Fault Location of Lightning Strikes Using Residual Analysis Based on an Adaptive Kalman Filter

Abstract

Lightning strikes are the significant cause of transient, faults, and outages in electric power transmission and distribution systems. The double-exponential function is proposed to simulate the decaying part of the measured lightning current. The transient characteristic difference (such as the gradient and waveform) between the common fault current and the lightning fault current is described in this paper. Thus, the criterion, which is to compare the peak value of transient current with the amplitude of transient current at the following 200 μs, is proposed to distinguish the short-circuit fault and the lightning stroke fault. The traveling wave location method, which uses the estimated residuals by the Kalman filter based on maximum likelihood (KF-ML), has been proposed for lightning location and fault location in various conditions, such as different lightning points, different fault points, and different lightning sides. The estimated residuals exhibit the sharp singularity when traveling wave front arrives. Thus, the lightning side and the fault side can be differentiated by comparing the time when the initial traveling wave reaches the two substations. Also, the lightning distance can be calculated by the time when the initial lightning traveling wave surge reaches the substations M and N, and the fault distance can be obtained by the time of lightning initial traveling wave arriving at the fault side and the reflection wave from fault point which gets from the lightning side. The effectiveness of exacting mutation feature using the proposed method has been demonstrated by the simulated lightning current and has also been tested with the accuracy of lightning location and fault location under various conditions.