Research on power transformer maintenance strategy based on case-based reasoning

Abstract

Abstract In the past, the development of transformer maintenance strategies mostly used information data from a single source, and the final strategy was too general, unable to intelligently and automatically achieve effective guidance for equipment operation and maintenance. Based on existing case sets, this article studies case representation, feature attributes and weights, and a maintenance decision algorithm based on case-based reasoning. Firstly, the structural similarity between the target problem and the source case is analyzed, then the similarity between the two attributes is analyzed, and finally, the global similarity is calculated to modify and reuse the maintenance strategy of historical cases and retain new cases. The results show that the fault characteristics of the power transformer can be accurately grasped, and a reasonable maintenance strategy can be formulated to reduce the maintenance cost.