Production Scheduling Expert System And Its Realization Technologies Based On Interactive Simualtion

The optimal scheduling of shop floor is an important measure to improve the enterprise competitiveness and its response ability to the market. However, in an increasingly complex production environment, there are some new characteristics in production scheduling, such as complexity, dynamic randomness, multi objectivities, and so on. The traditional scheduling optimization algorithms have been challenged. How to reflect the above-mentioned characteristics of production scheduling accurately and comprehensively has become a major subject in both academic circles and business circles.Aiming at the limitations of traditional scheduling optimization algorithms and the disadvantages in the Knowledge Base Scheduler(KBS)(such as the deficiency in knowledge acqusiton means, the deficiency in knowledge elicitation means, the limitness of application mode, and so on), this paper proposed a research on production scheduling expert system and its realization technologies. The main contents of the paper are as follows.It proposed an implement principle of the production scheduling expert system based on interactive simulation, presented the technical route and key technologies of the system realization, and put forward system working process and application mode based on Web Service.It proposed a construction process of manufacturing system simulation model and interactive simulation model, designed stability judgment algorithm and expert decision selection algorithm, and introduced the realization process of the interactive simulation by using an example.It detailedly studied the technologies to realize the production scheduling expert system. These technologies contain the following four sub-technologies.①Considering the complexity and diversity of production scheduling influence factors, a dimension reduction technology of production state vector was presented.②Aming at the multi-mode nature of variables in production control strategies, a computing technology combining Fisher Classifier with Regression RBF Neutral Network was proposed.③In view of the deficiency in knowledge elicitation means of traditional production scheduling expert system, a reference engine rule extraction technology based on EE-TREE algorithm and pre-text technique was put forward.④Focusing on the limitations of application mode of traditional productin scheduling expert system, the program realization of the production scheduling expert system that using double com object structure packaging and an application mode based on Web Service was proposed.Taking the production scheduling of engine production line in a motorcycle enterprise as research background, we verified the feasibility of the proposed principles and techniques. And the results showed that the production scheduling expert system based on interactive simulation technology is feasible and operational in practice. It provides a new approach for intelligent scheduling.