The Research Of Transformer Running Condition Assessment Based On Fault-tree Analysis

In recent years,with the rapidly development of power grid,the security of power system becomes more and more important,and the transformer is the core in the whole power system,the security of its running is also the important object for people attention,so,it is a great significance for grasp the operation condition of power transformer in real time.The general method for evaluate the power transformer operation is preventive test method,namely that doing a series of preventive test for transformer,then obtain the preventive test index and data which is relate the transformer,and using these indicators to build a reasonable evaluation model to judge the operation condition of transformer.However,this method is just can get the overall operation condition of transformer.In real life,we are more need to grasp the running status of transformer internal components so that the maintenance personnel can more quickly find out the weak fault components,therefore,this paper proposed an evaluation model for power transformer running status which is based on fault tree.Firstly,the transformer is divided into seven fault parts which according to the knowledge of fault tree,and find out the bottom event of each fault part,then find out the preventive test index which is relate each bottom event,and set up the transformer test index model is based on the fault tree.Secondly,it is necessary to calculate the weight of each index and each event.For quantitative indicators,this paper combine with improved analytic hierarchy process(IAHP)method and entropy weight method to determine the index weight;for qualitative indicators,using IAHP method to determine the index weight;for the determination of the weight of the fault tree bottom event,this paper use IAHP method and probability importance methods to determined.Finally,to determine the membership function of each index,this paper through the ridge type membership function to establish the membership function of each indicator.And,as the failure probability of transformer will along with the exponential growth when the operation time of transformer is more than 15 years,so this paper presents that the determination of transformer’s index membership function which is more than 15 years and no more than 15 years are different,and giving its membership function through relevant literature.For the qualitative index,accordingto the group decision theory to calculate the expert subjective weight,and combined with the triangle-trapezoid membership function to conclude the membership degree of each index.Then according to the weighted average model,we can get the operation condition of transformer’s child widgets and transformer.Through the results of the city’s transformer performance evaluation analysis,this method is reasonable and effective,and it also has a certain guiding significance for the state overhaul of transformer.