Study On Fault Diagnosis Technology And System Of The Traction System Of Metro Vehicles

For a long time, urban mass transit operational safety issues have been given close attention by all levels of government and public. And whether metro vehicles safety or not is closely related to passengers' safety. Furthermore, in recent years, the complexity of metro vehicles equipment is increasing, and the failure rate is rising. Therefore, how to diagnose the traction system of metro vehicles efficiently, rapidly and accurately is an important issue to be resolved.In this paper, based on the rigorous summarization to resent rail transit fault diagnosis technology in the world, we have done an in-depth research on operational vehicles in Shanghai Metro, conducted and analyzed the application and fault occurrence of the key equipment, and identified the rules and characteristics of fault knowledge. Based on the above needs analysis and research, this paper has designed a fault diagnosis integrated system of the traction system of metro vehicles which sets carborne, metro depot and control center in one. And we have established the functional structure of the sub-systems, and then the design of expert system is demonstrated in detail in this paper.Knowledge acquisition is the basis of fault diagnosis. This paper first analyzes the source of diagnosis knowledge and access methods, and its access methods have two ways: first, acquiring from knowledge engineers and equipment data; and the other from the FMEA form. Then we look production rules as an expert system shallow knowledge, and denote the failure of the traction system using production rules.Causality diagram can present causal relationship between failure intuitively, satisfies with probability theory rigorousness, and hasn't restriction for the topology of graphs. Therefore, based on the causality diagram theory, this paper has established causality diagram model about traction system, and look it as deep knowledge in expert system for diagnosis. In view of the diverse existing reasoning algorithms for causality diagram, this paper use a conventional causality diagram reasoning algorithm for authentication, then we find that this algorithm is very complicated. Therefore, this paper finally computes the causality diagram using an approximate reasoning algorithm. This algorithm is presented in computer program.Finally, this paper has developed the fault diagnosis system of the traction system of metro vehicles. First, functional modules and system databases are described, and then system operational interface is presented. Furthermore, this paper has combined with shallow knowledge and deep knowledge for diagnosis, which not only applies to current fault diagnosis, but also applies to reasoning the occurrence possibility of potential failures. The results demonstrated a preferable validity.