Research On Multidimensional Network Model Of Automotive Stamping Problem Solving Knowledge

Automotive industry is in a leading position in the manufacturing industry, it is one of the core industries guiding the rapid growth of national economy. It has the features of highly industry relevance and talent, capital, technology and other resources highly aggregated. Automotive industry is facing unprecedented fierce competition in the market, under the serious situation of opportunities and challenges, only leading technology and knowledge innovation can keep its competitive advantage. Knowledge resources has gradually become the strategic resources of automotive enterprises, knowledge management has become the source of power for sustainable development of enterprises.Stamping process is a very important part in the four automotive processing technologies, the quality of stamping parts and production efficiency directly affect the quality of the vehicle and the cost of production. In stamping process, halting problem is the key factor that influences the efficiency of stamping production. Therefore, the effective management of halting problem is the key to improve production efficiency and overall production operation ability. Frontline staff accumulated a large amount of factual data in the process of dealing with halting problems, These data contains empirical knowledge, also indirectly reflects management problems behind the halting problems. How to organize and use the knowledge effectively to provide decision supports for operational and management decision-makers, reduce halting problems from occurring is the core of the halting problem management.Based on knowledge management, started from the characteristics of halting problem solving facts, this paper established the Problem Solving Knowledge Multidimensional Network Model, which contains multiple dimensions and relations between dimensions, describes different size of correspondence between problem-related factors. Based on this model, we presented knowledge extraction and application methods to meet different decision-makers’needs for part of the problems to be solved in practical production decisions.In the proposing part of the model, we considered the important role problem contexts play in the study of problem, as well as decision-making needs in different sizes. Use the levels of the dimension network to represent different size of problem-related factors, Use correspondences between the dimension networks to represent relationships between the factors, thus, we established Problem Solving Knowledge Multidimensional Network Model which describes problem solving knowledge completely.In the application part of the model, centered on different dimension network, based on relationships between different dimension networks, we can provide support for solving specific problems in work site and enterprise strategic planning by extracting multi-angle, multi-granularity knowledge. This paper mainly studied three application modes, including multidimensional cognitive of problem context, analysis of the problem status, and importance evaluating of solutions. At last examples are analyzed by using real production data from a car manufacturing plant to verify the feasibility and practicability in decision-making of the Problem Solving Knowledge Multidimensional Network Model.