An Online Diagnosis System For Detecting Small-current Grounding Fault

With the development of computer and communication technology,to better meet the people’s growing demand for production and living,the State Grid Corp.is gradually promoting the construction of intelligent power grids.Specifically,a key goal of the intelligent power grids is to significantly improve its reliability.To improve power supply reliability,we firstly need to detect and deal with faults timely.Furthermore,80%of the faults are small-current grounding faults.This kind of fault is one of the most important types in power grids.Therefore,how to detect and deal with the small-current grounding fault as soon as possible has become one of the most important problems to solve in the construction of intelligent power grids.At present,most electric power companies still use low efficiency manual methods to analyze and diagnose small-current grounding faults.Therefore,the academic community has launched researches on automated diagnosis of small-current grounding fault.This kind of automation methods performs the fault diagnosis based on the line-specific information from the acquisition device.However,the typical automated methods usually consider only one kind of data for its diagnosis.As a result,these methods are still difficult to correctly locate the small-current grounding faults and thus cannot be applied in practice at this moment.To address these issues,this thesis designs and implements an online system for diagnosing and analyzing small-current grounding faults.The idea of this work is to combine the diagnosis results from different automated devices for diagnosing automated small-current grounding faults with the data of lines and bus bars for manual diagnosis to achieve a comprehensive judgment.By consulting the DS evidence theory,we proposed a criterion-confidence method based on the combined pluralistic data to improve the efficiency and the success rate of small-current grounding fault diagnosis.Specifically,the work of this thesis is summarized as follows:·Based on different data transmission protocols,we obtain the diagnosis results from various fault diagnosis devices and the operational data of lines and bus bars with necessary pre-processing.·We turn data into positive and negative criteria according to whether the data meets the characteristics of small-current grounding fault.By fusing the criterions,we got the probability of small-current grounding faults occurred in the line segments.By setting the threshold to that probability,we can conclude whether a small-current grounding fault is occurred in a particular line segment.·We develop a Web-side visualization system by using the Spring MVC framework.The system is able to display the line status,the fault-related information,and the fault diagnosis and positioning results on the webpage in real time.Jiangsu Electric Power Research Institute conducted grounding experiments to evaluate our proposed diagnostic system in the distribution network of the Guangling District in Yangzhou.In the experiments,the system successfully detected the smallcurrent grounding fault that occurred in the field.This system also showed the fault occurrence segment and the cause of the fault on the webpage.The experimental result shows our system has high diagnostic efficiency and meets the requirements of practical application.