Diagnosing Potentially Abnormal Attribute of Power Transformers Method

Abstract

Design Method: In this paper, the program integrates those diagnostic methods from the Big data, the ANSI/IEEE C57.104 specification, and the Key gas to develop a kind of diagnostic method in power transformer. Their main diagnostic element is the insulating oil dissolved gas, which is integrated through the "correlation coefficient" and "classical detection and estimation theory". To develop a novel method diagnoses the latent abnormality attribute of the power transformer. The new method is accomplished through clever loop control, which has been validated by repeated and extensive data testing. This method is superior to the existing traditional diagnosis method in type of latent abnormality of diagnose, and has an accuracy of 90% under the verification those actual cases obtained from Research Institute of Taiwan Electric Power Company. This method is based on the MATLAB application software to execute all steps. In addition to being easy to operate and displaying the results of test in text, it not only improves the accuracy of the diagnosis of the latent abnormal attribute inside the power transformer, but also prevents the accident of blackout equipment and affects the power supply. It can to say, it is a powerful tool for electrical equipment maintenance technicians to detect. Design Purpose: Based on this diagnostic method to improve the accuracy of the diagnosis of the latent abnormal attribute inside the power transformer by MATLAB application software. Design Effectiveness: The program has been tested from some actual cases to improve the accuracy of diagnosis is been risen 90% up.