基于粒子群算法-最小二乘支持向量机算法的磁化曲线拟合

Curve Fitting of Excitation Characteristics Based on Particle Swarm Optimization-Least Squares Support Vector Machine Algorithm

磁化曲线是强非线性函数,提高磁化曲线的拟合精度对含有铁磁材料的电气设备建模准确性至关重要。提出了一种基于粒子群算法-最小二乘支持向量机(PSO-LSSVM)算法的磁化曲线拟合方法。该方法用粒子群优化算法解决了最小二乘支持向量机(LSSVM)参数的选择问题。仿真结果显示PSO-LSSVM算法能获得最优的LSSVM参数,且采用PSO-LSSVM算法拟合的磁化曲线与实际测量的磁化曲线基本无偏差,拟合精度较高。

Magnetization curve was strongly nonlinear function. It was important to improve the accuracy of the magnetization curve fitting for the model of electrical equipment containing ferromagnetic material. Therefore, a method of magnetization curve fitting based on PSO-LSSVM algorithm was proposed. The method used particle swarm optimization algorithm to solve the LSSVM parameters selection problem. The simulation results showed that PSOLSSVM algorithm could obtain optimal LSSVM parameters and the magnetization curve used PSO-LSSVM algorithm has high fitting accuracy.