A multi-agent path planning algorithm based on hierarchical reinforcement learning and artificial potential field

Aiming at the problems of the path planning algorithm, such as slow convergence speed and low efficiency, a multi-agent path planning algorithm based on hierarchical reinforcement learning and artificial potential field was proposed. Firstly, the multi-agent operating environment was regarded as an artificial potential field, the potential energy of every point (the maximal rewards) was determine by the priori knowledge. Then, the update process of strategy was limited to smaller local space or high-level space of lower dimension to enhance the performance of learning algorithm by using the partial updates ability of hierarchical reinforcement learning. Finally, aiming at the problem of taxi, the simulation experiment of algorithm proposed by this paper was done in grid environment. To close to the real environment and increase the portability of the algorithm, this algorithm was verified in 3-dimension simulation environment. The results show that the convergence speed of the algorithm is fast, and the convergence procedure is stable.