robustness (evolution)SortingComputer scienceFinite element methodRobust optimizationGenetic algorithmTorqueMulti-objective optimizationMagnetKrigingMathematical optimizationControl theory (sociology)AlgorithmMathematicsEngineeringMechanical engineeringPhysicsThermodynamicsMachine learningArtificial intelligenceStructural engineeringChemistryGeneControl (management)BiochemistryElectric Motor Design and AnalysisInduction Heating and Inverter TechnologyPiezoelectric Actuators and Control