Ph.D.s' Dissertations A Study on the Integration of Process Planning and Scheduling
2015.07.29 12:16
| year | 1997 |
|---|---|
| author | Kidong Kim |
| Keyword | Integrated Manufacturing System, Process Planning, Scheduling, AND/OR Graph, Branch and Bound |
| Abstract | In most manufacturing environment, manufacturing activities can be classified into two main domains. One is the domain of manufacturing technology and the other is the domain of manufacturing planning and control. Design, process planning, and manufacturing activities belong to the former, and MPS, MRP and scheduling belong to the latter. Traditionally, activities of the above two domains were performed separately. However, they should be integrated to make manufacturing systems more efficient. In this thesis, we endeavor to integrate the process planning and scheduling activities as an attempt to integrate the two realms which traditionally had been treated separately. To draw up a plan of process planning and scheduling in real manufacturing environment is not an easy task because available time to plan could be limited and the shop status could change frequently. So we propose an architecture of integrated process planning and scheduling system which can settle the integrated process planning and scheduling problem within the allowed time even if shop situations change rather frequently. Also we propose solution algorithms for the problems derived by the system. The inputs to the proposed system (named as IPPSS, Integrated Process Planning and Scheduling System) are the shop status, the information about orders, the allowed time for the plan, and the AND/OR graph which represents the process plan of part. The IPPSS is composed of three main modules and one supplementary module. The main modules, which are directly related to process planning and scheduling are the emergency module, the iterative module, and the integrated module. The supplementary module can convert the data used in the integrated system to EXPRESS language and vice versa. Each of these three main modules can generate solution but the quality of solution from the integration module is better - i.e. more optimum - than iterative module, and also the iterative module better than from emergency module, etc. However the time to get solutions gets longer in reverse order. One of three main modules is selected based on the allowed time for the plan in this proposed system so that we can always plan the best solution for each condition. The emergency module solves the process planning and scheduling problem one by one. In this module, the AND/OR graph is modified to ´AND/OR graph with time window´. And then, an algorithm which finds the shortest path from the start node to terminal node is applied to the ´AND/OR graph with time window´. The iterative module generates the process plan and the schedule sequentially and fixes the process plan after verifying the performance of the schedule using simulation study. In this module, the process plan is generated by selecting the operations to be performed and the machines to manufacture parts with the objective of balancing the weighted load. And then, a simulation study is performed using dispatching rules and the makespan is determined from the result of the simulation study. The checking procedure checks the terminating condition and updates the weights of the machines if necessary. These procedures of process planning, simulation, and checking are repeated until the terminating condition is met. The process planning procedure uses branch and bound method, and the Lagrangian relaxation method is used to find the lower bounds. In the integrated module, the optimal solution of the process planning and scheduling is made. The branch and bound algorithm is used and the branching strategy is concretely related with the problem structure. The experimental results of the proposed system are also included. |
| c | PhD |
Download :
