Manifold Approximation-Based Probabilistic Roadmap Approach for Constrained Path Planning of Mobile Manipulators
Xinsheng Tang, Weiwei Shang, Jiahao Hu, Fei Zhang, Liubo Kong · IEEE/ASME Transactions on Mechatronics · 2025
Constrained path planning task requires finding a collision-free path connecting the start and goal configurations while satisfying constraints, which is a challenging research area. This problem is commonly encountered in repetitive tasks with small variations. However, existing methods incur high computational costs and struggle to balance planning speed and path quality. This article proposes a manifold approximation-based probabilistic roadmap method. This method is a sampling-based planning approach that combines the approximate representation of constraint manifold with probabilistic roadmap method. It can address the constrained path planning problems in repetitive tasks and generate high-quality path within a relatively short planning time. Simulation and experimental results in a chemical laboratory setting demonstrate that the proposed method offers faster planning speed and higher, more stable path quality compared to other methods.