The Study And Application Of Association Rules In Assessment Of Power Transformer Fault

As the hub of power system equipment,power transformer's safety relates to security and stability of power system equipment. In the transmission and distribution process, power transformer is energy conversion and transmission of the core , a source of national economic sectors and household energy,and it is the most important and most critical equipment. The safety of the operation of power equipment is first line of defense to avoid major accidents grid,which is the most critical defense systems for the power transformer equipment. Transformer serious accidents will not only damage their own cause,but also cause great economic losses to the social. Though many solutions have been applied to the diagnosis of transformers'failures, they still have many flaws. None of them can completely make the correct diagnosis in short time.The Operation of power transformer fault diagnosis is an important way to guide transformer maintenance. In this paper, fault diagnosis of transformer has prolems in the existing practical application ,so we introduce the association rules theory which from data mining,and introduce the gray-level evaluation method to assess the state of the transformer fault.By analyzing the causes of transformer failure,we find transformer fault is the result of many factors.,and there is complicated relationship between them .According to the multilevel nature of this information and the multiple factors , we establish an assessment of transformer fault diagnosis system,build transformer condition assessment matrix with least squares,and analysis it between the indicators of support and confidence of the calculation method to improve the use of grey, eliminate the difference between individual sample data and varied data, which leds to the phenomenon of the final results of the assessment is unreasonable, while avoiding the uncertainty of sample data.We improve the accuracy of fault assessment further. Example shows that a new resolution ratio and combination of data mining association rules theory of gray correlation assessment method is an effective method .Finally, the association rules of gray relational analysis is applied to transformer fault diagnosis and design a complete experimental procedures and programs, analyzing and comparing the results of experiments proved that this program transformer fault diagnosis in a higher correct rate.