摘要
介绍利用最小二乘支持向量机的回归理论对牵引电机磁化曲线进行拟合,从而建立准确的电力机车牵引电机模型的方法。针对最小二乘支持向量机参数选择耗时长的问题,提出一种基于三步搜索技术的参数选择方法。理论分析及仿真结果表明,该方法可优化选择最小二乘支持向量机的参数,并可提高最小二乘支持向量机的建模速度。将该方法用于电力机车牵引电机建模的参数选择,仿真结果表明,该方法建立的电力机车牵引电机模型精确度高,可用于对电力机车主电路性能及控制策略的研究。
Considering the nonlinear magnetization characteristics of the traction motor of the electric locomotive,the paper presents the least squares support vector machine(LS-SVM) to acquire the fitting of magnetization curve.In order to shorten the modeling time,the three step search method is used to get the optimized parameters of LS-SVM.Theoretical analysis and simulation results show that the method facilitates optimization of the parameters and shortening of the modeling time of LS-SVM.The traction motor simulation model is built on the basis of the improved LS-SVM.The simulation results show that the proposed traction motor model of the electric locomotive is of high accuracy and is applicable in research of the performance and control strategy of the main circuit of the electric locomotive.
出处
《铁道学报》
EI
CAS
CSCD
北大核心
2011年第3期23-27,共5页
Journal of the China Railway Society
基金
国家自然科学基金资助项目(51007023)
关键词
电力机车
曲线拟合
最小二乘支持向量机
三步搜索法
建立模型
electric locomotive
curve fitting
least squares support vector machine
three step search method
modeling