1990 Fiscal Year Final Research Report Summary
A Study on Simulation for Planning of Batch Process Operation
| Project/Area Number |
01550732
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
化学工学
|
|---|
| Research Institution | Toyohashi University of Technology |
|---|
Principal Investigator |
ONOGI Katsuaki Toyohashi University of Technology Associate Professor Faculty of Engineering, 工学部, 助教授 (80115542)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
INOMATA Toshimitsu Toyohashi University of Technology Research Associate Faculty of Engineering, 工学部, 助手 (30213193)
NISHIMURA Yoshiyuki Toyohashi University of Technology professor Faculty of Engineering, 工学部, 教授 (70023075)
|
|---|
| Project Period (FY) |
1989 – 1990
|
| Keywords | Batch Process / Discrete Event System / Petri Net / Simulation / Operation Planning / Control |
|---|
| Research Abstract |
The dynamic behavior of a batch process is represented by describing how materials/products flow through it. Such systems can be regarded as discrete event systems. The state transition of a discrete event system takes place when and only when one of the enabled events actually occurs. If the set of enabled events is open to choice, the state transition is nondeterministic. A choice of an event among the enabled events for a state transition is called a resolution for the transition. The dynamic behavior of a discrete event system is characterized by synchronization, concurrency and nondeterminism of event occurrence.
We first interpreted the problem of operating a discrete event system as
(1) Model building - to build a model which includes a nondeterministic state transition rule
(2) Controllability analysis - to find the set of nondeterminism resolution strategies which satisfy the specified policy
(3) Decision process evaluation - to evaluate the system performance for each nondeterminism resolution strategy
It is desirable for us to have an appropriate discrete event system mode which can deal with all of these problems. Any such models have, however, not yet been presented. We second proposed a discrete event system model based on Petri nets and an object-oriented approach. It consists of a bipartite graph, marking, and a nondeterministic marking transition rule.
We last developed a computer aided design and analysis system for discrete event systems on the basis of the above model. It offers an environment for interactive control of simulation process. Such an environment seems to be useful to operation planning problem.
|
