引用本文:范厚明,刘浩,刘鹏程,任晓雪.集货需求模糊的异型车同时配集货路径优化[J].控制理论与应用,2021,38(5):661~675.[点击复制]
FAN Hou-ming,LIU Hao,LIU Peng-cheng,REN Xiao-xue.Heterogeneous fleet vehicle routing problem with simultaneous deterministic delivery and fuzzy pickup[J].Control Theory and Technology,2021,38(5):661~675.[点击复制]
集货需求模糊的异型车同时配集货路径优化
Heterogeneous fleet vehicle routing problem with simultaneous deterministic delivery and fuzzy pickup
摘要点击 1506  全文点击 504  投稿时间：2020-06-22  修订日期：2020-11-29
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2020.00376
  2021,38(5):661-675
中文关键词  车辆路径问题  模糊需求  异型车辆  同时配集货  遗传变邻域算法
英文关键词  vehicle routing problem  fuzzy demand  heterogeneous fleet vehicle  simultaneous delivery and pickup  genetic variable neighborhood algorithm
基金项目  国家社科基金应急管理体系建设研究专项项目(20VYJ024)资助.
作者单位E-mail
范厚明* 大连海事大学交通运输工程学院 fhm468@163.com 
刘浩 大连海事大学交通运输工程学院  
刘鹏程 大连海事大学交通运输工程学院  
任晓雪 大连海事大学交通运输工程学院  
中文摘要
      针对集货需求模糊的异型车同时配集货车辆路径问题(HFVRPSDDFP), 基于先预优化再重优化的思路构 建模型. 预优化阶段根据可信度理论和车型选取方法为客户点分配车辆, 生成配送方案. 重优化阶段利用随机模拟 算法(SSA)确定客户集货需求, 对服务失败的客户点, 制定服务策略, 将模糊问题转化为确定型的异型车辆路径问 题(HFVRP), 并规划路径. 设计遗传变邻域算法, 通过测试确定邻域结构构造, 将自适应搜索策略应用到邻域搜索过 程中, 保证迭代前期收敛速度和后期全局搜索能力. 通过算例验证了本文模型及算法的有效性.
英文摘要
      The heterogeneous fleet vehicle routing problem with simultaneous deterministic delivery and fuzzy pickup (HFVRPSDDFP) is solved according to the idea of pre-optimization and re-dispatch in this paper. In the pre-optimization stage, vehicles are allocated to customers based on the credibility theory and the rule of vehicle type selecting, and the distribution scheme in the pre-optimization stage is generated. In the re-optimization stage, the stochastic simulation algorithm (SSA) is used to determine the customers’ pickup demand and the service strategy is formulated for the customer points that have failed to serve. The fuzzy problem is transformed into a deterministic heterogeneous fleet vehicle routing problem (HFVRP), and the route is re-planned. According to the characteristics of the problem, the genetic variable neighborhood algorithm is proposed. The neighborhood design is determined by repeated tests. The adaptive neighborhood search strategy is applied to the variable neighborhood search process in order to ensure the convergence speed in the early iteration and the global search ability in the later iteration. The effectiveness of the model and algorithm in this paper is verified by instances.