1993 Fiscal Year Final Research Report Summary
Optimal Tool Planning on Flexible Production System
| Project/Area Number |
04832008
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
社会システム工学
|
|---|
| Research Institution | The University of Tsukuba |
|---|
Principal Investigator |
SUZUKI Hisatoshi Tsukuba Univ., Institute of Socio-Economic Planning, Professor, 社会工学系, 教授 (10108219)
|
|---|
| Project Period (FY) |
1992 – 1993
|
| Keywords | Machining Center / FMS / Tool Change / Set-UpTime / Processing with no setup / Group Technolgy / Mathematical Model / Optimization Theory |
|---|
| Research Abstract |
A machining center is installed many tools in its tool magazine and can successively process a variety of items. Its associated Automatic Tool Changer (ATC) selects and changes a group of tools used to process each item automatically. When a processing itm is switched to another one, the machine is stopped and is set up another group of tools to process a subsequent item. Therefore, frequent switches of processing items bring a large increase in set-up time and a large decrease in machine productivity.
In such a situation, by making a group of items with many common tools and by setting all tools necessary to process all items in a group, the machining center can process these items with no tool change. A tool change is necessary only to switch a group of items. Each group of items is called "parts family" and its corresponding set of tools is called "tool module". The tool module system brings to reduce the number of tool changes and much loss time in their set-up.
In this research, we discuss an problem to find the optimal set of tool modules minimizing the number of tool changes. The problem is formulated on a bipartite graph and solved by mathematical programming techniques. The proposed algorithm is developed as a software in a NEC supercomputer.
We already developed an algorithm to find the optimal tool module brought the successively longest processing time without tool change. At last year we extended it to our new concerned problem. The extended algorithm was developed here and was implemented on a microcomputer.
At this year, many testing problems have been solved on the supercomputer to verify performances of the proposed algorithm. Through the results the algorithm has been improved more and more. Finally, applying the algorithm to a real world problem, it seems that our new algorithm is very effective and practical.
|
Research Products
(8 results)
