基于PT在线注入信号及神经网络跟踪小波变换模极大值变化的配电网故障定位方法  

Distribution network fault location method based on PT injecting signal and neural network fault diagnosis system tracking the character of wavelet transform modulus maxima

作　　者：蔡维[1] 张雪松[2] 刘承志[3] 杜继伟[3] 

机构地区：[1]广西电力工业勘察设计研究院,广西南宁530023 [2]平朔煤炭工业公司矿区动力中心,山西朔州036800 [3]西南交通大学电气工程学院,四川成都610031

出　　处：《继电器》2006年第23期44-48,53,共6页Relay

摘　　要：提出了一种配网单相接地在线故障定位的新方法。故障发生后从变电站母线在线向系统注入高频脉冲信号并获取线路上与注入信号相对应的反射行波,在电网结构已知的情况下,通过对反射波的分析进行故障定位。该方法不仅能检测故障点的电气距离,而且能确定故障所在的分支线路,还能通过调节注入信号的大小多次定位故障,提高定位的准确性。应用该方法进行故障定位不受接地电阻的限制。通过M atlab/S imu-link进行了建模与仿真分析,仿真结果验证了该方法的正确性。The principle of traveling wave based on fault location of power transmission lines is presented. The power net is injected high frequency impulse signal online from voltage transformer connected to substation bus after the fault of single-phase grounding take place, and then the echo traveling wave from the fault spot corresponding to the high frequency impulse signal is captured by the fault detector installed in the substation. As long as the network structure is known, fault location can be found by analyzing character of wavelet transform modulus maxima of echo traveling. With this method, not only the electrical distance from fault spot to the detector can be checked but also fault offset circuitry is confirmed. Fault location veracity can also be increased via adjusting injected signal magnitude time after time according to analysis results error. This method can avoid grounding resistance influence for a precision detection. The modeling and simulation of the proposed method are conducted by software MATLAB/SIMULINK. The simulation results show that the proposed method is feasible.

关 键 词：电压互感器 小电流接地系统 神经网络 小波变换 相模变换 C型行波故障定位 模极大值 

分 类 号：TM711[电气工程—电力系统及自动化]