基于深度学习的机器人移动路径规划算法OA

In response to the low efficiency of traditional deep Q network(DQN)applied to mobile robot path planning,an im-proved DQN algorithm is proposed.By combining the environmental perception ability of deep learning with the decision-mak-ing capability of reinforcement learning,intelligent path planning for mobile robots in complex environments is achieved.The improved DQN algorithm can be used to record obstacles encountered during the learning process,enhance the reward function,and design a dynamic exploration factor function.The proposed mobile robot path planning algorithm is compared with A*,rapidly-exploring random tree(RRT),and traditional DQN algorithms,and the results show that the improved DQN algorithm effectively enhances the path planning efficiency of mobile robots.