Collision-Free Path Planning for Cooperative Aerial Manipulators Under Velocity and Curvature Constraints

摘要

This paper presents a path planning problem with velocity and curvature constraints for cooperative aerial manipulators in obstacle environments. While potential-flow-based path planning approaches can help generate smooth paths in which cooperative aerial manipulators can avoid obstacles, they do not facilitate the consideration of velocity and curvature constraints. By contrast, while such constraints can be considered in the optimization approach, the approach often requires a heavy computational load. To overcome these drawbacks, we propose a simple approach based on curvature-constrained harmonic potential flow and a streamline-changing algorithm. We also handle the initial and final velocity problem by a smooth activation function, which is crucial for the tracking performance at the initial and final position. The proposed approach can be used to generate a smooth path for cooperative robots in real time. For the validation of the approach, the simulation results obtained with the proposed algorithm were compared with those acquired with an existing approach including dynamic movement primitives(DMPs). Furthermore, we performed load-carrying experiment using custom-made aerial manipulators.