.png){kind=icon}
.png){kind=icon}
Return
An asymptotically optimal sampling-based method for high-dimensional complex motion planning using novel efficient heuristics
.png){kind=icon}
PRE
AI
DOI：10.1016/j.robot.2026.105462.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
From mobile robot navigation to robotic arm grasping, motion planning around obstacles plays a key role in modern robotics. Sampling-based motion planning methods can effectively plan robot paths in high-dimensional spaces. However, it is difficult to balance efficiency and planning quality for the methods when facing both cluttered environments and high dimensions. Here, we present an asymptotically optimal planner, the Informed Heuristic Bi-directional Fast Marching Tree (IHBFMT*), that can enable a strategy of heuristic search based on a bi-directional exploring tree to efficiently and reliably find high-quality paths around obstacles. Using the recent informed sampling technique in path optimization, we design a path optimization strategy by reusing previous motion planning information to further accelerate obtaining high-quality paths. Through challenging simulations and the 6-DOF manipulator grasping experiments, we demonstrate that the proposed method can find better paths in less time to solve high-dimensional complex motion planning problems.
Keywords:
motion planning
sampling-based methods
high-dimensional spaces
heuristic search
path optimization
Journal
IF:
5.2
Papers:
639
Citations:
1.0W