基于BP神经网络的电抗器铁心柱轴向松动诊断

Fault Diagnosis of Axial Looseness of Reactor Core Column Based on BP Neural Network

油浸式高压并联电抗器是电力系统中重要的设备之一,而铁心柱轴向松动是高压并联电抗器常见的机械故障,可以通过测试振动信号对其进行有效诊断。探讨了高压并联电抗器振动机理,分析了并联电抗器振动信号的频谱特征。以油浸式高压并联电抗器模型为研究对象,提出了借助BP神经网络预测油浸式高压并联电抗器振动信号,将预测信号与实测信号相对比,诊断并联电抗器铁心柱松动的方法。研究结果表明,所提方法可以有效诊断油浸式高压并联铁心电抗器铁心柱轴向松动。

Oil-immersed high voltage shunt reactor is one of the important devices in power system.Axial loosening of core column is a common mechanical failure of high voltage shunt reactor,which can be effectively diagnosed by testing the vibration signal.The vibration mechanism of high voltage shunt reactor is discussed,and the spectrum characteristics of the vibration signal of shunt reactor are analyzed.Based on the model of oil-immersed high voltage shunt reactor,a prediction method for vibration signal of oil-immersed high voltage shunt reactor using BP network is presented.This method compares the predicted signal with the measured signal to diagnose the loosening of the core of the shunt reactor.The results show that the proposed method can effectively diagnose the axial loosening of core columns in oil-immersed high-pressure parallel core reactor.