The vacuum circuit breaker is applied to increasingly wide voltage class and its application domain is more and more widespread. The vacuum interrupter is a core component of vacuum circuit breaker; its reliability service is a reliable guarantee of power grid. During normal working hours, the vacuum level is required to maintain above a given value. Otherwise there will be an accident. Therefore it is very important to study the vacuum measurement and the remaining lifespan. But the previous research has been devoted to improve measurement methods of vacuum; little attention is paid to the remaining lifespan and the lifespan management of vacuum interrupter.In this paper, the expert system is used to realize vacuum measurement and lifespan management on vacuum interrupter.Firstly, a variety of measurement methods and influence factors of the vacuum level are studied. Inspiration, deflation, proliferation, seepage and air leakage in the store and work are analysed. The experience data and formula are collected and organized; the knowledge base of vacuum degree is established.Secondly, making use of learning capability of BP neural network, vacuum value is forecasted.The algorithms of BP are compared, the convergence of trainlm is faster than other algorithms, so trainlm is more suitable to forecast vacuum level.Thirdly,the change rule of vacuum level is discussed, the accepted leakage rate that meets the requirement of vacuum lifespan and the management curve of vacuum lifespan are given. Through this curve, not only change tendency of vacuum level may be seen, but also whether the vacuum lifespan is qualified can be judged.Finally, the simulation of human thoughts has been carried on to fuzzy reason to the determination of vacuum level; flowchart of expert system on vacuum measurement and lifespan management is given. According to flowchart, expert system is divided into data read, vacuum level computation, error and credibility judgement, remaining lifespan calculation and read in the database. Prototeyp of expert system about vacuum measurement and lifespan management is builded. And in credibility judgment, the credibility of measurement methods is synthetically judged by using the fuzzy theory and rough sets theory. The foundation is provided to establish the expert system and realize offline monitor. If expert system about vacuum measurement and lifespan management is applied, measurement in net and off-line can be achieved. Reliability of the grid will not be reduced, this method is superior to online monitor in the application and promotion, and the need of condition maintenance on switching device is met, the foundation is established to realize intelligent vacuum switch. |