Multiobjective Optimization of a Six‐Phase Axial‐Flux Sandwich PM Motor via Sensitivity‐Guided Hybrid Particle Swarm Optimization

This paper introduces a novel six‐phase axial‐field flux‐switching sandwiched permanent magnet machine with a twisted structure (AFFSSPM‐TS), designed to overcome key limitations in conventional flux‐switching machines. The proposed configuration features a dual‐stator, dual‐rotor arrangement with modular U‐shaped stator segments and circumferentially opposed PMs, enhancing flux‐focusing effects and torque density. A twisted rotor structure is incorporated to significantly reduce cogging torque and back‐EMF harmonics while maintaining high average torque. For efficient multiobjective optimization, a sensitivity‐guided hybrid particle swarm optimization (SG‐HPSO) algorithm coupled with a Kriging surrogate model is developed, enabling rapid and accurate exploration of the design space while reducing computational cost. Sensitivity analysis identifies key parameters, which are then optimized hierarchically. Experimental validation via a fabricated prototype confirms substantial performance improvements: a 14% increase in average torque, a 41% reduction in cogging torque, and an 18% improvement in full‐load efficiency compared to baseline designs. The proposed AFFSSPM‐TS machine demonstrates strong potential for high‐performance applications such as electric vehicles, offering enhanced torque density, reduced torque ripple, and improved fault tolerance.