蚁群算法在交通信号实时控制中的研究及应用
|
摘要
城市交通拥堵是当今社会普遍关注的热点问题,尤其是道路交叉路口作为城市交通的重要环节,是产生交通拥堵的主要地点。众所周知车辆在交叉口上的平均延误时间太长,所以对其交通流进行合理分析和控制,是缓解交通拥堵的重要手段和途径,也是研究城市交通问题的一个关键切入点。城市道路交叉口作为城市路网中的重要组成部分,它的通畅与否直接影响到整个城市交通网络能否安全、有效的运行。因此本文以此为出发点,开展了对城市交通信号控制的研究,重点研究了用蚁群算法对交通信号进行优化的实时配时问题。
本文针对目前城市交通问题的现状,首先分析和讨论了交通信号控制的基本理论,对交通信号控制的发展过程、主要参数和控制类型进行了分析。其次,重点研究了城市交通信号实时控制模型的建立,分析了现有实时控制模型的参数,并针对其不足,建立了以车辆延时最少、平均停车次数最少和交通路口通行能力最大为目标函数的单个交叉口的点控制优化模型。之后,在具体分析了相位差、周期、绿信比和排队容量等交通信号控制参数的基础上,建立了使两个路口周期相等,以车辆延时最少、交通路口通行能力最大为目标函数的两个交叉口的线控制优化模型。最后,用泊松分布进行随机模拟生成车流量数据,使用蚁群算法对点控制和线控制模型中的控制变量进行了优化。蚁群算法是优化领域中新出现的一种仿生进化算法,它具有随机搜索、信息正反馈、分布式计算和全局优化等优点。文中我们在Matlab环境中对车流量进行了10个周期的仿真实验,并设计相应的蚁群算法,给出了基于MATLAB 7.0的具体实现,得出了优化后的目标函数值和优化后的各相位绿灯控制时间。仿真试验表明,所得结果优于经典方法,降低了交叉口的总延误时间和停车次数,提高了通行能力。
此外,本文提出了在单个交叉口的点控制和两个交叉口的线控制研究的基础上,可以扩展到在五个交叉口的面控制方面进行探讨。并从道路交通网状结构的假设和缓冲区的设置两个方面提出了自己的看法。所以说本文提出的方法对于我国城市交通信号实时控制的研究和建设具有很好的参考和实用价值。
Urban traffic jams is a hot problem commonly concerned around the society. As the important part of urban traffic, intersections are the distributing center that caused jams. We know that the average stop delay of vehicle at the intersections is too long. So rationally analyzing and controlling the traffic flow in the intersection is not only an important means and ways to relieve the traffic jam but also a key respect to study urban traffic problem. As the important part of urban traffic network, the arterial traffic affects the whole urban traffic. In this paper, urban traffic signal control is studied, and application of ant colony algorithm in traffic signal real-time control is mainly studied.
View of the current status of the urban traffic problems, firstly, the basic theory of traffic signal control is discussed. And the development process, the main parameters and control types of traffic signal control is studied. Then urban traffic signal real-time control model is mainly studied. In summing up the parameters and shortage of exist real-time control model, a single intersection optimization control model is presented, which is to minimize vehicle time delay and average number of stops and maximize traffic capacity. Besides, the traffic signal control parameters including offset, cycle length, split and queue capacity is analyzed. Two intersections line control optimization model is presented, which is to minimize vehicle time delay and maximize traffic capacity. In the model the cycle length of two intersections is equal. Finally, ant colony algorithm is bionic evolutionary algorithm in optimization of the field, which has the characters of random search, information on positive feedback, distributed computing and global optimization etc. Through poisson distribution generating flow data, a simulation is conducted in a single intersection optimization control mode and line control optimization model by using ant colony algorithm in MATLAB, which run for 10 cycles in normal traffic condition. The optimized objective function value and green time of signal phase is presented. Simulation experiments show that the results can be better than that of classical methods, and can reduce total delay and stops and improve traffic capacity at intersections.
