| Project/Area Number |
01460122
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
機械力学・制御工学
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
NAKANO Eini The Faculty of Engineering, Tohoku Univ., 工学部, 教授 (90198151)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KIMURA. h Hiroshi The Faculty of Engineering, Tohoku Univ., 工学部, 助手 (40192562)
|
|---|
| Project Period (FY) |
1989 – 1991
|
| Project Status |
Completed (Fiscal Year 1991)
|
| Budget Amount *help |
¥4,000,000 (Direct Cost: ¥4,000,000)
Fiscal Year 1991: ¥400,000 (Direct Cost: ¥400,000)
Fiscal Year 1990: ¥300,000 (Direct Cost: ¥300,000)
Fiscal Year 1989: ¥3,300,000 (Direct Cost: ¥3,300,000)
|
|---|
| Keywords | Geometical Model / Walking Robot / Knowledge Acquisition by Trials / 知識獲得 / 知識利用 |
|---|
| Research Abstract |
First, this research intended the automatic modeling of the object of which geometrical model is unknown. We constructed the robot vision system and proposed the new method for the automatic modeling using it, which is called "visual point selecting cone insertion method". It makes modeling effective and causes more precise model. Moreover, in the case of the objects overlapped on the screen, we proposed the new method for dividing them by getting the 3D division information actively, which is called "visual point selecting image partition method". We succeeded in making the individual models of the overlapped objects automatically.
By the other hand, we constructed a leg-wheel mobile robot which had mechanically separated wheels and legs to utilize advantages of the two mechanisms. We made this robot walk on the stairs and the irregular terrain and avoid the obstacles on the flat terrain using the ability of recognition about the environment for the planning and the control, which was developed at the automatic modeling research. This robot can detect the information about the environment automatically and can acquire the knowledge about the motion plan against the new situation.
When we divide the environment into the several parts, each part may become relatively simple. By such way, the walking robot which has the appropriate intelligence can act to the unknown environment autonomously. We have concluded that the knowledge acquisition of the walking robot is possible by the automatic modeling of the unknown environment and by the autonomous adaptation to the environment.
|
