Development of a method for recognizing the type of defect based on the results of dissolved gas analysis using a set of diagnostic criteria

Abstract

Ensuring a reliable supply of electricity to both the households and industry is the main task of the power industry. One of the factors that affects the operational reliability of electric power equipment is the ageing of equipment, which is observed in every country. The most used among non-destructive methods of diagnosing the condition of high-voltage oil-filled equipment is the dissolved gas analysis. Using the values of gas ratios, gas percentages, as well as the values of gas ratios to the gas with the maximum content, most methods of interpreting the results of dissolved gas analysis can recognize electrical and thermal type defects, and some methods allow to recognize also combined type defects. However, the reliability of defect recognition by different methods for the same data is significantly different. The use of only one of the diagnostic criteria to recognize defects of different types does not always allow to make the correct diagnosis. Thus, the issue of developing a method for recognizing the type of defect based on the results of the dissolved gas analysis with the simultaneous use of a set of diagnostic criteria is relevant and has practical significance. Since the values of diagnostic criteria in the diagnostic space differ for different types of defects, it is advisable to use deterministic recognition methods, in particular distance recognition methods. To recognize the type of defect, it is proposed to use the method of minimum distance to the set, which allows not only to recognize the type of defect, but also to identify the same type of object with the closest gas content and with known causes of the defect. For the practical implementation of the method, an algorithm has been developed based on the joint use of the key gas method and the ratios of characteristic gases. The given example of practical use of the developed method showed its high efficiency in comparison with the existing methods of interpreting the results of the dissolved gas analysis.