UAV Path Planning Based on APF-Q-Learning

With the broadening of UAV application fields, the working environment of UAVs has become more and more complex. Intensive, dynamic and non-convex are the main characteristics of the obstacle environment under the new demand, and the complex obstacle environment brings great challenges to the working operation and flight of UAVs. This paper puts forward a reinforcement learning algorithm named APF-Q-learning algorithm, which is the combination of the artificial potential field (APF) method and the Q-Learning algorithm, and the reward function is designed to make the value function table converge faster. The simulation results also show that the proposed algorithm can better solve the problems of local optimum and slow convergence of the value function.