| Power Industry is one of basic industries. Study the diagnosis of power transformer incipient fault is most important measures to ensure safe operation of power equipments and advance development of the national economy.Based on " Research on CMS2000 Integration Transformer Substation State Monitor System " (Electron Information Industry Development Fund Project of Ministry of Information Industry). This study aimed to the transformer incipient fault diagnosis, the defect of the current diagnosis systems was discussed; a fault diagnosis artificial intelligence module and corresponding fault diagnosis system was accomplished with fuzzy classification theory. At last, the fault diagnosis system will be a part of CMS2000 Integration Transformer Substation State Monitor System.At the present time, three ratios method has come into widespread use in the domain of diagnosis of power transformer incipient fault. This method is at variance with the known facts of transformer faults in some times because of the code absence and absolutes' ratios. To the question of defect of currently fault diagnostic techniques, this dissertation introduced the factor of created gases into the field of transformer incipient fault diagnosis with vast amounts of online DGA data and fuzzy theory and performed modules of fuzzy measurement, fuzzy distinguish. At last, a fault diagnosis system based on these researches was accomplished. Good practice dictates that the methods of based on fuzzy measurement and fuzzy distinguish modules is more efficiency than three ratios method and the method based on fuzzy multi-layers classification module with enough sample data is better than all of them. |
