Research On Grounding Grids Breaking Points Diagnostic Method Based On Information Fusion

Grounding grids play a vital role to ensure the stable operation of the electrical power system and the safety of operators in power substations and plants. Corrosion phenomenon of grounding grids buried in soil exists generally. Serious corrosion of grounding grids will induce the grounding body rupture, inadequate hot stabilities, grounding parameters worsen, and so on. Accodingly the safety and stability of the power system will be in danger directly. Therefore, it is quite urgent and important to develop a method to diagnose breakpoints in grounding electrode without excavation or power off in time, in order to take corresponding defensive measures and prevent the hidden troubles happening.By analyzing weaknesses of existing two diagnosis approaches for breakpoints in grounding grids, which are based on circuit network method and electromagnetic field method respectively, some improvement measures are proposed. Also, this paper studies the application of Monte Carlo method in the evaluation of circuit network diagnosis algorithm and the testability analysis of breaking points diagnosis for grounding grids, which are usful for reasonable utilization of the diagnosis algorithm and diagnosis results in the subsequent chapters. Then, simulations for breakpoint diagnosis in grounding electrodes are carried out by using the two apprcaches respectively. The results show the disagreement in some special diagnosis of breaking points in grounding grids.In order to improve the diagnosis efficiency and accuracy, information fusion technology is introdued, for the first time, into breakpoints diagnosis of grounding grid. The model of fault diagnosis based on information fusion is built. Based on it, D-S evidence theory is applied to breakpoint diagnosis in grounding electrodes. Taking the results from the two above-mentioned approaches as sources of the evidences and using membership function to realize probability assignment of evidence bodies, fusion diagnosis is performed. Simulation results show that the proposed method is more effective than the previous.