基于支持向量化与混沌搜索的新型直线电机参数优化

Parametric Optimization of Novel Cylinder Linear Motor Based on SVM and Chaotic Search

以一种新型圆筒直线电机为研究对象,推导出了适合其结构特征的电磁方程;采用变分方法,建立了电机轴对称涡流场的有限元模型;通过仿真计算,得出了电机的启动推力和电流;样机试制和输出性能测试表明了有限元模型的正确性。基于正交和随机混合试验设计方法安排有限元仿真试验,获得了用于非线性电磁建模的样本空间;通过支持向量机对结构参数与计算结果的输入输出关系进行非参数回归建模,然后用混沌搜索方法对回归函数进行优化,得出最优目标结构参数。有限元仿真验证表明,基于支持向量机和混沌搜索的优化方法用于研究新型直线电机电磁结构参数是可行的。

Through analyzing the structure of a novel cylinder linear motor, the axisymmetric electromagnetic equation, which fits for its structure characteristics, was established and then, the axisymmetric vortex FEM model was set up with the variation theory. Firstly, the pushing forces and currents were calculated from the simulation results. The model linear motor was made with optimum parameters, and the testing results show that the FEM model is correct. 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,by means of SVM nonlinear regression, the input and output relation among parameters with calculation results was reconstructed. Then we opti- mized the structure parameters with chaotic search method, and the optimum structure parameters was obtained. At last, FEM simulation validates that the parameter optimization method based on SVM and chaotic search is feasible for researching the new cylinder type linear motor.