Moreover, based on the research of a single intersection optimization control mode and line control optimization model, so five intersections network control model in the paper can be discussed. And I have some ideas on the assumptions of traffic network and the buffer zone set up. The research findings of the dissertation contribute to a further study into the urban traffic signal real-time control which fits the real traffic situation in China.
本文针对目前城市交通问题的现状,首先分析和讨论了交通信号控制的基本理论,对交通信号控制的发展过程、主要参数和控制类型进行了分析。其次,重点研究了城市交通信号实时控制模型的建立,分析了现有实时控制模型的参数,并针对其不足,建立了以车辆延时最少、平均停车次数最少和交通路口通行能力最大为目标函数的单个交叉口的点控制优化模型。之后,在具体分析了相位差、周期、绿信比和排队容量等交通信号控制参数的基础上,建立了使两个路口周期相等,以车辆延时最少、交通路口通行能力最大为目标函数的两个交叉口的线控制优化模型。最后,用泊松分布进行随机模拟生成车流量数据,使用蚁群算法对点控制和线控制模型中的控制变量进行了优化。蚁群算法是优化领域中新出现的一种仿生进化算法,它具有随机搜索、信息正反馈、分布式计算和全局优化等优点。文中我们在Matlab环境中对车流量进行了10个周期的仿真实验,并设计相应的蚁群算法,给出了基于MATLAB 7.0的具体实现,得出了优化后的目标函数值和优化后的各相位绿灯控制时间。仿真试验表明,所得结果优于经典方法,降低了交叉口的总延误时间和停车次数,提高了通行能力。
此外,本文提出了在单个交叉口的点控制和两个交叉口的线控制研究的基础上,可以扩展到在五个交叉口的面控制方面进行探讨。并从道路交通网状结构的假设和缓冲区的设置两个方面提出了自己的看法。所以说本文提出的方法对于我国城市交通信号实时控制的研究和建设具有很好的参考和实用价值。
Urban traffic jams is a hot problem commonly concerned around the society. As the important part of urban traffic, intersections are the distributing center that caused jams. We know that the average stop delay of vehicle at the intersections is too long. So rationally analyzing and controlling the traffic flow in the intersection is not only an important means and ways to relieve the traffic jam but also a key respect to study urban traffic problem. As the important part of urban traffic network, the arterial traffic affects the whole urban traffic. In this paper, urban traffic signal control is studied, and application of ant colony algorithm in traffic signal real-time control is mainly studied.
View of the current status of the urban traffic problems, firstly, the basic theory of traffic signal control is discussed. And the development process, the main parameters and control types of traffic signal control is studied. Then urban traffic signal real-time control model is mainly studied. In summing up the parameters and shortage of exist real-time control model, a single intersection optimization control model is presented, which is to minimize vehicle time delay and average number of stops and maximize traffic capacity. Besides, the traffic signal control parameters including offset, cycle length, split and queue capacity is analyzed. Two intersections line control optimization model is presented, which is to minimize vehicle time delay and maximize traffic capacity. In the model the cycle length of two intersections is equal. Finally, ant colony algorithm is bionic evolutionary algorithm in optimization of the field, which has the characters of random search, information on positive feedback, distributed computing and global optimization etc. Through poisson distribution generating flow data, a simulation is conducted in a single intersection optimization control mode and line control optimization model by using ant colony algorithm in MATLAB, which run for 10 cycles in normal traffic condition. The optimized objective function value and green time of signal phase is presented. Simulation experiments show that the results can be better than that of classical methods, and can reduce total delay and stops and improve traffic capacity at intersections.
Moreover, based on the research of a single intersection optimization control mode and line control optimization model, so five intersections network control model in the paper can be discussed. And I have some ideas on the assumptions of traffic network and the buffer zone set up. The research findings of the dissertation contribute to a further study into the urban traffic signal real-time control which fits the real traffic situation in China.
