Research On Abstract And Thinking Case In The Management Of Automobile Stamping Halting Problem

Stamping is one of the most important processes in automobile production. In stamping production, the production line will be forced to stop due to various reasons, it is called halting problem, it leads to a reduction in production efficiency and an increase in the cost of production. A lot of experience will be generated when the staffs deal with halting problems, if we can identify, acquire these knowledge and manage the knowledge effectively, halting problem will be greatly improved.This paper analyzed current status of the researches on halting problem, proposed to use the thought of CBR to realize the effective combination of halting problem management and knowledge management. On the basis, we analyzed the needs of front-line staff, the characteristics of halting problem itself, and the inadequacies of applying traditional CBR directly to the halting problem, two kinds of models are put forward:abstract case and thinking case, and they are applied to halting problem. Specific contents are as follows:(1) Abstract case model. Abstract case model is used to provide summary and general knowledge. This paper firstly put forward the concept and model representation of abstract case, gave a detailed analysis of the characteristics of abstract model, and described the process of building abstract case. In addition, we designed abstract case retrieval strategy according to the advantages of the combination of abstract case and specific case. Finally, examples show that abstract case reasoning is not only more efficient than traditional CBR in retrieving, but also proved that abstract case has stronger heuristic, it can improve knowledge level of the users, and stimulate their innovative ability.(2) The design of case similarity algorithm. Abstract case and thinking case reasoning cannot be separated from case similarity calculation. Therefore, this paper analyzed the case attributes in halting problem, designed similarity calculation method which conforms to the characteristics of halting problem data considering content and structure, then verified the rationality of the method though examples.(3) Thinking case model. Thinking case is used to provide the thinking of the problem. Thinking case model is used to solve the problem that one halting problem may have many similar cases or no similar case. This paper drew lessons from CCBR model, and put forward the framework of thinking case reasoning. We analyzed factors that influence thinking in halting problem, and designed corresponding ranking algorithm. Then we proved the feasibility of thinking case though specific examples.The research content of this paper has certain significance. In terms of practical significance, halting problem is effectively solved, production efficiency is improved, and the knowledge level of the staff is improved. In terms of theoretical significance, from the perspective of CBR, we expanded CBR ideas; from the perspective of problem-solving, we extended the perspective from the current problem to the problem class, and put forward the idea of combining the summary of knowledge and concrete examples to solve problems. From the perspective of knowledge management, this paper deepens the study of knowledge management to some extent. Deepen the study on the knowledge of "how to do " to the thinking process of "how to think".