Historical Similar Ticket Matching and Extraction used for Power Grid Maintenance Work Ticket Decision Making

Abstract

In order to play to the important reference value of historical maintenance work ticket in power grid for making new work ticket and realize multi-directional decision support, a method of intelligent matching and extraction of work ticket text is proposed. Firstly, the text of power grid security measures information is preprocessed with the power field knowledge. Aiming at the problem of non-standard text representation, an improved two-level bag of word (BOW) model with main word and auxiliary words is proposed. Then, Term frequency-inverse document frequency (TF-IDF) method is introduced and used to extract text features. Finally, cosine similarity method is used to calculate the multi-variable similarity between the critical information of maintenance equipment and the historical scenes. The problems of word order inversion and multi-word one meaning can be eliminated by the proposed method, and then the matching efficiency and precision can be improved. Hence, with the improved historical similar ticket matching and extraction method, the dispatchers and operators can get more comprehensive decision support when making new work tickets. The effectiveness of the proposed method is validated on several cases based on an actual power grid.