一种基于改进蚁群算法的垃圾车辆低碳收运路径优化方法  

An Improved Ant Colony Algorithm Based Low-carbon Collection Path Optimization Method for Waste Vehicles

作　　者：李婷婷 邓社军[1] 陆曹烨 王勇 廖华军[4] LI Ting-ting;DENG She-jun;LU Cao-ye;WANG Yong;LIAO Hua-jun(School of Architectural Science and Engineering,Yangzhou University,Yangzhou Jiangsu 225009,China;Nantong Branch of China Design Group,Nantong Jiangsu 226000,China;School of Economics and Management,Chongqing Jiaotong University,Chongqing 400074,China;Beijing Chaotu Software Co.,Ltd.,Beijing 100015,China)

机构地区：[1]扬州大学建筑科学与工程学院,江苏扬州225009 [2]华设设计集团股份有限公司南通分公司,江苏南通226000 [3]重庆交通大学经济与管理学院,重庆400074 [4]北京超图软件股份有限公司,北京100015

出　　处：《公路交通科技》2023年第5期221-227,246,共8页Journal of Highway and Transportation Research and Development

基　　金：国家自然科学基金项目(71871035);扬州市科技计划项目(YZ2021169)。

摘　　要：为实现城市生活垃圾的高效收集和低碳运输,加快实现城市交通运输体系“双碳”目标,基于车辆路径问题的建模求解方法,进行了具有车辆容量限制及时间窗约束的垃圾车辆低碳收运路径优化研究。为合理利用现有资源,根据垃圾时空分布规律,采用K-means聚类算法对城市垃圾收集的关键节点进行了分区重构。为准确测算车辆收运过程中的碳排放量,引入碳排放系数与碳税参数,采用“燃油-碳排”转化法来衡量企业的碳成本,构建了以车辆运输成本、固定成本和碳排放成本之和最小化为目标的车辆调度与路径优化模型。为加快算法的收敛速度并提高求解质量,提出了改进蚁群算法,首先改进了蚁群算法的状态转移方式和信息素更新方式,引入最大最小蚁群算法与2-opt局部搜索算法,进行了早熟收敛性判断。最后基于浙江省湖州某地垃圾收运实例数据,与遗传算法和粒子群算法进行了对比分析,验证了所提模型和算法的有效性。结果表明:与遗传算法和粒子群算法相比,蚁群算法具有较强的鲁棒性与并行性;与实际收运方案相比,基于分区的改进蚁群算法所得到的收运方案总成本可减少35.7%,其中碳排放成本可减少25.3%。研究成果可为实际收运过程中的车辆路径选择提供参考,提高收运效率,实现节能减排的目的。In order to realize the efficient collection and low-carbon transport of municipal garbage and accelerate the realize the“dual-carbon”goal for urban transport system,based on the modeling and solving method of vehicle routing problem,the low-carbon collection and transport path optimization of garbage trucks with capacity constraint and time window constraint is proposed.In order to make rational use of the existing resources,the key nodes of urban garbage collection are partitioned and reconstructed by K-means clustering algorithm according to the spatio-temporal distribution of garbage.In order to accurately calculate the carbon emissions during vehicle collection and transport,by introducing carbon emission coefficient and carbon tax parameter,by using the“fuel+carbon emission”conversion method to measure the carbon cost of enterprise,the vehicle scheduling and path optimization model with the goal of minimizing the sum of vehicle transport costs,fixed costs,and carbon emission costs is constructed.In order to accelerate the convergence speed of the algorithm and improve the solution quality,an improved ant colony algorithm is proposed.First,the state transfer and pheromone updating of ant colony algorithm are improved,and the premature convergence is judged by introducing the max-min ant colony algorithm and the 2-opt local search algorithm.Finally,based on the example data of garbage collection and transport in a certain area of Huzhou City,Zhejiang Province,the comparative analysis with genetic algorithm and particle swarm optimization algorithm is conducted,and the effectiveness of the proposed model and algorithm is verified.The result shows that(1)compared with genetic algorithm and particle swarm optimization,ant colony algorithm has strong robustness and parallelism;(2)compared with the actual collection and transport scheme,the improved ant colony algorithm based on partitioning can reduce the total cost of the scheme by 35.7%with a carbon emission cost reduction of 25.3%.The research result c

关 键 词：城市交通 车辆路径优化 改进蚁群算法 垃圾收运 绿色低碳 

分 类 号：U492.22[交通运输工程—交通运输规划与管理]