Autonomous multi-swarm optimizer for dynamic environments inspired by the partial synchronization in coupled chaotic oscillator networks
| Project/Area Number |
17H06552
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Soft computing
|
|---|
| Research Institution | Utsunomiya University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-08-25 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
Fiscal Year 2018: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2017: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
|
|---|
| Keywords | 群れ形成 / 最適化問題 / 群知能 / 力学系 / 粒子群最適化 / 適応制御 / カオス / 動的最適化 / 非線形力学系 |
|---|
| Outline of Final Research Achievements |
Optimizers finding multiple optimal or quasi-optimal solutions in a single run are required especially for the real-world problems in dynamic environments. In this research, we proposed a novel multi-swarm optimizer, in which multiple subswarms can dynamically be organized around the optimal solutions. In particular, we modified particle swarm optimization by implementing a simple gravitational force between particles. In the proposed method, the particles not only form multiple subswarms but also escape from the corresponding subswarm, by which the proposed method was able to find near and distant optimal (or quasi-optimal) solutions. The mechanism of these search motions was revealed by analyzing the effects of the simple gravitational force. Furthermore, based on this analysis, the exploitation and exploration capability of the proposed method were controlled. As the results, the performance of proposed method was remarkably improved.
|
| Academic Significance and Societal Importance of the Research Achievements |
予測不可能な環境下における最適化問題では，複数の最適解もしくは実用上十分に優良な解を並列に探索する必要がある．従来，最適解を探索するエージェントを複数の群れに分割する手法が提案されているが，多くの場合，その群れの数や1つの群れあたりのエージェント数をユーザが事前に定める必要があった．
本研究では，群れが自律的に創発される新たな手法を提案し，従来は不可欠であったユーザによる群れ構造の定義を不要にした．さらに，提案手法はパラメータ数と計算コストが従来法よりも小さいにも関わらず，より高い性能を実現可能であることを明らかにした．本成果は多様な実問題におけるリアルタイムな最適化の実現に寄与するものである．
|
Report
(3 results)
Research Products
(2 results)
