Research Of Detection Method On Small Current Fault Section Signal Based On Wide Area Measurement

In our country, due to neutral point non-effectively grounding operation mode with high supply reliability, it has been widely used in distribution network. In resonant grounding system, arc suppression coil can countervail the fault current, but makes the fault current more weak, making it harder to detect the fault signal. Obtaining the accurate line fault data timely is one of important functions of the distribution network on-line monitoring system. Which can not only provide data support for all kinds of fault location algorithms, but also can contribute to extracting fault feature and distinguish different fault category. Thus, it is of big importance to study the topic of single-phase ground fault signal detection in resonant grounding system.Single-phase grounding fault sector signal detection method in resonant grounding system of distribution network based on wide area measurement technology have been researched in this paper. Firstly, single-phase grounding fault in resonant grounding system has been analyzed, the zero-sequence phasor theory and its calculation method have been introduced, resonant grounding fault characteristics has been analyzed, fault signal distribution in upstream region and downstream region of the fault point has been described, single-phase grounding fault location criterion in resonant grounding system has been worked out. Then, it has analyzed the influence of the imbalance zero-sequence current and signal acquisition device to the signal detection, and proposed the solutions to each issue. Whereafter, it has introduced a zero-sequence signal acquisition device, the core component of which is a microprocessor based upon ARM Cortex-M3, accomplishing synchronization signal detection via GPS timing, obtaining zero-sequence voltage and zero sequence current data by ADC sampling and fast fourier transform. Then, time synchronization algorithm in wireless sensor network has been studied. And an algorithm DMTS based on Weibull Distribution model has been proposed, introducing time delay parameter to ameliorate the original model, using Weibull Distribution to model the time parameters, optimizing the original algorithm. Finally, for the sake of signal de-noising, a singular value decomposition signal de-noising algorithm based on genetic algorithm has been proposed in this paper. This algorithm constructs a Hankel matrix with the original signal, and utilizes GA to optimize the matrix structure. Then it conducts singular value decomposition transformation on the matrix. Then, it works out the number of useful singular values by K-medoids clustering algorithm, and reconstructs the signal with the method of conducting inverse singular value decomposition transformation on the values and their corresponding vectors. Simulation results show that this algorithm has obvious effects.