基于改进遗传算法的可变信息标志选址优化研究  

Optimization for Location of Variable Message Sign Based on an Improved Genetic Algorithm

作　　者：张荣辉 马壮林[1] 党永乐 徐婷[1] ZHANG Ronghui;MA Zhuanglin;DANG Yongle;XU Ting(School of Automobile,Chang′an University,Xi′an 710064,China)

机构地区：[1]长安大学汽车学院,西安710064

出　　处：《交通信息与安全》2018年第6期113-122,共10页Journal of Transport Information and Safety

基　　金：国家自然科学基金项目(51208052;51878066);国家重大科技研发计划项目((2018YFC0807505);陕西省自然科学基础研究计划项目(2017JM5084);教育部人文社会科学研究项目(18YJCZH130);中央高校基本科研业务费专项资金项目(300102228202)资助

摘　　要：随着城市化进程的不断加快,交通拥堵已成为亟待解决的社会问题,实践表明可变信息标志(VMS)布设位置对于诱导拥堵路段上的交通流量、缓解交通拥堵、提高整个交通诱导系统的效率至关重要。在考虑道路交通流量、交通拥堵指数以及可变信息标志发布信息的衰减程度对VMS选址影响的基础上,考虑可替代路径的诱导比率、绕行时间以及布设VMS所需的成本对VMS选址的影响,然后综合VMS的效用和总成本2个因素,采用最大效用与最小成本之比的形式构建了VMS选址优化的双目标规划模型,分别采用传统的遗传算法和改进的遗传算法对构建的模型进行求解,最后以西安市南二环路为例进行实例分析。结果表明,当设置4块VMS时效果最好,最优布设路段为10,13,15,17。与传统的遗传算法相比,改进的遗传算法陷入局部收敛的次数由21次减小到6次,且收敛速度更快;模型的最优解与现状VMS的布设情况相比函数适应度值更大,表明其布设效果更好,验证了模型的有效性。Traffic congestion has become an urgent problem to be solved with acceleration of urbanization process.Actual data shows that location of variable message sign(VMS)is crucial for inducing traffic flows on congested road segments,reducing traffic congestion and improving efficiency of entire system of traffic induction.Based on road traffic flow,traffic congestion index,and impacts of information attenuation degree on location of VMS,a bi-objective programming model is developed.Induction rate of alternative paths,bypass time,and cost of deploy VMS are taken into consideration.Combining utility and the total cost of set VMS,the model shows a ratio of maximum utility to minimum cost.Traditional genetic algorithm and improved genetic algorithm are used to find solutions.The road segment of Nan′er Huan Road in Xi′an is taken as a case study.The results show that when set four VMS,the outcome is best,and the optimal location of VMS is section 10,13,15,and 17.Compared with traditional genetic algorithm,the improved genetic algorithm has faster speed of convergence,and number of local convergence reduces from 21 to 6.The optimal solution in the developed model has a larger value of function fitness than actual layout,which indicates better location of VMS.Effectiveness of the model is valid.

关 键 词：交通工程 可变信息标志选址 可替代路径 绕行时间 双目标规划模型 遗传算法 动态变异概率 

分 类 号：U4-9[交通运输工程—道路与铁道工程]