引文
[1]徐吉万,徐冬玲.城市交通的计算机控制和管理[M].北京:测绘出版社,1988
[2]沈建武.城市交通分析与道路设计.武汉:武汉大学出版社,2001
[3]薛珊荣,徐吉谦,陈敏尧.城市交通工程与街道规划设计[M].北京:中国建筑工业出版社, 1988
[4]美国运输部联邦公路局.交通控制系统手册[M].北京:人民交通出版社,1987
[5]杨锦东,杨东援.城市信号控制交叉口信号周期时长优化模型[J].同济大学学报,2001,29(7):789～794
[6]李建新,毛保华.混合交通流环境下有信号平面交叉口通行能力研究[J].交通运输系统工程与信息,2001,1(2):119～123
[7] WEBSTER.F.V. Traffic Signal Settings [J]. Road Research Laboratory Technical Paper,1958,39(1):1~39
[8] Ma J M, Zhang H T, Ren F T, Rong J. Study on Optimizing Traffic Operation of Signalized Intersection[J]. Journal of Highway And Transportation Research And Development, 2001
[9] Mechant D K, Nemhauser G L. A Modeland Algorithm for Dynamic Traffic Assignment Problem [J]. Transpn. Sci, 12,1978
[10] Newell G.F. Comments on Traffic Dynamics. Transpn.Res.-B,1989
[11]杨佩昆,张树升.交通管理与控制.北京:人民交通出版社,1995
[12]吴晓华.城市平面交叉口交通信号的智能控制技术研究[硕士学位论文],武汉:武汉理工大学,2006
[13]臧利林.城市交通信号优化控制算法研究:[博士学位论文].山东:山东大学,2007
[14] Dorigo M, Gambardella L M. Ant Colony System: A Cooperative Learning Approach to the Traveling Salesman Problem. IEEE Trans on Evolutionary Computation, 1997,1 (1). 53~66
[15]闻育,吴铁军.基于蚁群算法的城域交通控制实时滚动优化.控制与决策,2004年9月
[16] Solnon.C. Ants can solve constraint satisfaction problems. IEEE Transaction on Evolutionary Computation,2002.6(4).345~358
[17] M. Randall. A Parallel Implementation of Ant Colony Optimization. Journal of Parallel and Distributed Computing, 2002,62. 1420~1440
[18] Nagel.K, Pieck.M, Simon.P.M, etc. Comparison between three different traffic Micro-simulations and reality in Dallas[R].Los Alamos Unclassified Report LA-UR 1998.92~196
[19]颜艳霞,李文权.单点交叉口信号实时配时模型及蚂蚁算法.公路交通科技,2006年11月
[20]刘瑛,李跃武.规划法优化单个交叉口配时[J].湖南交通科技,2002
[21]王炜,高海龙,李文权,等.公路交叉口通行能力分析方法[M].北京:科学出版社,2001
[22]庄焰,曾文佳.信号交叉口延误计算模型研究.深圳大学学报理工版,2006年10月: 309～313
[23]李建斌,高成修.城市道路网络多交叉路口交通信号实时优化控制模型与算法.系统工程,2004年10期
[24]马子斌,商学岭.道路交叉口处信号灯的黄灯时间和优化配时问题[J].数学的实践与认识,1995:8～14
[25]陈琳,刘翔,孙优贤.单交叉路口交通流的通用多相位智能控制策略.浙江大学学报2006年11期
[26]姜启源,谢金星,叶俊,等.数学模型(第三版).北京:高等教育出版社,2003
[27]徐勋倩,黄卫.单路口交通信号多相位实时控制模型及其算法.控制理论与应用, 2005年03期
[28]高成修,陈克兵.交叉口信号灯实时动态时间分配模型[J].武汉大学学报(理学版), 2003,49:297～300
[29]王殿海.交通流理论[M].北京:人民交通出版社,2002
[30]刘运通,石建军,熊辉.交通系统仿真技术[M].北京:人民交通出版社,2002
[31]商蕾,高孝洪,蒋汉平.城市微观交通模型建模及仿真实现[J].武汉理工大学学报(交通科学与工程版),2003,27(4):498～503
[32]马建明,荣建,任福田,等.信号交叉口微观仿真模型研究.计算机仿真,2001,18(4):58～60
