基于SVM和遗传算法的新型直线电机结构参数优化

建立了一种新型圆筒直线电机的有限元模型,并验证了有限元模型的正确性。基于正交和随机方法安排了有限元仿真试验,获得了用于非线性电磁建模的样本空间。进一步采用支持向量机回归建模方法,建立了直线电机的非线性数学模型,而后采用遗传算法进行了直线电机结构参数多目标优化,得出最优结构参数组合。有限元仿真验证表明,基于支持向量机和遗传算法的优化方法用于研究该种新型直线电机电磁结构参数优化是可行的。

Parameter optimization of novel cylinder type linear motor based on SVM and genetic algorithm

The FEM(Finite Element Method) model of a novel linear motor was established,and then its correction was proofed.To obtain sample space used to establish the nonlinear regression model,the FEM simulation tests were arranged with the methods of orthogonal and random experimental design.Moreover,the nonlinear mathematical model of a novel linear motor was established using nonlinear regression method of the Support Vector Machine(SVM).Then the optimum structure parameters was obtained by optimizing the structure parameters with Genetic Algorithms(GA).At last,the FEM simulating results show that the parameter optimization method based on SVM and GA is feasible for researching the new cylinder type linear motor.