【摘要】 针对车物流网络问题，结合多式联运背景，对多目标下的路径选择问题进行研究；针对同时接受多个运输任务的应用场景，对含有区间参数的路径选择问题进行研究；首先，考虑运输过程中的运输时间和转换成本的不确定性，以运输的费用、运输总时间、运输碳排放量为目标，考虑节点的班期约束和收货人的时间窗约束，建立多任务的路径优化模型；然后设计带全局存档的NSGA-Ⅱ多目标进化算法，为了更好地迭代寻优，采用基于优先度排序的间接编码的方法，并通过Python编程求解一汽物流汽车运输路径，最后通过模糊决策方法选择合适的运输方案。结果表明：带全局存档的NSGA-Ⅱ多目标进化算法相较基础NSGA-Ⅱ算法提高了16%计算效率，运输决策者可以根据对各个子目标的重要性在Pareto解集中选择最佳方案。更多还原

【Abstract】 Based on the problem of vehicle logistics network and combined with the background of multimodal transportation, this paper studied the path selection-making problem under multi-objective. Aiming at the application scenario of accepting multiple transportation tasks at the same time, the path selection problem with interval parameters was studied. First, considering the uncertainty of transportation time and conversion cost in the transportation process, taking the transportation cost, the total transportation time and the transport carbon emission as the target, the shift constraint of the node and the time window constraint of the consignee were considered and a multi-task path optimization model was established. Then NSGA2 multi-objective evolutionary algorithm with global archiving was designed. For better iterative optimization, the indirect coding method based on priority ranking was adopted, and FAW logistics vehicle transportation path was solved by Python programming, and finally the appropriate transportation scheme was selected by fuzzy decision-making method. The results show that NSGA2 multi-objective evolutionary algorithm with global archive improves the computational efficiency by 16% compared with the basic NSGA-Ⅱ algorithm, and the transportation decision maker can choose the best scheme in the Pareto solution set according to the importance to each sub-target.更多还原