Research On Fault Diagnosis And Technological Service Support System Of TBM

TBM as the main force of subway construction is playing an increasingly important role in urbanization construction today. Because the TBM is a large and complex equipment which include machines, electrical and hydraulic, it is considerable difficulties to find and remove breakdowns merely depending on the experience of experts.Trouble shooting in good time and minimizing downtime caused by maintenance will bring huge economic benefits. In this reason, this thesis makes study of the fault diagnosis and technological service support about TBM and develops a system.Knowledge is the foundation to develop expert system and the system’s practicality and stability is affected by completeness and reliability of knowledge. Knowledge acquisition is one of the difficulties in the process of expert system development. A rough set data mining technology for feature weight calculating and rule acquisition is selected to over come the shortages of some knowledge attaining methods existed before. The key technologies include data discretization based on SOM and attribute reduction based on genetic algorithm and heuristic value reduction gorithm. Feature weight becomes more reasonable and rule becomes more condensed by improving algorithm.The Intelligent Fault Diagnosis model and Technological Service Support model of TBM are the core contents of this paper. Intelligent Fault Diagnosis model is presented based on CBR and RBR in this paper. The computing method of euclidean distance and the reasoning mechanism of multi-step coregistration are used in CBR. RBR expresses knowledge producing rules, uses reasoning machanism of positive direction. Finally CBR and RBR are integrated sequentially. Technological Service Support model applies a fuzzy analytic hierarchy process to schedule tasks and establishes personnel scheduling methods with regard to single-objective tasks, multi-objective tasks, rush-repair tasks and parallel tasks.This paper programmed a Intelligent Fault Diagnosis System of TBM based on CBR and RBR and a Technological Support System based on the theory of sequencing tasks and assigning personnel, with the technology of obj ect-oriented design and various kinds of data accessing technologies in Visual Basic and MATLAB mixed programming environment. The system is able to mine and data diagnose interactively and offer Technological Support and maintain system. At last the reliability of the system is verified through experiments. Results show that this system runs faster and has more pertinence and diagnosis results conform with situations.