多式联运路径问题的一种混合算法  

A hybrid algorithm for multimodal transport routing problem

作　　者：裴骁 芦有鹏[1] 刘斌[1] 罗海燕[1] 许乃萱 PEI Xiao;LU You-peng;LIU Bin;LOU Hai-yan;XU Nai-xuan(College of Traffic and Transportation,Lanzhou Jiaotong University,Lanzhou 730070,China)

机构地区：[1]兰州交通大学交通运输学院,甘肃兰州730070

出　　处：《甘肃农业大学学报》2020年第5期235-244,共10页Journal of Gansu Agricultural University

基　　金：国家自然科学基金项目(71671079);兰州交通大学青年科学基金项目(2018018).

摘　　要：【目的】针对多式联运过程中运输路径和运输方式的优选问题,以运输成本最少为目标,建立了含运到时限约束和路径容量约束的组合优化模型.【方法】针对模型设计了双信息素蚁群-遗传混合算法进行求解,并引入最大最小蚂蚁策略、自适应灾变算子等操作,适时规避局部最优解,以提高算法优化能力.【结果】将混合算法求得的结果与单遗传算法、蚁群算法进行比较,混合算法较另两者的收敛稳定性平均提升15.534%和12.537%,且求得的20次最小费用平均值分别降低了3.18%和1.38%.【结论】采用上述方法探讨运到时限、容量约束以及货物运量对路径选择的影响,从而为多式联运经营人决策提供了参考.【Objective】Aiming at the problem of optimal transport routes and modes selection in multimodal transportation,a combined optimization model with time-bound constraints and path capacity constraints is established.【Method】A dual pheromone ant colony-genetic hybrid algorithm is designed for the model,introduce the maximum and minimum ant strategy,annealing penalty function and other operations to avoid the local optimal solution and improve the capabilities of the algorithm.【Result】Compare the hybrid algorithm with the single genetic algorithm and ant colony algorithm,the results show that the convergence stability can be improved 15.534%and 12.537%on average,and the average of 20 minimum expenses decline 3.18%and 1.38%,respectively.【Conclusion】With the proposed method,the effects of arrival time,capacity constraints,and cargo volume on route selection are explored,which provide a reference for multimodal transport operator decisions.

关 键 词：多式联运 路径优化 遗传算法 蚁群算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]