| Budget Amount *help |
¥17,420,000 (Direct Cost: ¥13,400,000、Indirect Cost: ¥4,020,000)
Fiscal Year 2023: ¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
Fiscal Year 2022: ¥5,070,000 (Direct Cost: ¥3,900,000、Indirect Cost: ¥1,170,000)
Fiscal Year 2021: ¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000)
Fiscal Year 2020: ¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
|
|---|
| Outline of Final Research Achievements |
This study aims to understand human balance control by modeling the human motor control system and implementing it into a humanoid robot. Focusing on the motion synergy (the planar covariation), which is a characteristic of human gait, we introduced this feature as a reward in reinforcement learning. This approach suggested that the planar covariation is one of the factors characterizing human walking motions, such as heel-strike and toe-off walking. Inspired by human jumping movements, we devised a method to generate jumping motions for robots using linear model predictive control. This method includes inequality constraints ensuring the center of pressure remains within the support area and equality constraints preventing the robot from rotating during the flight phase. The jumping motion was successfully realized in both simulations and actual robots.
|
