Research On Fault Line Selection Device For Small Current Grounding System Based On Multi Criteria Fusion

Due to the different voltage levels in China’s distribution network,the magnitude of fault current caused by grounding fault is not the same,so different neutral grounding modes should be adopted.Therefore,according to the national standard,a 3～66k V-voltage distribution network should adopt neutral non-effective grounding,while single-phase ground fault takes up the biggest part in the distribution network fault.As the rising necessity of China and its people,it is getting complicated to construct medium and lower voltage distribution.In the case of single-phase ground fault,it becomes harder to pick out faulty line because of various interference like environment.However,the traditional line selection device based on a single signal,such as transient or steady state components,cannot fit the requests of fault line selection in the new distribution network.Consequently,it is significant to study an integrated line selection algorithm to overcome the complex environment and complex signals.This paper introduces the concept of multi-criteria with multiple algorithms into fault line selection of distribution network.The details are as the following.1.Analyze the coefficients of each line when single-phase grounding fault comes about in distribution network,and infer the equation.And build a stimulation model,then analyze and verify the result to lay a theoretical basis for the method of selecting a single line.2.Study the line selection method based on each component,and analyze the principles of wavelet packet analysis,active power,fundamental wave amplitude phase comparison method and quintuple harmonics respectively.Finally,fault measure models are constructed respectively as per the characteristic quantities extracted from zero sequence current by different line selection methods,to provide data driving for the following integrated line selection algorithm.3.Study the integrated line selection algorithm intelligent algorithm as the central,expound the theory and context of BP(back propagation)neural network,and apply genetic algorithm to improve its weight and threshold.Lastly,employ the traditional line selection method to obtain various fault,based on which the model imitates the complicated condition in distribution network.After evaluating the GABP algorithm with BP algorithm and single line selection method,the virtues of the integrated line selection algorithm projected in this paper is demonstrated.4.The core part of fault line selection device is designed.The dual chip based on FPGA+ARM is adopted.ADC128S022+FPGA is used as the data processing module and ARM as the main control module to ensure the performance of the integrated line selection algorithm.The device is designed based on functional modularization,which is easy to be modified in the later stage.Finally,the hardware platform is tested in the 10 k V distribution laboratory of Yantai Dong fang Wisdom Electric Co.,Ltd.,and the feasibility of the device designed in this paper is proved by the wave recorded on the oscillograph.