Delivery Route Optimization Based on Deep Reinforcement Learning

With the rapid development of fast food industry, the research on the algorithm of delivery problem becomes increasingly important. The delivery problem of takeout is essentially the optimal path problem, while the traditional algorithm optimization of delivery path has been unable to meet the needs of modern takeout development. Based on this, this paper carries out the research on the delivery path optimization based on Deep Reinforcement learning algorithm. In this paper, we use the improvement method Heuristics in Deep Reinforcement learning to optimize the delivery path. In addition, we compare this method with the traditional tabu search algorithm, three distribution locations are selected and compared from two aspects of delivery time and customer satisfaction. The results show that using Deep Reinforcement learning algorithm to optimize delivery path can effectively reduce delivery time and improve delivery efficiency.