Research On Remote Diagnosing And Virtual Maintaining System Of High Speed CRH Train

Modern high-speed CRH train has developed into an intelligent large complex system integrated with complex structures and high technology contents. Its reliability is directly high related to the safety and efficiency of the high speed and overloaded railway transportation. Some problems and limitation, such as low informationization level, deferred diagnosing and mantainence, difficult to find and locate the fault in complex system, and low sharing rate of the knowledge, exist in the condition monitoring and fault diagnosis of CRH trains. It is vital and meaningful to design and develop a remote diagnosis and virtual maintenance system of CRH trains.First of all, based on the analysis of the data transmission modes and interface specification in the process of dara transmission, data of the CRH train running condition involved in SDR data package and LDR data package were devided according to their characteristics. New data packages and data access layers were established and special tablespaces and spatial indexs were used to assure high efficiency in data storing and querying.Next, the limitations of the conventional Web knowledge management mode and the problems that exist in storing of the CRH trains fault case record were discussed. Ontology was used to description of the fault cases, and the advantages and disadvantages of different ontology description languages and reasoning mechanisms were compared in this paper. Structure Ontology (StrOnto) and Fault Ontology (FaultOnto) of CRH train fault diagnosis were established to standardize the description of the CRH train faults and the reusability and sharing of the knowledge were also improved a lot.Then, systems and structures of CRH train were divided into five levels based on Classification of CRH Train System, and the according encoding rules were established. Common failure modes, failure mechanisms and the typical structure of the CRH trains were analysed. Fault trees of different typical systems of CRH train were established based on deep analysis of the complex relationships between large amounts of faults. Fault tree qualitative analysis and quantitative analysis on different systems were also carried through to improve the diagnosis efficiency.Finally, the remote condition monitoring and virtual maintenance system was designed and developed for the CSR2train that produced by CSR Qingdao Sifang Co., Ltd. Functions such as remote real-time running condition monitoring and query and statistics of the historical faults with muti-condition way were realized efficiently in this system. The filling out, submiting and storing of the daily fault cases were also standardized with the use of this system. Dhree-dimensional models of the CRH train was generated and combined with fault tree analysis to realized virtual maintenance.