Knowledge-based master production scheduler for the computer-integrated manufacturing system.

摘要

This dissertation investigates applying the manufacturing resources planning (MRP II) system concept and Artificial Intelligence techniques to solve the master production scheduling problem in a computer integrated manufacturing (CIM) environment. The master production scheduling problem is defined as: disaggregating the production plan to determine the requirements for final products by date and quantity. Its objectives are to support the production plan, meet all the demand, control inventory level, and control resource usage level. In order to develop an intelligent master production scheduler, a knowledge based system (KBMPS) was developed.;The control algorithm proposed in this system includes eight phases. They are: (1) Net demand calculation; (2) Aggregation production plan validation; (3) Item disaggregation; (4) Inventory constraint checking; (5) Rough cut capacity planning; (6) Capacity requirements planning; (7) Component requirements planning; (8) Rescheduling analysis.;The KBMPS knowledge base is constructed with a three-level hierarchy, which includes: six rule categories, sixteen rule sets, and fifty-three rules. Also, it has two types of knowledge: algorithmic knowledge which can assess the system states and conduct the scheduling process, and analysis knowledge which can detect constraint failure and suggest appropriate solutions.;The KBMPS system is able to perform just as an experienced master production scheduler can. In addition, the capacity requirements planning, component requirements planning, and rescheduling analysis functions enhance the system's capability and guarantee the output of a feasible master production schedule.