| Project/Area Number |
13650446
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
System engineering
|
|---|
| Research Institution | KOBE UNIVERSITY |
|---|
Principal Investigator |
HATONO Itsuo Kobe University, Information Processing Center, Associate Professor, 総合情報処理センター, 助教授 (10208548)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
UEDA Kanji Kobe University, Faculty of Engineering, Professor, 工学部, 教授 (50031133)
藤井 信忠 神戸大学, 工学部, 助手 (80332758)
|
|---|
| Project Period (FY) |
2001 – 2002
|
| Project Status |
Completed (Fiscal Year 2002)
|
| Budget Amount *help |
¥3,200,000 (Direct Cost: ¥3,200,000)
Fiscal Year 2002: ¥1,600,000 (Direct Cost: ¥1,600,000)
Fiscal Year 2001: ¥1,600,000 (Direct Cost: ¥1,600,000)
|
|---|
| Keywords | Scheduling / Interactive / Ucertainty / Reinforcement Learning / order-to-make production |
|---|
| Research Abstract |
Today the manufacturing demand changes dynamically due to the variety of consumers' needs and the quick change of trends. On the other hand, the opportunity of the cooperation between companies or different departments in the development and the manufacturing of products increases, because of the companies' globalization and the separation of works.
In the presented research work, an interactive scheduling system which has following two main characteristics is proposed. First one is a scheduling method which can adapt to dynamical manufacturing environments and obtain the schedule quickly. Second one is to build an interactivity which can cooperate among companies or different departments.
As the indispensable method for the interactive sch eduling system, we propose the scheduling method which takes into account uncertainties. We develop the scheduling method which adjusts a margin of due date of each product while taking into account the future fluctuation of the manufacturing environment.
The proposed interactive scheduling method is applied to DVD make-to-order manufacturing system which is an example of dynamical manufacturing environments. We develop the scheduler using data from real factory to confirm the feasibility of the proposed method. From experimental results, we observe the increase of the number of accepted order and the decrease of tardiness.
To keep a margin of due date appropriately, we should change the size of initial margin according to the manufacturing condition. We propose reinforcement learning as a method that sets initial margin size, and we confirm the increase of accepted order in experimental results.
Obtained results show the effectiveness of the proposed scheduling method by taking into account uncertainties.
|
