Mobile Robot Motion Planning in Multi-Resolution Lattices with Hybrid Dimensionality

Safe and efficient path planning for mobile robots in large dynamic environments is still a challenging research topic. In order to plan collision-free trajectories, the time component of the path must be explicitly considered during the search. Furthermore, a precise planning near obstacles and in the vicinity of the robot is important. This results in a high computational burden of the trajectory planning algorithms. However, in large open areas and in the far future of the path, the planning can be performed more coarsely.In this paper, we present a novel algorithm that uses a hybrid-dimensional multi-resolution state × time lattice to efficiently compute trajectories with an adaptive fidelity according to the environmental requirements. We show how to construct this lattice in a consistent way and define the transitions between regions of different granularity. Finally, we provide some experimental results, which prove the real-time capability of our approach and show its advantages over single-dimensional single-resolution approaches.