关键词: 物流工程, 绿色车辆路径模型, 遗传算法, 城市配送系统, 时间依赖

Abstract: This paper studies the vehicle routing problem in urban distribution systems considering traffic congestion and environmental pollution. The time dependence of the problem lies in the following aspects: different vehicle speeds under the road congestion lead to different transportation time at different departure times, thus resulting in great differences in transportation costs and environmental pollution. Therefore, this paper proposes a time-dependent green vehicle routing model to reduce transportation costs and environmental pollution by the optimization of vehicle routes and departure time. The model minimizes the total transportation cost including the fuel consumption cost, and the fuel consumption is measured based on the Comprehensive Modal Emissions Model. The innovation of the model lies in the defining of the situation that vehicles are allowed to wait at the nodes to choose the right time to avoid congestion, that is, by optimizing the vehicle routes and the departure time at each node on the path to find the cost optimal transportation scheme. In this paper, a nested genetic algorithm is proposed to solve the model. The outer genetic algorithm optimizes the path, and the inner genetic algorithm optimizes the vehicle departure time at each node of the path. The key parameters of the algorithm are debugged by response surface analysis method, and the best parameters for the model are obtained. The performance test results show that the algorithm is efficient. Finally, numerical experiments are carried out based on PRPLIB database. The experimental results show that the fuel consumption and total cost can be reduced to a certain extent by allowing the vehicle to wait at the customer's place and choose the right time to start off. In addition, the introduction of fuel consumption elements in the objective function can greatly reduce the fuel consumption of the decision-making scheme and reduce environmental pollution.

Key words: logistics engineering, green vehicle routing model, genetic algorithm, urban distribution, time dependence