[2]沈建武.城市交通分析与道路设计.武汉:武汉大学出版社,2001
[3]薛珊荣,徐吉谦,陈敏尧.城市交通工程与街道规划设计[M].北京:中国建筑工业出版社, 1988
[4]美国运输部联邦公路局.交通控制系统手册[M].北京:人民交通出版社,1987
[5]杨锦东,杨东援.城市信号控制交叉口信号周期时长优化模型[J].同济大学学报,2001,29(7):789～794
[6]李建新,毛保华.混合交通流环境下有信号平面交叉口通行能力研究[J].交通运输系统工程与信息,2001,1(2):119～123
[7] WEBSTER.F.V. Traffic Signal Settings [J]. Road Research Laboratory Technical Paper,1958,39(1):1~39
[8] Ma J M, Zhang H T, Ren F T, Rong J. Study on Optimizing Traffic Operation of Signalized Intersection[J]. Journal of Highway And Transportation Research And Development, 2001
[9] Mechant D K, Nemhauser G L. A Modeland Algorithm for Dynamic Traffic Assignment Problem [J]. Transpn. Sci, 12,1978
[10] Newell G.F. Comments on Traffic Dynamics. Transpn.Res.-B,1989
[11]杨佩昆,张树升.交通管理与控制.北京:人民交通出版社,1995
[12]吴晓华.城市平面交叉口交通信号的智能控制技术研究[硕士学位论文],武汉:武汉理工大学,2006
[13]臧利林.城市交通信号优化控制算法研究:[博士学位论文].山东:山东大学,2007
[14] Dorigo M, Gambardella L M. Ant Colony System: A Cooperative Learning Approach to the Traveling Salesman Problem. IEEE Trans on Evolutionary Computation, 1997,1 (1). 53~66
[15]闻育,吴铁军.基于蚁群算法的城域交通控制实时滚动优化.控制与决策,2004年9月
[16] Solnon.C. Ants can solve constraint satisfaction problems. IEEE Transaction on Evolutionary Computation,2002.6(4).345~358
[17] M. Randall. A Parallel Implementation of Ant Colony Optimization. Journal of Parallel and Distributed Computing, 2002,62. 1420~1440
[18] Nagel.K, Pieck.M, Simon.P.M, etc. Comparison between three different traffic Micro-simulations and reality in Dallas[R].Los Alamos Unclassified Report LA-UR 1998.92~196
[19]颜艳霞,李文权.单点交叉口信号实时配时模型及蚂蚁算法.公路交通科技,2006年11月
[20]刘瑛,李跃武.规划法优化单个交叉口配时[J].湖南交通科技,2002
[21]王炜,高海龙,李文权,等.公路交叉口通行能力分析方法[M].北京:科学出版社,2001
[22]庄焰,曾文佳.信号交叉口延误计算模型研究.深圳大学学报理工版,2006年10月: 309～313
[23]李建斌,高成修.城市道路网络多交叉路口交通信号实时优化控制模型与算法.系统工程,2004年10期
[24]马子斌,商学岭.道路交叉口处信号灯的黄灯时间和优化配时问题[J].数学的实践与认识,1995:8～14
[25]陈琳,刘翔,孙优贤.单交叉路口交通流的通用多相位智能控制策略.浙江大学学报2006年11期
[26]姜启源,谢金星,叶俊,等.数学模型(第三版).北京:高等教育出版社,2003
[27]徐勋倩,黄卫.单路口交通信号多相位实时控制模型及其算法.控制理论与应用, 2005年03期
[28]高成修,陈克兵.交叉口信号灯实时动态时间分配模型[J].武汉大学学报(理学版), 2003,49:297～300
[29]王殿海.交通流理论[M].北京:人民交通出版社,2002
[30]刘运通,石建军,熊辉.交通系统仿真技术[M].北京:人民交通出版社,2002
[31]商蕾,高孝洪,蒋汉平.城市微观交通模型建模及仿真实现[J].武汉理工大学学报(交通科学与工程版),2003,27(4):498～503
[32]马建明,荣建,任福田,等.信号交叉口微观仿真模型研究.计算机仿真,2001,18(4):58～60
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |