基于改进SVM的高速铁路10 kV电力贯通线短路故障自动识别  

Automatic Identification of Short-Circuit Faults in High-Speed Railway 10 kV Power Transmission Lines Based on Improved SVM

作　　者：章波 ZHANG Bo(Construction Headquarters of Hangzhou Railway Hub Project,China Railway Shanghai Group Co.,Ltd.,Hangzhou,Zhejiang 310008,China)

机构地区：[1]中国铁路上海局集团有限公司杭州铁路枢纽工程建设指挥部,浙江杭州310008

出　　处：《自动化应用》2025年第6期150-152,共3页Automation Application

摘　　要：电力贯通线的信号具有非线性和非平稳的特性,导致对短路故障的识别效果难以得到保障,为此,提出基于改进SVM的高速铁路10 kV电力贯通线短路故障自动识别研究。借助VMD构建原始的高速铁路10 kV电力贯通线信号的约束变分问题,并通过引入二次惩罚项和拉格朗日乘法算子求解,将复杂信号分解为多个具有特定中心频率和带宽限制的IMF;识别其短路故障状态时,采用了基于一对一(1-a-1)的支持向量机(SVM)算法,综合各分类器的输出,确定最终故障状态。测试结果表明,设计方法有效识别了设置的6种不同短路故障,具有良好的应用效果。Due to the nonlinear and non-stationary characteristics of signals in power transmission lines,it is difficult to ensure the recognition effect of short-circuit faults.Therefore,a research on automatic identification of short-circuit faults in highspeed railway 10 kV power transmission lines based on improved SVM is proposed.Using VMD to construct the constrained variational problem of the original high-speed railway 10 kV power transmission line signal,and solving it by introducing a quadratic penalty term and Lagrange multiplication operator,the complex signal is decomposed into multiple IMFs with specific center frequencies and bandwidth constraints.When identifying its short-circuit fault state,a Support Vector Machine(SVM)algorithm based on one-to-one(1-a-1)was used to synthesize the outputs of various classifiers and determine the final fault state.The test results show that the design method effectively identifies six different short-circuit faults and has good application effects.

关 键 词：改进SVM 自动识别 约束变分问题 二次惩罚项 拉格朗日乘法算子 1-a-1支持向量机 

分 类 号：TM41[电气工程—电器]