基于非支配排序遗传算法的沥青路面大中修养护决策  

Decision-making in Asphalt Pavement Repair Maintenance Program Using a Non-dominated Sorting Genetic Algorithm

作　　者：马士宾 徐梓菲 刘月钊 宋玮卓 MA Shi-bin;XU Zi-fei;LIU Yue-zhao;SONG Wei-zhuo(School of Civil Engineering and Transportation,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学土木与交通学院,天津300401

出　　处：《科学技术与工程》2023年第34期14815-14823,共9页Science Technology and Engineering

基　　金：天津市交通运输科技发展项目(2022-01)。

摘　　要：为了更好地协调沥青路面大中修养护方案决策时成本、环境和性能之间相互制衡的矛盾,提出了一种基于非支配排序的遗传算法(non-dominated sorting genetic algorithmⅢ,NSGA-Ⅲ)来优化沥青路面大中修养护方案决策。针对实际公路工程项目中多重约束条件下的养护需求,建立了以碳排放量、寿命周期成本及路面使用性能为目标的多目标优化模型,采用NSGA-Ⅲ算法求解模型,有效解决各目标间的非支配关系,并得到养护方案的Pareto解集;在多属性决策中引入广义马氏距离与组合赋权方法,对Pareto解集进行排序决策,并以承德市某公路为例进行实例演示;划分3种不同的权重等级和六种权重策略,分析对不同目标赋予不同权重等级时对最优养护方案的影响。研究结果表明:NSGA-Ⅲ算法可以快速、高效地捕捉多目标解集,当对不同目标赋予相同的权重等级时,相比于只考虑单目标与双目标的养护方案,其最优解在碳排放量上分别降低24.45%、28.41%;寿命周期成本分别降低14.43%、17.95%;PCI值分别降低1.07%、3.68%。该方法相比于传统的决策方法可以更准确、更迅速地得到可持续性的养护方案。In order to better reconcile the contradiction between cost,environment and performance in the decision making of asphalt pavement medium and heave maintenance scheme,a non-dominated sorting genetic algorithmⅢ(NSGA-Ⅲ)was proposed to optimize the decision making of asphalt pavement medium and heave maintenance scheme.The NSGA-Ⅲalgorithm was used to construct a linear hyperplane and set reference points to effectively solve the non-dominated relationship among the objectives and obtain the Pareto solution set of the maintenance scheme.Generalized mahalanobis distance and combination weights were introduced in the multi-attribute decision-making method to sort the optimal solutions,which were used to rank and decide the Pareto optimal solution set.A road in Chengde was used as an example to demonstrate the optimization example.Three different weight classes and six weighting strategies were delineated to analyze the effects on the optimal maintenance solution and the optimal solution when different weight classes were assigned to different objectives.Compared to the conservation scenarios that only consider single-objective and dual-objective,when different objectives are given the same weighting level,the optimal solutions reduce carbon emissions by 24.45%and 28.41%,life-cycle costs by 14.43%and 17.95%,and PCI values by 1.07%and 3.68%,respectively.The method can obtain sustainable conservation solutions more accurately and quickly than traditional decision-making methods.

关 键 词：道路工程 沥青路面大中修 养护方案决策 多目标优化模型 NSGA-Ⅲ算法 

分 类 号：U418.6[交通运输工程—道路与铁道工程]