Fault diagnosis expert system for diesel locomotive based on fault diagnosis binary tree

作者：Wang Zihao^*; Mo Yimin

来源：China Railway Science, 2008, 29(5): 98-101.

• Before and after
• Fault characteristics
• Fault diagnosis
• Fault diagnosis expert systems
• Fault probabilities
• Heuristic search
• Heuristic searches
• Knowledge base
• Matching ratio
• Rule base
• Valuation functions

Based on the fault characteristics of diesel locomotive, the binary tree of the fault diagnosis process was studied. Then, the knowledge base was divided into three independent and interrelated databases: the fault phenomena base, the fault cause base and the fault diagnosis rule base. Under the control of knowledge base management module, the before and after components of a fault rule, which was formed by the knowledge extracted respectively from the fault phenomena base and the fault cause base, were constituted and saved in the form of "IF&mellip;THEN&mellip;" in the fault diagnosis rule base. Lastly, a fault matching ratio based mechanism of heuristic search, which combined with experience of experts and probability of faults, was adopted. By modifying the valuation function, the priority research of greater fault probability can be realized, and the efficiency of fault diagnosis can be enhanced. At present, the fault diagnosis expert system for diesel locomotive, which based on fault diagnosis binary tree, has been operated successfully in each locomotive depot of Nanchang Railway Bureau and achieved the anticipated goal.

• 出版日期2008