基于小波分析的绝缘架空导线局部放电信号识别方法

A Method for Identifying Partial Discharge Signals of Insulated Overhead Conductors Based on Wavelet Analysis

现有的信号识别方法马修斯相关系数较低、识别效果差,因此研究基于小波分析的绝缘架空导线局部放电信号识别方法。运用安全光栅传感器(Sensor to Logic Converter,SLC)设备采集数据,提取局部放电类型的脉冲特征参数,利用小波变换对绝缘架空导线局部放电信号故障数据进行去噪。在反向传播(Back Propagation,BP)网络中对得到的数据进行预处理,根据误差反向传播对BP网络进行训练,提取不同种类的局部放电信号特征量完成识别。实验结果表明,所提方法的马修斯相关系数较高,可达到73%,能够对绝缘架空导线局部放电信号进行有效识别。

Due to the low Matthews correlation coefficient and poor recognition performance of existing signal recognition methods,a wavelet analysis based method for identifying partial discharge signals of insulated overhead conductors is studied.Using Sensor to Logic Converter(SLC)equipment to collect data,extract pulse characteristic parameters of partial discharge types,and use wavelet transform to denoise fault data of partial discharge signals in insulated overhead wires.Preprocess the obtained data in the Back Propagation(BP)network,train the BP network based on error backpropagation,and extract different types of partial discharge signal feature quantities to complete recognition.The experimental results show that the proposed method has a high Matthews correlation coefficient,which can reach 73%,and can effectively identify partial discharge signals of insulated overhead wires.