An expert system for adaptive part routing in computer integrated manufacturing
This research introduces the RES, a Routing Expert System for solving part routing problems in computer integrated manufacturing systems (CIMS). The RES consists of three parts: (1) knowledge base. (2) heuristic knowledge, and (3) meta knowledge. The knowledge base contains all the static and dynamic information of the CIMS. The heuristic knowledge consists of a heuristic search strategy that deci- des the best routes for parts in real time. The meta knowledge acts like a human supervisor. It assesses the performance and decisions made by the heuristic knowledge, and based on the behavior the system, may make overriding decisions so that the best performance of the system can be achieved. In order to implement the RES, a knowledge based simulation model for the CIMS is also developed. Both the RES and the knowledge based simulation model are written in PROLOG and are run on an IBM PC-AT. When compared to random routing policy, the RES is clearly far more superior even when the meta knowledge part of the RES is disconnected. An interesting result observed is that when the random routing policy is added with the meta knowledge, significant improvements of the system can be achieved. The expert system deve- loped has the potential to be used for the control system of a truly unmanned CIMS.
School Location:USA - Ohio
Source Type:Master's Thesis
Keywords:routing expert system computer integrated manufacturing cims prolog
Date of Publication:01/01/1987