Research On Full Dimension Intelligent Decision Support System Of Transformer

In recent years, with the development of national economy and improvement of living standards, the scale of power grid has been expanded rapidly, which has put forward a higher requirement for the quality and reliability of power supply. And Transformer is an important equipment in power gird, and its safe and stable operation has a direct impact on the power grid. Therefore, the decision analysis research of transformer has a decisive significance on the improvement of the power supply quality and power gird reliability.This dissertation carries on some researches, including transformer state assessment, risk assessment, maintenance decision-making, fault diagnosis and insulation analysis, on the basis of study of present research situation of transformer decision analysis. Considering all the results, we put forward a full dimension intelligent decision support system of transformer, which contains three aspects:full dimension of data, full dimension of device and full dimension of decision analysis. In order to fully utilize the system, we construct a decision support information system based on the full dimension system by applying information technology. Since the implement of the decision support information system, it has brought huge economic and social benefits, and also significantly improved the efficiency of operation and maintenance management, which proves the effectiveness of the full dimension intelligent decision support system. The main conclusions of this dissertation are as followings:(1) The state assessment of transformer is the basis of decision analysis. For the complex structure of transformer, there are many factors relating to the operation state, and it's hard to evaluate the operation state of transformer only by one single factor. Therefore, we propose a novel state assessment model of transformer based on the fuzzy comprehensive evaluation and the improved DS evidence theory. It can effectively eliminate the uncertainty among multi-source information, and provide scientific support for the maintenance decision-making of transformer.(2) Through the in-depth study of the failure mode, its harmfulness analysis and maintenance decision-making influencing factors, this dissertation puts forward a risk assessment model based on FMECA and a maintenance decision-making knowledge base.(3) The timely diagnosis of incipient faults of transformer has an extremely important significance on the stable operation of power grid. And the dissolved gas analysis is considered to be best means to the diagnosis of oil-immersed transformer. In order to overcome the defects of the traditional dissolved gas analysis and improve the diagnosis accuracy of existing fault diagnosis models, this dissertation proposes a fault diagnosis model based on improved BP neural network by applying chemical reaction optimization and fuzzy c-means method. Through the actual case, it concludes this model can achieve a higher diagnosis accuracy.(4) Transformer insulation life is one of the important indicators of transformer operational reliability, and closely related to the aging of insulation materials. Therefore, this dissertation, considering insulation reliability indexes and analytic hierarchy process indexes, proposes an insulation assessment model based on entropy weight fusion method, which improves the accuracy and rationality of transformer insulation assessment. On the basis of the insulation assessment model, this dissertation puts forward a novel transformer lifetime prediction model based on improved grey prediction algorithm, and finally forms the insulation analysis system of transformer based on health index. Through the actual case, it concludes the system can predict the insulation aging trend of transformer more accurately.(5) Based on the research of transformer state assessment, risk assessment, maintenance decision-making, fault diagnosis and insulation analysis, this dissertation proposes a full dimension intelligent decision support system of transformer, including full dimension of data, full dimension of device and full dimension of decision analysis. In this system, we integrate the asset data, real-time monitoring data and inspection data of transformer, and use decision analysis full dimension model and the results of comparison among similar equipment to provide a theoretical analysis for the maintenance decision-making. This system can improve the accuracy and rationality of decision analysis and the level of transformer operation and maintenance work. And in order to fully utilize the system, we construct a decision support information system based on the full dimension system by applying information technology.