2006 Fiscal Year Final Research Report Summary
Whole-body grasping of humanoid robots
| Project/Area Number |
16360128
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent mechanics/Mechanical systems
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology |
|---|
Principal Investigator |
KAGAMI Satoshi National Institute of Advanced Industrial Science and Technolog, Digilal Human Research Center, Senior Research Scientist, デジタルヒューマン研究センター, 研究チーム長 (30344196)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
NISHIWAKI Koichi National Institute of Advanced Industrial Science and Technolog, Digilal Human Research Center, Research Scientist, デジタルヒューマン研究センター, 研究員 (50359650)
MIYAKOSHI Seiichi National Institute of Advanced Industrial Science and Technolog, Digilal Human Research Center, Research Scientist, デジタルヒューマン研究センター, 研究員 (20358120)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Keywords | Humanoid Robot / Hand / Redundant Manipulator / Inverse Kinematics / Manipulability / Object Recognition ＆ Positioning |
|---|
| Research Abstract |
This research aimed for realizing autonomous grasping of objects by using vision, tactile, and force sensors on full body humanoid with 2 legs and 2 arms.
Research topics described in the following were carried out :
1.Developemnt of hands that have force sensors, can support robot's own weight and can be implemented on a full size humanoid.
2.Memory based inverse kinematics solver that takes manipulability into consideration. The solver was developed for generating whole body reaching trajectories that maintain high manipulability and keep away from the joint angle limitations.
3.Target object detection and position measurement system by using stereo vision. Matching between shape model of objects and measured data was executed quickly by using rendering function of graphics hardware and correlation method. In addition, feature (edge) based matching system was developped also by using the graphics hardware in order to achieve more precise positioning and robustness to the partial occlusion.
4.Whole body balance control system while exerting force to target objects.The system consists of 1[ms] cycle impedance control of hand position, 20[ms] cycle dynamically stable walking pattern generation which get the feedback of force measured at wrists, and automatic adjustment of stepping position in order to maintain long-term balance.
|
