The popularization of unmanned aerial vehicle (UAV) technology has made the obstacle avoidance and path planning system crucial. Traditional obstacle avoidance relies on manual operation by pilots and struggles to adapt to complex environments. In contrast

the sensor-based autonomous obstacle avoidance technology enables UAVs to avoid obstacles autonomously by using various sensors. In recent years

reinforcement learning

as an emerging artificial intelligence technology

has been widely applied to UAV obstacle avoidance. It allows UAVs to interact with the environment and automatically learn the optimal flight strategies. There is no need to hard-code control signals; only action feedback is required. This feedback guides UAVs to choose better control behaviors. Through iteration

the best strategy is found to achieve precise flight. Therefore

reinforcement learning has brought innovation to UAV obstacle avoidance and expanded the application of UAVs in various fields.