基于高频等效模型的分合闸线圈绝缘故障评估  

Insulation Failure Evaluation of Opening/Closing Coil Based on High Frequency Equivalent Model

作　　者：钱帅伟 彭彦军 王璐 周雄 滕本科 钟来胜 Qian Shuaiwei;Peng Yanjun;Wang Lu;Zhou Xiong;Teng Benke;Zhong Laisheng(Guilin Power Supply Bureau,Guangxi Power Grid Co.,Ltd.,Guilin,Guangxi 541002,China)

机构地区：[1]广西电网有限责任公司桂林供电局,广西桂林541002

出　　处：《机电工程技术》2023年第6期197-201,共5页Mechanical & Electrical Engineering Technology

基　　金：广西电网公司科技项目资助(0403002020030103BY00066)。

摘　　要：为有效诊断断路器分合闸线圈的绝缘故障,提出一种线圈匝间短路故障位置的评估方法。建立分合闸线圈的高频等效电路模型,在模型首端注入高频方波脉冲信号,在末端测得电压响应特征曲线。计算不同匝间短路位置下的特征曲线小波包—能量谱。结果表明:随着短路位置靠近线圈末端,特征曲线能量分布向低频集中。以特征曲线的小波包—能量谱为输入,短路位置为输出,基于概率神经网络建立诊断模型并进行仿真。结果表明:该模型故障定位正确率为85.7%,具有一定的可行性。In order to diagnose the insulation fault of circuit breaker,an evaluation method of the fault position of coil is proposed.The high frequency equivalent circuit model of the opening/closing coil is established,by injecting the high frequency square wave pulse signal at the first end of the model,the voltage response characteristic curve can be measured at the end.The characteristic curve wavelet packet-energy spectrum at different inter-turn short-circuit positions is calculated.The results show that the energy distribution of the characteristic curve concentrates towards the low frequency as the short circuit position is near the end of the coil.With the wavelet packet energy spectrum of characteristic curve as input and the short circuit position as output,a diagnostic model is established based on probabilistic neural network and simulated.The results show that the correct rate of fault location is 85.7%,and the method is feasible.

关 键 词：分合闸线圈 匝间短路 高频等效模型 故障定位 小波包-能量谱 概率神经网络 

分 类 号：TM561[电气工程—电器]