In order to improve the efficiency and economic benefits of rural logistics transportation, a new planning method for the optimal movement path of logistics transportation vehicles is proposed by using deep reinforcement learning. Firstly, the road and logistics information data in rural areas are collected and applied to rural logistics transportation path planning, vehicle scheduling and other practical work. Second, deep reinforcement learning is used to model the environment so that the algorithm can learn and find the optimal path. On this basis, the optimal vehicle movement path model with time window is constructed, and a reasonable distribution path scheme is selected for each vehicle to minimize the total transportation cost of all vehicles. Finally, the optimal movement path planning algorithm for rural logistics transportation vehicles is designed. The examination outcomes demonstrate that this technique outperforms the conventional method when it comes to both overall transportation expenses and duration, subsequently boosting the efficiency and financial returns of rural logistics transportation.