基于有限元的永磁无刷直流电机转矩优化设计

摘要

在分析永磁无刷直流电机电磁转矩的基础上,以抑制电机转矩脉动为目标,提出了有限元与最小二乘支持向量机( LS-SVM)及粒子群优化算法相结合的永磁无刷直流电机设计方法,建立了永磁无刷直流电机有限元模型.经有限元计算分析,得到了极弧系数、永磁体厚度和气隙长度对电机转矩脉动的影响曲线,生成了LS-SVM训练样本,经训练得到了永磁无刷直流电机LS-SVM模型,通过有限元计算和LS-SVM模型拟合结果的比较,验证了模型的准确性.利用粒子群优化算法对永磁无刷直流电机LS-SVM模型进行优化,得到了极弧系数、永磁体厚度和气隙长度的最优参数,经有限元计算结果验证了优化的有效性.%In order to keep down torque ripple. The thesis analyzed the torque of permanent magnet brushless DC motor (PMBLDCM) and put forward a new method of finite element and least squares support vector machine ( LS-SVM) and partide swarm optimization ( PSO). A infinite element model of PMBLDC has been found. After the anylisis of model, the influence of pole embrace and thickness of PM and gas length on torque ripple. The training sample has been gained. The LS-SVM model of PMBLDCM found. After contrast of two results, the accuracy of model was verified. Then optimizing model of PMBLDCM LS-SVM was obtained. The optimizing parameter of PMBLDCM has been found, the validity of model was verified.