Power equipment defect intelligent diagnosis platform based on knowledge map

Abstract

The defect data of electric power equipment generated during the production and operation of the electric power field is widely used for judging the defect degree of the equipment, assisting the inspection personnel to store production certificates, and facilitating the on-site maintenance of maintenance personnel. During the operation of electric power equipment, it is necessary to timely and effectively judge the defect degree of the equipment as to avoid a series of cascading failures caused by the untimely treatment of critical defective equipment. It affects the power production efficiency. This paper proposes an effective method for entity disambiguation. It finds that the knowledge map of power equipment defects lacks updating measures, and gives specific updating methods for different updating reasons. The model uses the attention mechanism to extract the word importance features in the defective text, uses the enhanced coding method to recode the digital information, and the fusion layer fuses the global semantic features, digital features, and word importance features; Finally, combined with knowledge atlas, the text description of power defects with given structure is realized. The model designed by comparison can more accurately identify the entity information in the defect text. Meanwhile, it can also achieve an end-to-end analysis effect of the defect degree of the power equipment. The research part based on historical defect text uses the improved knowledge mapping technology of power equipment defect text, which explores the practicality of this technology.