Algorithms for rough terrain trajectory planning
Résumé
| This paper deals with motion planning on rough terrain for mobile robots. The aim is to develop eecient algorithms, suitable for various types of robots. On rough terrain, the planned trajectory must verify several validity constraints : stability of the robot, mechanical limits and collision avoidance with the ground. Our approach relies on a static and kinematic model of the robot. EEcient geometric algorithms have been developed, taking advantage of each vehicle's speciicities. Motion planning relies on incremental search in the discretized connguration space and uses eecient heuristics based on terrain characteristic to limit the size of search space. Simulation results present trajectories planned in a few seconds. The second part takes into account uncertainties to improve trajectory robustness: uncertainties on the terrain model and the position of the robot. The adaptation of the previous algorithms allow to nd robust trajectories, without excessive time increase.
Domaines
Robotique [cs.RO]Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...