| Project/Area Number |
11680462
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
社会システム工学
|
|---|
| Research Institution | Hiroshima Shudo University (2001)
Hiroshima Shudo Junior College (1999-2000) |
|---|
Principal Investigator |
SAKAI Takahama,Setuko Faculty of Commercial Sciences, Hiroshima Shudo University, professor, 商学部, 教授 (60186989)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KAIO Naoto Faculty of Economic Sciences, Hiroshima Shudo University Professor, 経済科学部, 教授 (80148741)
HIROMITSU Seijirou Faculty of Economic Sciences, Hiroshima Shudo University Professor, 経済科学部, 教授 (90043827)
KODAMA Masanori Faculty of Economic Sciences, Hiroshima Shudo University Professor, 経済科学部, 教授 (20028989)
YONEDA Kunihiko Faculty of Economic Sciences, Hiroshima Shudo University Professor, 商学部, 助教授 (10201865)
FUJITA Tatehiko Faculty of Economic Sciences, Hiroshima Shudo University Professor, 短期大学部, 教授 (50105649)
|
|---|
| Project Period (FY) |
1999 – 2001
|
| Project Status |
Completed (Fiscal Year 2001)
|
| Budget Amount *help |
¥3,500,000 (Direct Cost: ¥3,500,000)
Fiscal Year 2001: ¥1,200,000 (Direct Cost: ¥1,200,000)
Fiscal Year 2000: ¥1,100,000 (Direct Cost: ¥1,100,000)
Fiscal Year 1999: ¥1,200,000 (Direct Cost: ¥1,200,000)
|
|---|
| Keywords | nonlinear optimization method / multiobjective optimization problem / constrained optimization problem / Pareto optimal / genetic algorithm / degeneration / structural optimization / mutant genc / Vector Simplex法 / Pareto最適解 / 対話的最適化 / チームモデル / MGGA / α制約法 / Simplex法 / 規則学習 / パラメータ・チューニング / ファジィ制御 / 直接探索法 / Powell法 |
|---|
| Research Abstract |
In this study, the research on the development of optimization method for the general and nonlinear multiobjective optimization problem of which the differentiability was not guaranteed was made to be a purpose. The research 3 years is approximately classified into the following three groups :
(1) The research on optimization methods for general constrained single objective nonlinear optimization problems,
(2) The research on optimization methods for general unconstrained multiobjective nonlinear optimization problems,
(3) The research on the other general optimization methods.
In the researches (1), the first, we proposed the transformation method "α constraint method" converting a constrained nonlinear optimization problem into an unconstrained one. Next, we proposed two methods, "the α constraint Powell method" which combined Powell and α constraint method and "the α constraint Simplex method" combined Simplex and α constraint method, and compared these method with Penalty method. From
… More
the comparation, it was shown that the α constraint method was a general-purposive transformation method in search of the solution at high speed under the condition of satisfying given constraint satisfaction.
In the research (2), we proposed "the Vector Simplex method" in search of the approximate set of all Pareto optimal solutions for a multiobjective nonlinear optimization problem (MOP). In MOP, only the partial order relation can not be guaranteed among the values of objective functions. The Vector Simplex method expanded Simplex method which was optimization method for nonlinear and single objective optimization problem as an optimization method for the multiobjective optimization problem, and it is an optimization method for MOP only using partial order of the objective function value vector. Bydivision and segmentation of the search space and serial generation of search points, it was improved in order to obtain the approximate set of all Pareto optimal solutions which were uniformly distributed. And, we also carried out the development of the interactive operation, since the intention of the decision maker is differently reflected.
In the research (3), the general optimization method "Team Model" which modeled the learning by team was proposed, and in addition, "Genetic Algorithm (GA^d) with Degeneration" which introduced damaged gene for the structural optimization of the model was proposed with "genetic algorithm (MGGA) which introduced the mutant gene". It was shown that each method is respectively more effective than the conventional genetic algorithm. And, using the TD method, the research on the acquisition of optimal strategy of each player in the competitive situation in the game society was also carried out. Less
|
