Production Rules Application In Electrical Equipment Fault Diagnosis

At present, all aspects of society and economy develop at significant rate. The massive requirement towards energy stimulates the fast development of electrical industry. While the system capacity and electrical power network is increasing significantly, the effect that is caused by electrical equipments unexpected faults may bring people much inconveniences in their usual lives and production. To keep system economics and stability, electrical industry mainly adopts these measures, which tries to promote the performance of electrical equipments, and the precautious experiments and planning maintenance schemes. However, this kind of maintenance scheme can't meet with the requirement of electrical industry's significant development. To adapt to the electrical industry's transmission to market economy, a new equipment maintenance strategy is proposed, e.g. electrical equipment protective maintenance.This paper introduces the development history and its present situation of electrical equipment maintenance scheme including post-incident maintenance, planning maintenance and protective maintenance, clarifies that status maintenance is the correct direction of this field. Moreover, this paper deeply discusses the implementation strategy, technical supporting measures and future development of electrical equipment maintenance. At last, the target research is decided to build an electrical equipment fault diagnosis expert system based on various status information.The technical support measures include reliability central maintenance (RCM) , status inspection and fault analyzing diagnosis technology, maintenance information management. The maintenance information management system is based on computerized information management to implement the overall collection and management of equipment relative static/dynamic status information. The status inspection and fault diagnosis technology is based on maintenance information system and artificial intelligence to solve the target problem. The latter is the core technology to implement the situational maintenance system. The paper introduces artificial intelligence and its study fields, furthermore, deeply discusses the expert system and its implementation mode, emphasizes how to select the proper structure of expert system and how to describe the relative knowledge at a rational way. Based on all above, the paper analyzes the basic engineering requirement about electrical equipments fault diagnosis and proposes the framework and its implementation ways of target system.Following the principle of software engineering, based on the real requirement in situational maintenance, this system uses equipments' online situation inspection data and offline basic data, applies the distributed database, creates the knowledge base containing fault set, symbol set and rules set, mainly applies the production rules in implementing the intelligent induction about electrical equipment faults. Finally, it works well in reality. In addition, the project also implements the equipment fault forecasting using production rules and the adapted exponential approximation algorithm.This paper mainly focuses upon the creation of knowledge base, production rule-based diagnosis mechanism. With the basic technology, such as database, artificial intelligence and software engineering, a target system comes into being in reality. It can solve the real problems intelligently and practically. Therefore, it is valuable in both theoretical fields and engineering fields.