Yaojin Fan, Jiayu Li, Yufei Liu, Bo You, and Liang Ding.
Human–robot collaboration, path planning, legged robot, game theory
Legged robots are recognised as among the most viable configurations for traversing unstructured outdoor terrain, owing to their inherent stability and enhanced maneuverability. Nevertheless, the realisation of autonomous path planning within dynamic and heterogeneous outdoor environments presents significant challenges, thereby necessitating human–robot collaborative approaches as an effective solution. To address these complexities, a novel human– robot collaborative path planning methodology is proposed in this paper. Initially, a specialised human–robot collaborative path planning architecture is developed for deployment in unstructured terrain scenarios. Subsequently, to mitigate conflicts arising from divergent path selection criteria during the collaborative planning process, a game-theoretic human–robot path fusion method is introduced. Furthermore, to accommodate scenarios wherein environmental perturbations render the predetermined path infeasible based on assessment outcomes, a hybrid path planning optimisation framework is formulated to facilitate local path replanning and refinement. The efficacy of the proposed human– robot collaborative path planning approach is demonstrated through comprehensive human-in-the-loop experimental validation.
Important Links:
Go Back
