Case-based Reasoning Fault Diagnosis Method Of Train Control RBC System

Chinese high-speed rail develops rapidly in recent years. So far, China has the most comprehensive high-speed rail system technology, the strongest integration capability, which leads a new trend of the development of high-speed rail in the world. RBC, which is the core ground equipment of CTCS-3 level train control system, ensures the stable train operations because of its high reliability and safety. Therefore, it’s significant for researches about RBC fault diagnosis.It’s difficult to maintain the equipment due to the concealment, diversity, and overlapping of RBC system. It is far from meeting the needs of fault diagnosis efficiently by a modern high-speed rail system which only relies on personnel experience of maintenance staff. In this thesis, Case-based Reasoning technology is used to support the fault diagnosis of RBC which is based on the analysis of the present maintenance of RBC and experience accumulated during the RBC fault process, which makes a technician diagnoses quickly and accurately, and ensures the safe and efficient operation of the train. The main findings of this thesis are as follows:First, the principle of Case-based Reasoning is introduced in the thesis, using a case representation method which based on fusion object-oriented technology and database technology to express the RBC fault, which formed a RBC case library. The next part analyses the structure and functions of the RBC system, organizes the RBC case library with the hierarchical organization method, and constructs the reasonable organization structure of RBC cases.Secondly, in order to avoid the problem that all the feature attribute of the case would have the same impact on case similarity, it is decided to use the Rough Set theory to reduce the attributes after comparing the common weight calculation method. Moreover, the importance level of attribute is used to extract RBC feature attribute weights, and subjective weights is also taken into consideration, then the final weights of the attributes are obtained.Thirdly, the fusion of weighted K-NN algorithm and cosine function algorithm of similarity calculation method is used to calculate the similarity between the cases after the study of common similarity calculation methods, meanwhile a practical example is used to verify the method; Then the RBC fault phased nearest neighbors method retrieval model is constructed according to RBC case organization structure.At last, C# language is used to develop RBC fault diagnosis system with the aid of the Microsoft Visual Studio 2013 platform. Meanwhile, the design of function module and the flow of diagnosis of the whole system are introduced in detail, where the database is designed by Microsoft Access 2010. Then, a specific example is expressed to demonstrate the effectiveness of the fault diagnosis system.It is proved that the Case-based Reasoning technology can help solving the problem in the RBC fault diagnosis and improving the efficiency of the field maintenance personnel’s work through the research of this thesis.