Research On Comprehensive Diagnosis System About The Health State Of Running Transformer

Large power transformers play an important role in power system equipment,whether it runs properly directly related to the security and stability of the power system. Assessment of the health state of the power transformer is an important step in the Condition-Based Maintenance(CBM).Using the existing state information to assess to the health status of the transformer, has an important practical significance to improve the operation reliability of the transformer.On the basis of the summary and analysis of the exiting transformer condition assessment methods as well as detection and fault diagnosis technology,research and analysis the health state assessment of power transformers are involved in some state information. In this paper I take the assessment information into five parts which are attachment information,operating information, trial information, fault defects (maintenance) information and inspection information to establish the system of status evaluation of transformer. Using percentile ratings scoring method to determine the status information, and unified quantify the state information.This paper combines the actual situation of the current transformer, using gray theory and unascertained rational number theory to build a comprehensive assessment of the health status of the transformer model.The power transformer is seen as a typical gray system, using gray-clustering decision analysis methods to build the condition assessment model of the power transformer, the model of the power transformer divided the health state into five gray class,and established the whitening weight function of gray class.Because of the wide variety of performance parameters of the transformer, and the different impact of various parameters on the health status of the transformer,so take advantage of unascertained rational number to determine the index weight in the model.On this basis, introduce the information entropy theory, and improved indicators of variation is greater than the impact of the results of the assessment,and more accurately reflect the potential transformer exception status information,making the state assessment results more closer to the actual state of transformers. At the same time make the assessment model more practical.On the basis of model and the weights, research and analysis some state information, which are used to assess the health state assessment of power transformers, and show the gray-clustering decision steps of assessment health state of the transformer。Comprehensive diagnosis platform of health status for operating transformer was developed based on the model,using VB to design the front interface, meanwhile using SQL server2000as the back-end database.Verified by the later operational state and related tests,the platform developed in this thesis can make an effective and accurate evaluation on the status of running transformer,the assessment results can be used as reference for maintenance scheduling.