Research On The Evaluation Method For Power Transformer Condition Based On Set Pair Analysis

In recent years, with large capacity, high voltage transformer bringing intoservice, how to ensure their safe operation becomes a vital problem of power system.The introducing of Condition-Based Maintenance (CBM) solved the problem,simultaneously has also created the huge economic efficiency and the socialefficiency. As the measurable transmission and transformation equipment of powersystem, the reliability of transformer is vital to the safety and stability of powersystem. According to the site data of monitoring measurement, it is of practicalimportance for CBM to find out the operation state of transformer timely andaccurately. With the unceasing development of the on-line inspection monitoring, thedomestic and international scholars have taken the initiative investigations andstudies of transformer condition evaluation, and also made quite a number ofimportant achievements. But the algorithm has a high complexity, and is incapableof dealing with indefinite information with aims. Based on the work of the formers,realizing the difficulty of scale of index and combining Fuzzy Analytic HierarchyProcess(FAHP),the paper presents a Evaluating Model of affiliation degree andset-pair state for the variety and uncertainty of state parameters.Representative status parameters were chosen to assess the condition of thetransformer by combining the new development of the on-line inspection monitoring,and a model of evaluation of affiliation degree is established. This paper hasprovided an across-the-aboard evaluation on the condition of the transformer fromthe aspects of electrical test, parameter of oil characteristic, state parameter ofdissolved gases in oil. In view of the needs of Evaluating Model of affiliation degreeand set-pair state, according to Guide for Condition Evaluation of Oil-immersedPower Transformers(Reactors),combining the existent achievement and the expertexperience, this paper has provide an evaluation classification of assessment indices.Adopting the FAHP to confirm the weight distribution of every index, makesconsistency checking convenient and straightforward. Based on fuzzy degree ofconnection, a new method was proposed to determine the discrepancy coefficient inSet Pair Analysis. Using this method, the discrepancy coefficient i should beanalyzed in discrepant, identical and contrary aspects.Then the model is deployed in the evaluation practice of transformer. Compared with fuzzy comprehensive evaluation, all factors can fuse into an organic whole. Thecase study of actual transformer indicates that the assessment method is simple andconvenient with accurate results which have practicability and reference value fortransformer maintenance.