Multi-objective optimization design of dual-winding permanent magnet motor based on sensitivity analysis

Abstract The primary challenge in reducing electromagnetic vibration through structural parameter optimization lies in efficiently and systematically adjusting multiple parameters without compromising the motor’s output performance. To address this problem, the optimization objective function is formulated based on key performance indicators of the DWPMSM (Dual-winding Permanent Magnet Motor), including the peak radial electromagnetic force amplitude, average torque, and torque ripple. Then, the sensitivity analysis method is used to evaluate the importance of design parameters to the optimization objectives. The design parameters with weak sensitivity retain the initial value, and the design parameters with strong sensitivity are selected in the next optimization design. Subsequently, the EA algorithm is employed to optimize the motor design model, resulting in an improved set of motor dimensions. Next, the EA algorithm is applied to refine the motor optimization model, yielding an optimized set of motor dimensions. To confirm the success of the optimization, a finite element analysis is performed to compare the motor’s torque, torque ripple, and radial electromagnetic force both before and after the optimization process.