Approach With Information Fusion Of Multiple Data Resources To Fault Diagnosis Of Power Grid

As the industrialization developed, the need of electric power is increasing day by day. Meanwhile, energy is increasingly brought to the attention of the country and the people. Nowadays, the size of the grid is expanding, which means that the power grid in providing people with more clean energy but the probability of failure is becoming more and more high. When a fault happens in the grid, so many monitoring devices will upload vast amounts of alarm information to the control center and it resulting dispatchers often too late to deal with and judge. For this reason, study a system which can help dispatchers determine the fault component in grid is very necessary.Based on the existing literature, some fault diagnosis methods was studied in this paper. They are temporal cause-effect net of variable structure, fuzzy cause-effect net and multiple resources information fusion. Temporal cause-effect net of variable structure is a diagnosis method which based on the traditional theory of cause-effect net, combined with the idea of failure mode classification and using the alarm information by the temporal constraint network processing. The proposed method is simple, clear and has a faster speed of diagnosis. It has a good online application prospect.Fuzzy cause-effect net is a diagnosis method when protection and circuit breaker refusing action and cannot use temporal constraint network method in the pretreatment of alarm information. Besides, the causal intensity evaluation method of fuzzy cause-effect net between associated node was also studied in this paper. The method of fuzzy cause-effect net is given in the form of fault-confidence of each suspicious faulty component, more in line with people’s habits of judgment. In response to complex failure scenarios, a kind of fault diagnosis method based on multiple resources information fusion was studied in this paper. The method analyzes the current record data of suspicious faulty component to extract the characteristic. Then improved D-S evidence theory is used to fusion the characteristic of current record data of suspicious faulty component and characteristic of digital information, determine the faulty component through the result of data fusion.This method is effective to avoid the insufficiency of only use digital information, it through the fusion of multi-source information, provides a more accurate judgment to suspicious faulty component.Finally, several conclusions are obtained based on the research outcomes, and directions for future research indicated.