Research On Construction Method And Evaluation For Fault Knowledge Trees In Automobile Stamping Workshop

As research continues, knowledge management has been applied widely. In the automobile manufacturing enterprises, knowledge management has became mature abroad, while still in the exploratory stage domestically. The automobile manufacture enterprises develop rapidly in recent years, which has accumulated a lot of knowledge. How to carry on the effective management of knowledge to realize the value of knowledge and improve the competitiveness of the enterprises, is the problem of domestic auto industry which needs to be solved the urgently.Knowledge management in automobile fault enterprises is a new research subject. The data and knowledge expression in automatic fault process has distinct characteristics, which need specific methods to manage by researchers. This paper conducts the research on knowledge management method of automobile stamping workshop, by building malfunction knowledge tree in order to visualize the knowledge manage process. This paper conducts the following research:(l)This paper puts forward a semi-automatic knowledge tree building method by deeply analyzing the characteristics of reasons for fault of stamping workshop. As the keywords of single-pass fault of stamping workshop are less, the faults have linkage between each other and fault records are somewhat vague and lack of standardization, the early collating and analyzing of fault knowledge is very difficult. By making a segment and exaction of fault knowledge records in stamping workshops, this paper concludes that it is suitable to build a fault knowledge tree by correlation analysis. This paper adopts a FP-Growth algorithm to analyze the fault knowledge with the assistance of experts. In stamping workshop, the faults of processes and production lines are different, so as the reasons, so stamping workshop can generate fault knowledge tree to meet the needs of procedures and equipments. This paper constructs the fault knowledge trees which face op10and op20procedures, and fault knowledge tree which are suitable for all procedures and equipments, then analyze them.(2) Fault knowledge tree needs update and improvement consistently to retain the synchronization of knowledge tree and manufacture knowledge. This paper puts forward an evaluated method of fault knowledge tree which based on Fuzzy Analytical Hierarchy Process FAHP. Through evaluating the nodes, branches and new nodes of fault knowledge tree, achieve the improvement and innovation of fault knowledge tree. By introducing evaluation indicators of fault knowledge innovation, determine the weight of the factors in the evaluation indicators system by FAHP method and complete comprehensive evaluation of improvement of fault knowledge tree through the multistage examine and verify strategy. Through the feedback of system function design, optimize the knowledge tree consistently, for the purpose of knowledge management and knowledge innovation in stamping workshop.