Research On Early-warning And Forecasting System Algorithm For Distributor Fault Arc Based On Information Fusion

This paper comprehensively analyses the question in the current arc fault warning system. Based on the study statement of home and abroad, in this paper propose a arc fault precise warning system, using information fusion. This system involves before arcing sound early warning and after arcing arc fault waning. Arc sound warning system takes the pre-arc signal which detect from many FOM fiber sound signal to information fuse, as warning. Then get the accurate warning signal detection. Arc fault warning system bases warning signal on the arc light, voltage and current. The three type of information is on the basis of high sensitivity of the ultraviolet sensor SG01 series, WB series voltage sensor and SML series current sensor. Fuse the three type of information on the recognition scale, then, to get the accurate warning signal.This paper systematically studies function model and fusion structure model on different level the information fusion system. This paper respectively designs the parallel structure arc sound detection fusion model and arc light, voltage and current decision level character fusion structure model. Comparing the many fusion rule in the distributed detection fusion, this paper adopts the arc sound warning information fusion algorithm based on Bayesian minimum risk criteria. Contrasting with the D-S evidence theory and DSmT object identification fusion algorithm, comprehensively considering the fault arc character of the arc light, voltage and current, this paper introduce fault arc identification fusion algorithm based on DSmT.The paper tries to use MATLAB simulation to analyze arc sound early-warning fusion algorithm, which based on Bayesian minimum risk standard. And use practical examples to analyze fusion algorithm, which based on DSmT of arc fault recognition. The result shows that these algorithms have good performance in fault detection. The result also indicates that the design of the early-warning system in this paper, enhance the efficiency and accuracy of the arc fault early warning system comprehensively. At the same time it reduces false-alarm and fail rate of the system, and lower the conflict between sensitivity and false positives, so as to guarantee the stability and reliability of the system.