Method Based on Rough Set and Improved Petri Net for Transformer Faults Diagnosis

Abstract

Oil chromatographic analysis is widely used in transformer fault diagnosis, but it is difficult to establish accurate maping relationship between the parameter space and the state space, and there is information complexity. This paper adopts the combined diagnostic model of rough sets and petri nets, firstlysimplifies the complex system which contains complicated discrete information by rough set to solve the state space limitations of Petri network; and improves petri network based on mining association rules, adopts the correlation matrix and state equation method to improve the reasoning speed, at the same time turns the diagnosis into matrix operations to change complex calculations to simple math which has certain applicability. Finally the algorithm is applied to gas chromatographic analysis in transformer oil, the calculation results is the same with IEC three ratio method, which proves that this method can quickly and accurately to judge the running state of transformer, so as to improve the safety, stability and economic operation of the flat water transformer.