Method of intelligent thermographic diagnostics of power transformers

Abstract

The utility model relates to the field of electrical engineering and power engineering, namely to methods of intelligent technical diagnostics and monitoring of the condition of oil-filled electrical equipment, in particular to methods for assessing the condition of power transformers. The objective of the proposed invention is to increase the reliability of assessing the technical condition and the accuracy of predicting the residual life of power transformers by taking into account the statistical structure of thermal fields and automating the decision-making process. The stated problem is solved by introducing new features into the diagnostic method based on the formation of a thermal profile of equipment using infrared thermography data and the selection of a region of interest (ROI): a thermal information function (TIF), which characterizes the continuous density of the distribution of temperature probabilities, and a structural similarity coefficient ( ), which determines the degree of deformation of the thermal profile relative to the reference state. The final diagnostic decision (Health Index) is formed using a fuzzy logic system, the rules of which are adapted to regulatory requirements and current industry regulatory requirements for the diagnostics of oil-filled electrical equipment. The proposed method enables the early detection of hidden defects and "cold" anomalies (drop in oil level, clogged cooling system) while reducing human error in the interpretation of results and reducing diagnostic data processing time by 65-70%. According to expert assessments, this method provides increased defect recognition reliability compared to traditional point temperature analysis methods. Thus, the set of features newly introduced in the proposed method makes it possible to solve the problem to which the invention is directed, while ensuring the achievement of the required technical result.