Design Of Power Transformer Condition Assessment System

Transformer in the grid is one of the most important and expensive equipment, if it fails, can make a huge assets loss, power loss and bad social influence. At present, our country has been more transformer operation more than 20 years, the operation of transformer is facing increasingly serious such as equipment failure and insulation aging, increasing the probability of an accident, therefore, to efficiently evaluate the state of the power transformer and the in-depth research of fault diagnosis, guide the transformer running state of maintenance and repair, to reduce the risk of failure, has important theoretical and practical significance.This paper guides in collect a large number of technical standards, procedures, expert experience and transformer actual running status data, on the basis of further study of the power transformer condition assessment index system, evaluation methods and decision-making criteria. Transformer comprehensive evaluation based on fuzzy reasoning model is put forward, the fuzzy membership function is constructed to describe the evaluation index factor model. Considering the determination of membership function with someone for factors, will lead to the result is not objective, this paper adopted by the ACO algorithm is combined with fault tolerance factor improved BP neural network to train optimization of membership functions, in order to make the results more credible.This paper has finished the preparation and verification of the assessment system software, for a small amount of sample data, the verification results correspond with actual state basic transformer, proved the feasibility of the evaluation method.