RRT-GPMP2: A Motion Planner for Mobile Robots in Complex Maze Environments

TL;DR

RRT-GPMP2 tackles motion planning for mobile robots in complex maze environments by integrating GPMP2-based global trajectory optimization with RRT-based local re-planning. The GPMP2 component uses probabilistic inference on a vector-valued GP prior with obstacle costs, represented and solved via a factor-graph, while RRT provides collision-free refinements where the global path fails. The combined approach significantly reduces total planning time compared with standalone RRT and yields comparable path lengths, yielding smoother trajectories in practice. Demonstrations in MATLAB mazes and ROS-based maritime simulation with a WAM-V 20 USV validate the method's practicality and potential for real-world autonomous navigation in cluttered environments.

Abstract

With the development of science and technology, mobile robots are playing a significant important role in the new round of world revolution. Further, mobile robots might assist or replace human beings in a great number of areas. To increase the degree of automation for mobile robots, advanced motion planners need to be integrated into them to cope with various environments. Complex maze environments are common in the potential application scenarios of different mobile robots. This article proposes a novel motion planner named the rapidly exploring random tree based Gaussian process motion planner 2, which aims to tackle the motion planning problem for mobile robots in complex maze environments. To be more specific, the proposed motion planner successfully combines the advantages of a trajectory optimisation motion planning algorithm named the Gaussian process motion planner 2 and a sampling-based motion planning algorithm named the rapidly exploring random tree. To validate the performance and practicability of the proposed motion planner, we have tested it in several simulations in the Matrix laboratory and applied it on a marine mobile robot in a virtual scenario in the Robotic operating system.

Figures (9)

• Figure 1: Detailed demonstration of the selected USV platform in the Gazebo simulation environment in ROS bingham2019toward. The USV platform is floating on the ocean surface and it will perform a path-following mission to validate our proposed motion planner.
• Figure 2: Flow diagram of the proposed RRT-GPMP2 motion planner.
• Figure 3: Demonstration of the proposed RRT-GPMP2 motion planner and RRT global planning in the first and second motion planning problems Mukadam-IJRR-18Dong-RSS-16dong2018sparserrt-code-2014. Sub-figures (a) and (b) demonstrate the GPMP2 global planning and the RRT local re-planning in respective in the first motion planning problem. Sub-figure (c) demonstrate the RRT global planning in the first motion planning problem. Sub-figures (d) and (f) demonstrate the GPMP2 global planning and the RRT local re-planning in respective in the second motion planning problem. Sub-figure (f) demonstrate the RRT global planning in the second motion planning problem. The start position is represented in green, the goal position is represented in red, the GPMP2 generated path is represented in purple, the RRT generated path is represented in blue and the RRT tree branches are represented in cyan.
• Figure 4: Comparisons of the total time cost between the proposed RRT-GPMP2 motion planner and RRT global planning in both motion planning problems. Data in the figure is measured in milliseconds.
• Figure 5: Comparisons of the path length between the proposed RRT-GPMP2 motion planner and RRT global planning in both motion planning problems. Data in the figure is measured in meters.