一种求解带约束多式联运问题的群智能算法  

A Novel Swarm Intelligence Optimization Algorithm for Solving Constrained Multimodal Transportation Planning

作　　者：梁晓磊[1] 李文锋[1] 张煜[1] 

机构地区：[1]武汉理工大学物流工程学院,武汉430063

出　　处：《上海交通大学学报》2015年第8期1220-1230,共11页Journal of Shanghai Jiaotong University

基　　金：湖北省国际合作项目(2011BFA012);国家自然科学基金项目(71372202);十二五科技支撑计划项目(2014BAH24F03);湖北省自然科学基金项目(2014CFB875)资助

摘　　要：针对如何有效运用群智能算法求解多式联运问题,设计了一种针对群智能优化算法的个体解码方式,提出了一个有效的个体编码与多式联运方案的映射模型.在该映射模型中设计了基于比例的流量分配方式,实现了个体编码信息向初步流量分配方式的解码;同时构建了局部流量调整策略,进行不可行方案修复,提高了解码方案的有效性.而后,提出了一种变邻域粒子群算法,将社会网络演化特征引入进行粒子群算法的种群拓扑和邻域调整,以改善个体在搜索过程中的交互模式.基于解码策略,采用改进算法对多式联运问题进行求解,并与3种新型群智能算法进行对比.通过实例分析,该编码策略可以有效应用于多式联运问题求解.同时,变邻域粒子群优化算法的收敛效率和性能优于对比算法.In order to apply the swarm intelligence（SI）algorithms effectively to solve the multimodal transportation planning problem,this paper proposed a decoding strategy to build a mapping modal between individual representation of SI and multimodal transportation schedule.In the modal,a method for traffic assignment in a multimodal transportation network was provided to decode each individual to an initial schedule.Then,a strategy for local traffic adjustment was applied to mend these initial schedules to improve the success rate of decoding.A developed particle swarm optimization（PSO）algorithm was also proposed to solve the planning problem compared with three other state-of-art swarm intelligence optimization algorithms.A novel way that applies social network evolution behavior to adjust the swarm topology and individuals＇ neighborhood and promote the interaction modal among individuals was introduced in the proposed algorithm.The numerical test of an operational problem shows that the decoding strategy is efficient in solving the multimodal transportation planning problem and the proposed algorithm has a superior performance on the terms of convergence speed and solution accuracy in comparison with the selected algorithms.

关 键 词：解码策略 群智能 粒子群 多式联运 

分 类 号：U169.62[交通运输工程]