Development of Expert System for Dissolved Gas Analysis

HANDLE Web Site Open Access

Bibliographic Information

Other Title
  • 油中ガス分析エキスパートシステムの開発
  • ユチュウ　ガス　ブンセキ　エキスパート　システム　ノ　カイハツ

Search this article

Abstract

Dissolved gas analysis has been extensively as a method of maintenance of oil-filled electrical equipment. If an abnormality occurs in the internal oil-filled electrical equipment such as transformers, diagnostic experts diagnose to abnormal with gas pattern method. However, this method requires the expert knowledge, experience and labor of experts. Therefore, we have developed an expert system to automate the diagnosis of experts had been. Expert system has two types of knowledge base. One is an empirical model. This model is modeled based on expert knowledge and experience of diagnosis, make the imitation of conventional diagnostic techniques. Another is the prediction model. This model is modeled based on internal abnormal diagnosis by Fuzzy Decision Tree. Internal abnormal diagnosis by Fuzzy Decision Tree is possible to identify anomalies with an accuracy of 80% or more. By using an expert system, abnormal points and the probability can be predicted from the results of dissolved gas analysis, an objective and quantitative diagnosis when conventional methods was difficult become possible.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top