基于深度强化学习的农村物流运输路径自动选择研究

Research on Automatic Selection of Rural Logistics Transportation Route Based on Deep Reinforcement Learning

常规方法求解运输路径自动选择模型时,选择的运输路径的运输总成本较高,因此,提出基于深度强化学习的农村物流运输路径自动选择方法。通过分析农村流通供应链的特殊性确定目标函数,并基于目标函数设计运输路径自动选择模型,采用深度强化学习算法求解,设置状态-动作空间、奖励函数等要素完善动作选择过程,输出模型最优解,从而实现运输路径自动选择的目的。在实验论证中,所提方法所选运输路径的平均运输成本为8.35万元,相比对照方法更低。结果表明,设计的方法能够有效规划物流运输路径,降低运输成本。

When the conventional method is used to solve the automatic transportation route selection model,the total transportation cost of the selected transportation route is high.Therefore,the method of automatic selection of rural logistics transportation route based on deep reinforcement learning is proposed.By analyzing the particularity of the rural circulation supply chain,the objective function is determined,and based on the objective function,and an automatic transportation path selection model is designed based on the objective function.Deep reinforcement learning algorithm is used to solve the problem,and state action space,reward function and other elements are set to improve the action selection process.The optimal solution of the model is output,thereby achieving the goal of automatic transportation path selection.In the experimental demonstration,the average transportation cost of the transportation path selected by the proposed method is 83500 yuan,which is lower than the control method.The results show that the method can effectively plan the logistics transportation path and reduce the transportation cost.