A Study On Fault Diagnosis Based On Fault-tolerant Neural Network

Fault diagnosis is a new multi-subject-crossed technique and it meets the needs of project reality .It has been developed rapidly during the past 40 years, and it has brought huge benefit.Information fusion is a subject formed recently. It has been researched and applied in many fields. There is much available information in fault diagnosis. Only when the available information is used, can the precision and credibility be improved. So fault diagnosis is a process of information fusion.The fault diagnosis model is put forward in this paper. Based on the analysis of single neural network characteristic, the model of integrated neural network is put forward. At the same time, the way to establish model, the principle to compose and the strategy to realize are given in this paper. The result based on the simulation of the example show it is a effective way by using integrated neural network to deal with the characteristic information.How to reduce the misdiagnosing rate and improve dependability has great meanings in the process of the fault diagnosis. Traditional diagnosis system don't consider network output error under the fault situation,so the systematic fault-tolerance is limited greatly. So the dependability of the neural network itself seems more and more important, and it needs to analyse and design the theory and method of the neural network of high dependability urgently.In this paper we introduce the frame of fault-tolerant neural network and do simulation on fault-tolerant neural network. The result based on the simulation of the example show fault-tolerant neural network can keep performance well when the network runs well. When the network breaks down , the system have good performance too.The system improves the fault-tolerance of fault diagnosis greatly.