基于支持向量机-蚁群算法的电气化铁路牵引负荷参数辨识

Identification of Traction Load Parameters of Electrified Railway Based on Support Vector Machine-ant Colony Algorithm

牵引负荷建模是含牵引供电系统的电力系统仿真中的重要组成部分,准确且符合实际的牵引负荷模型可以极大地提高电力系统仿真结果的可用性。近年来,随着牵引负荷相关实测数据的快速增长,一些基于实测数据的牵引负荷概率模型被提出,但是针对于这些概率模型的参数辨识方法的有效性还未被深入地研究,基于此对牵引负荷概率模型的参数辨识方法进行了深入分析,并根据分析结果提出了一种新的参数辨识方法:支持向量机-蚁群算法。该方法使用支持向量机对蚁群算法进行寻优指导,解决了原算法随机范围不可控的问题,使得算法能够将随机性控制在一个更为合理的范围内,实例仿真结果验证了该算法具有较高的准确性与通用性。

Railway traction load modeling is an important part of simulation of a power system that includes traction power supply system.Accurate and practical traction load model can greatly improve the applicability of power system simulation results.In recent years,with the rapid growth of traction load-related measured data,some probability models of traction load based on measured data have been proposed.However,the effectiveness of parameter identification methods for these probability models has not been thoroughly studied.Based on this,this paper conducted an in-depth analysis of the parameter identification method of the probability model of the traction load,and proposed a new parameter identification method based on the analysis result:support vector machine-ant colony algorithm.The algorithm used support vector machine to optimize the ant colony algorithm,and solved the problem of the uncontrollable random range of the original algorithm.The algorithm can control the randomness in a more reasonable range.The simulation results show the high accuracy and generalization performance of the algorithm.