轮毂电机多目标优化设计与温升估计

According to the high torque density requirement for electric drive system of special vehicle,a multi-physics-based multi-objective optimization design and temperature rise estimation method for in-wheel electric machine is proposed to effectively enhance the peak torque and efficiency,reduce the torque ripple and prevent the in-wheel electric machine from overheating. The electromagnetism finite element model and loss models of in-wheel electric machine are established based on the vehicle mission profile. Non-dominated sorting genetic algorithm-Ⅱ(NSGA-Ⅱ) is applied to optimize the peak torque, torque ripple,efficiency,and heat exchange area of winding. Based on the key geometry parameters and losses characteristics obtained,a temperature rise estimation model for electric wheel lumped parameter thermal network including the electric machine is established to estimate the temperature rise and distribution characteristics under typical working conditions. The accuracy of the temperature rise estimation model is validated through a testbench. The result shows that the peak torque and its efficiency of optimized in-wheel electric machine are increased 5. 2% and 1. 15%,respectively. The root mean square error of the estimated temperature is less than 4. 3 ℃ compared with experimental result,and the calculation effort is dramatically reduced.