基于行波反射法的SF6断路器分合闸线圈匝间短路故障定位研究

Research on Fault Location of SF6 Circuit Breaker Opening/Closing Coil Based on Traveling-wave Reflection Method

提出将行波反射法用于断路器分合闸线圈的匝间短路故障定位,通过有限元仿真计算分合闸线圈的电容、电感矩阵,基于ATP-EMTP建立线圈的波阻抗模型。在绕组首端注入低压方波脉冲信号,获得发生匝间短路故障时的首端反射波特征曲线。采用覆盖法计算了反射波特征曲线的盒维数、关联维数与信息维数。结果表明:随短路位置变化,特征曲线盒维数未呈现明显的变化规律,关联维数与信息维数逐渐降低,且各层绕组匝间短路之间的关联维数具有良好的区分度。将短路故障位置划分为4段,以盒维数、信息维数与关联维数作为输入参数,使用BP神经网络进行短路故障的定位。仿真结果表明:测试集故障定位准确率为100%,该方法具有可行性。

Traveling-wave reflection method is used into locating the inter-turn short circuit fault of the SF6 circuit breaker opening/closing coil,the capacitance and inductance matrix of the opening/closing coil are calculated through finite element simulation,and the wave impedance model of the coil is established based on ATP-EMTP.Low voltage square wave pulse signal is injected into the start end of the winding,and the reflection wave is obtained at the start end.The box dimension,correlation dimension and information dimension of reflection wave characteristic curve are calculated by using covering method.The result shows that with the change of short circuit position,the dimension of the feature curve box does not show obvious change rule,the correlation dimension and information dimension gradually decrease,and the correlation dimension between the interturn short circuit of each layer has good differentiation.The location of short-circuit fault is divided into four sections,with the box dimension,information dimension and correlation dimension as input parameters,and BP neural network was used to locate short-circuit fault combined with fractal dimensions.The simulation results show that the accuracy of fault location in test set was 100%,indicating that the method is feasible.