| Paper Details: | Downloads: 1224 |
| Serial Number: | P1131131674
|
| Title: | LRF Assisted Force-based Walking for Hexapod Robot COMET-IV
|
| Authors: | Mohd Razali Daud and Kenzo Nonami
|
| Abstract: | This paper describes the method employed in vision assisted force control for a hexapod robot, named as COMET-IV. Previously, the control algorithm has used fixed stable range value with alternate Pull-back ON/OFF to avoid leg stretching or clamping over the angles limitation that has been outlined for COMET-IV system, as proposed in [4]. However, the method almost fails during walking in ‘sometimes changes’ ground level terrain. This article introduced a method, named as Grid-based Walking Assistant for Legged Robot (GWALR) to enable the control algorithm dynamically alters the stable range according to the ground level. The mentioned method uses the obstacles geometric determined based on the 3D point cloud data acquired using rotating LRF for that purpose. The geometric of the obstacles in the environment is represented using the occupancy grid cells, which are categorized into three categories depends on the height of the obstacles exist in each cell. The distances of the obstacles relative to the center of the robot’s body are computed using a customized edge detection method, named as Grid-cell Edge Detection (GCED) method. The experiment results show the performance of the proposed methods are successfully assisted the position-based force control algorithm to maintain the body in horizontal condition at all times during walking.
|
| Keywords: | Grid-based mapping, Grid-cell edge detection,
Vision assisted force control.
|
| Journal/Conference: | International Journal of Automation, Robotics and Autonomous Systems
|
| Volume: | 11
|
| Issue: | 1
|
| Submission Date: | 8/2/2011 12:00:00 AM
|
| Review Date: | 10/10/2011 12:00:00 AM
|
| Publishing Date: | 10/26/2011 12:00:00 AM
|
| Article Downloads: | 1224
|
| Download: |
|