Harmonic artificial potential field path planning for tracked robots in unstructured environments

Path planning is crucial for autonomous mobile robots, particularly in complex terrains like those in space exploration. While Artificial Potential Fields (APF) offer quick, reactive responses to dynamic environments, they often face issues like local minima and sub-optimal paths. Harmonic Artificial Potential Fields (HAPF) address these limitations with smoother and more complete solutions but are computationally intensive. This paper presents a hybrid approach that combines APF’s responsiveness with HAPF’s global completeness, formulating a path optimization framework that reduces sub-optimality. It also accounts for off-road tracked vehicle dynamics, including terramechanics and non-holonomic constraints. Using a real-case planetary exploration scenario, results demonstrate that the proposed algorithm provides a more comprehensive approach to path planning by taking into account various factors. For example, when compared to an existing planner from the literature, the proposed HAPF-based planner generates a path that optimizes the expected cost based on distance to obstacles, path length, average slope, and sinkage, with an 85% reduction in overall path cost.