Neural Network And Fuzzy Reasoning In Breakdown Diagnosis Applied Research

Along with the technical rapid development as well as the large-scale complex system in industrial production widespread application, causes the conventional fault diagnosis technology to satisfy the people more and more with difficulty the reliable request which proposed to the large-scale complex system, but more and more receives the favor based on the fuzzy reasoning and the artificial neural networks intelligent diagnosis technology, is the large-scale complex system failure diagnosis development key direction.This article introduced the fuzzy logic and the artificial neural networks elementary theory, then unifies both carries on the superiority supplementarily, has constructed the fault diagnosis method which one kind of artificial neural networks and the fuzzy reasoning technology unify. The examination variable partial data send in the neural network to carry on processing first, its obtained preliminary diagnosis result recombination other correlation observed value after fuzzy reasoning system processing, obtains the finally diagnosis result, finally by the principle explained the system gives the breakdown diagnosis basis. This method obviously reduced the fuzzy reasoning section construction complexity, strengthened the artificial neural networks solubility. Passes the superheat nitric acid cooling system the example to prove that, unifies the fuzzy reasoning technology and the neural network technology applies in the breakdown diagnosis system is feasible, also is effective.