The Development Of Logistics Discrete Event Simulation Platform Based On Expert System Modelling Tools

Expert system is an important branch of artificial intelligence. Expert system has a wide range of theory and practice progress after its development more than 40 years since Fagan Baum developed the first expert system Dendel. Expert system is a unique methodology to describe system at all, ie regarding the whole system as one composed of various facts and rules based on the rule-based describing methodology. Among these elements, "facts" forms the framework of the system that describes the static character at specific moment while "rules forms the core, making the system run smoothly. Existing facts produce new facts and remove or modify the old facts, while new facts invoke new rules. As a result, the final solution of problem given through facts will be reached by interaction between facts and rules.Modelling tool for expertise system is the achievement that expertise system gain as it's mature to some extent. In the initial stage of the expertise system, all modules in it need to be rebuilt from scratch including reasoning machine, knowledge library, human - machine interface every time a new system is developed. Later on, a new solution that some general modules such as reasoning machine, human - machine interface integrated into a development tool is adopted. By doing so , these tools can be used directly and developers can be absorbed in building the core modules the specific expertise composed of, such as knowledge library without paying attention the the implementation of other general modules. After the existence of such solution, efficiency of developing expertise system is greatly promoted and the usage of such system is widened. With a complete set of development language provided, users can programme with the development language ie the modelling tool because these programming language are integrated already. Based on this, the modelling tool can be regarded as specific programming language.The system whose status changes only on some discrete point - in - time can be called discrete event system. As many systems in daily life can be simulated by discrete event system and such system is more suitable for simulation by computer compared with continuous system, it's widely applied. Logistics discrete event system simulation is an important branch of discrete simulation which mainly focus on the real-time status and regularity of temporary entity( represent material, product and customers in respective system) in permanent entity (equipment and desk), thus building and optimization advise are given through it. System description is the base for system simulation. Limited status machine and Petri network are usually adopted which is characterized by complex expression of knowledge and difficulty in system maintaining. Focusing on these shortcomings, this article proposes a new methodology for forming logistics discrete evewnt system after studying respective character and internal relationship between expertise system and discrete event system. The methodology mentioned above describes system through rules and build the logistics simulation system through expertise system modelling tool. Last but not least important, the simulation effect and excellent expansibility is proved by testing and function can be easily extended in the future.