Research on collaborate optimization of unmanned vehicles based on improved RRT*-APF algorithm

In the process of path planning of unmanned vehicles in complex environments, the traditional RRT* algorithm (Rapidly-Exploring Random Tree, RRT) has problems such as blind search, slow convergence, so on. To address this issue, this paper proposed a kind of fusion RRT*-APF algorithm (Artificial Potential Field, APF). It incorporates the attractive force from the goal and the repulsive force from obstacles into the node expansion, which guides node to rapidly expand toward the target point. Moreover, the Analytic Hierarchy Process (AHP) is adopted to determine the most appropriate factors influencing node expansion, as well as the coefficients affecting the attractive force and the repulsive force. By combining relevant performance metrics to establish an evaluation function, the algorithm with the highest score is selected for path planning based on various scenarios and maps. Ultimately, a robot experiment platform is built to evaluate the feasibility and stability of the improved algorithm by applying it to a real mobile robot scenario. Results demonstrate that the improved RRT*-APF algorithm, compared to the RRT algorithm, and the enhanced RRT-APF algorithm, produces smoother paths with reduced length and faster search speeds during the exploration process.