Behavior Acquisition of Multi-legged Robots by CPG and Reinforcement Learning in Multi-task Environment
| Project/Area Number |
15K00355
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Soft computing
|
|---|
| Research Institution | Matsue National College of Technology |
|---|
Principal Investigator |
HORIUCHI Tadashi 松江工業高等専門学校, 電子制御工学科, 教授 (50294129)
|
|---|
| Research Collaborator |
KISHIMOTO Ryoichi
SASAKI Hikaru
|
|---|
| Project Period (FY) |
2015-04-01 – 2018-03-31
|
| Project Status |
Completed (Fiscal Year 2017)
|
| Budget Amount *help |
¥4,680,000 (Direct Cost: ¥3,600,000、Indirect Cost: ¥1,080,000)
Fiscal Year 2017: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2016: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2015: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
|
|---|
| Keywords | 知能ロボティクス / ニューラルネットワーク / 強化学習 / 行動獲得 / 複数タスク |
|---|
| Outline of Final Research Achievements |
In this research, we realized that real multi-legged robots themselves acquired behaviors to perform multiple tasks such as avoiding the obstacles and reaching the target area, by using Central Pattern Generator (CPG) and reinforcement learning. First, we realized each learning module to perform each task, by using CPG and reinforcement learning. Next, we constructed behavior learning system for multi-task environment, by adding the upper module to integrate the outputs of lower learning modules. We implemented this behavior learning system for six-legged robot, four-legged robot and snake-like robot. Through the experiments for three kinds of real robots, we confirmed the effectiveness of our behavior learning system using CPG and reinforcement.
|
Report
(4 results)
Research Products
(34 results)
