RRT-Trained DDQN Algorithm for Pathfinding Problem of Automated Guided Vehicle

This study proposes an innovative path planning method that using the Rapidly-exploring Random Tree (RRT) to guide the training of the Double Deep Q Network (DDQN), to improve the learning efficiency and convergence speed of Deep Reinforcement Learning (DRL). The experimental results show that this method significantly improves the performance of DDQN in complex and high-dimensional task environments, providing a new perspective for the field of deep reinforcement learning. This research demonstrates the potential of combining classic path planning algorithms and modern deep learning technologies.