Research on Transformer Condition Evaluation Based on Improved Fuzzy Comprehensive Evaluation

To realize the evaluation for the running condition of transformers accurately is not only the basement of condition-based maintenance, but also the need for distribution network intelligent reform. This paper focuses on transformer state estimation in distribution system. Various state qualities which affect transformer operation status, the access to acquire state qualities and maintenance needs in actual operation were well analyzed. On the basis of analysis, the transformer was divided into several component units and by the adoption of the theory of fuzzy mathematics, a feasible multi-level fuzzy comprehensive evaluation model for transformer state estimation was established. Based on the analysis of traditional AHP method, in this paper, the degree of impact on operation state of distribution transformer by different state qualities was considered and the weights of state factors in estimation model were determined through the introduction of triangular fuzzy numbers, which further reduces human subjective impacts. At the same time, taking significant information lost in the judgment of distribution transformer final status by the principal of maximum degree of membership into account, the data was processed and corrected with gray related degree theory to improve estimation accuracy. Through an example analysis and comparison with the traditional method, availability was showed by the new method, it can provide the basis for CBM of transformers.