A Path Planning Method Based on Collaborative Learning for Multi-Robot with Connectivity and Obstacle Avoidance Constraints

摘要

Cooperative path planning method is one of the research hotspots in the field of multi-agent. In this paper, we study the problem of cooperative path planning, in which the robots under the constraints of communication connectivity and collision avoidance, and choose the destination location independently. It is difficult to find the optimal solution efficiently in sparse solution space by large-scale complex environments. To overcome the problem, in this paper, a novel collaborative learning algorithm, named Park Algorithm (PA) is proposed, which makes full use of exploration experience to improve the exploration efficiency. First, an objective function with three goals including path length, communication and collision is designed. Second, Marker Matrix is designed to evaluate the exploration value of the grids in the map, and each robot uses its own and teammates’ historical experience to update Marker Matrix. Our method can make full use of the exploration experience, realize self-learning and mutual learning between the robots, accelerate the exploration of solution space and avoid falling into local optimal solutions. Finally, the effectiveness and superiority of the proposed method are verified by simulation analysis. The results show that our method has a better performance compared with Genetic Algorithm (GA), Particle Swarm Optimization Algorithm (PSO) and Ant Colony Algorithm (ACO).