Robot path planning in the presence of obstacles

Voir Plus Voir Moins

Afficher le Résumé

The design of algorithms that handle On-line motion planning of robot manipulators working in an unknown or known but changing environment is investigated in this thesis. Finding exact solutions to the motion planning problem is a very complex and tedious task. Trying to obtain such solutions on-line is very difficult and may not be useful for practical situations. The approach taken in this thesis is to use heuristic shortcut methods in order to find some acceptable not be largely generalized.solutions, although these solutions may The robot system and the exterior world are modeled by a col lection of convex polygons. All the information IS storedm the World Model where any change in the robot environment is reported. The planning is carried out in the joint space (or the Configuration space: C-space). An approximate mapping of the real obstacles into the C-space is performed. The forbidden regions or C-space obstacles are then used by the path planner in order to find a safe path connecting the initial and the final configuration of the robot system in the free C-space. Simulation results have proven that the path planning process is computationally practical for the planar robot arms.