基于LSSVM算法的10kV配电线路故障诊断模型

A Fault Diagnosis Model for 10 kV Distribution Lines Based on LSSVM Algorithm

常规的故障诊断模型以故障线路等值参数识别为主,无法深度剖析配电线路的故障位置,影响故障诊断准确性。因此,研究基于最小二乘支持向量机(Least Squares Support Vector Machine,LSSVM)算法的10 kV配电线路故障诊断模型。分析10 kV配电线路故障信息特征参数,找出分层性、冗余性、间断性以及不规范性等故障信息,短时间内收集大量的故障信息,为故障诊断提供基础条件。基于LSSVM算法建立线路故障诊断对数损失函数,对10 kV配电线路故障位置进行精准定位,通过对数损失函数将故障诊断总体误差控制在合理水平,满足线路故障诊断需求。通过仿真实验,验证了该诊断模型的诊断准确性较高,能够应用于实际生活。

The conventional fault diagnosis model mainly focuses on the identification of equivalent parameters of fault lines,which can not deeply analyze the fault location of distribution lines and affect the accuracy of fault diagnosis.Therefore,the fault diagnosis model of 10 kV distribution lines based on Least Squares Support Vector Machine(LSSVM)algorithm is studied.Analyze the characteristic parameters of fault information of 10 kV distribution lines,find out fault information such as layering,redundancy,discontinuity and irregularity,and collect a lot of fault information in a short time,which provides basic conditions for fault diagnosis.Based on LSSVM algorithm,the logarithmic loss function of line fault diagnosis is established,and the fault location of 10 kV distribution line is accurately located.The overall error of fault diagnosis is controlled at a reasonable level through the logarithmic loss function,which meets the requirements of line fault diagnosis.Through the simulation experiment,it is verified that the diagnostic model has high diagnostic accuracy and can be applied to real life.