考虑时变风险的成品油二次配送路径优化  

Route optimizing on secondary distribution of refined oil considering time-dependent risk

作　　者：韦金银 郭琪 史彬 鄢烈祥 WEI Jinyin;GUO Qi;SHI Bin;YAN Liexiang(School of Chemistry,Chemical Engineering and Life Sciences,Wuhan University of Technology,Wuhan 430070,Hubei,China)

机构地区：[1]武汉理工大学化学化工与生命科学学院,湖北武汉430070

出　　处：《化工进展》2020年第4期1597-1604,共8页Chemical Industry and Engineering Progress

基　　金：国家自然科学基金(21878238)。

摘　　要：成品油属于易燃易爆炸危险品,其配送过程一旦发生事故后果严重。考虑到配送过程人群聚集程度、载重和车速的动态变化,本文建立时变风险度量模型,运用BP神经网络预测车速,预测了城市高速路段和主干路路段的车速。在此基础上,构建了同时考虑配送成本与时变风险的成品油二次配送双目标路径优化模型,并提出改进的多目标列队竞争算法进行求解。本文还采用改进的多目标列队竞争算法求解文献报道的实例,求得非支配解完全支配文献报道中的非支配解,验证了算法的有效性。某武汉石化企业实例研究表明,预测的车速具有较高的准确度,所提出的方法能同步考虑配送过程的时变风险和配送成本,求得的Pareto最优配送路径集可以为企业决策者提供多样化的选择方案,具有较好的应用价值。Refined oil belongs to inflammable and explosive dangerous products.Once an accident occurs in the distribution process,the consequences are serious.Considering the dynamic changes of population aggregation,load and speed in distribution process,a time-dependent risk measurement model was established.The BP neural network was used to predict vehicle speed,and the speed of expressway and main road in Wuhan was predicted.On this basis,a bi-objective route optimization model of refined oil secondary distribution route considering both distribution cost and time-dependent risk was proposed,and an improved multi-objective line-up competition algorithm was proposed to solve the problem.The improved multi-objective line-up competition algorithm was used to solve the examples reported in the literature,and the obtained non-dominated solutions completely dominated the non-dominated solutions in the literature reports,which verified the effectiveness of the algorithm.A case study of Wuhan petrochemical enterprise showed that the predicted speed had a high accuracy,the proposed method could simultaneously consider the time-dependent risks and distribution costs in the distribution process,and the obtained Pareto optimal distribution routing set could provide diversified options for enterprise decision-makers,and had good application value.

关 键 词：时变风险 成品油配送 预测 优化 算法 

分 类 号：TQ015.9[化学工程]