D. Um
Sensor-based planning, randomized sampling, unknown environment motion planning, collision avoidance
Path planning in unknown environments such as unexplored planets or underwater is not only a challenging but also daunting task, especially for a multi-degree manipulator. The most feasible solutions proposed so far are either modeling unknown environments in real time or applying randomized model-based planner (MBP) to tackle unknown environments. As for the latter case, however, no significant study has been carried out as to what are the issues for conversion of a randomized MBP to solve an unknown environment planning problem. In this paper, we investigate essential technical issues when tackling a manipulator planning problem in unknown environments using a senson-based planner (SBP). Second, we empirically study the sensing range effect of a sampling-based planner when used in an unknown environment as well. Noticeably evident in the study was that the larger sensing range exponentially diminishes the time of convergence, thus ensuring probabilistically faster planning of a manipulator in unknown environments.
Important Links:
Go Back
