【作者】 陈林；

【导师】 杨兆升；

【作者基本信息】 吉林大学 ， 交通信息工程及控制， 2006， 硕士

【摘要】 本论文在当今计算机、通信及系统工程等技术的发展支持下,依据城市交通发展需求及交通控制系统的发展方向,主要研究了一种配时方案实时选择城市交通信号自适应控制系统。论文首先介绍了研究背景、必要性和文章的内容组织架构,通过阐述城市交通信号控制的分类、发展方向,指出自适应控制系统是今后一段时间的研究重点。系统结构、相位模型、交通数据模型及与管理目标、控制策略和信号配时参数优化算法是影响系统控制效益的几个关键因素,本文的研究重点是相位模型和信号配时参数优化算法。相位模型研究提出相位编码概念,便于交叉口微观分析,简化了相位执行指令,对叠加相位可行性进行分析,总结出相位设计规则经验参考,提出了一种公交优先相位设计方法,最后给出了相位生成目标指数分析,目的在于对众多可行相位相序择优选择。信号配时参数优化采用的算法可描述为“跟随交通变化趋势的小步距动态调整算法”,小步距调整避免了配时参数的抖动,连续的同方向调整可以反映变化的趋势,短时间内和趋势不一致的随即交通流不会导致配时参数的调整。通过本论文的研究为城市交通自适应控制的研究提供了一种可行的思路和方法,分布式实时配时参数选择控制系统具有信息处理速度快、系统资源消耗低、控制效益较高,控制规模大等优点,对研制我国自主产权的城市交通控制有一定的参考价值。更多还原

【Abstract】 Since 1990’s, the upsurge in ITS (Intelligent Transportation Systems) researchhas been raised all over the world. Advanced Traffic Signal Control System is avery important part of ITS. It is not only the important traffic flow informationcollecting subsystem, but also the core subsystem to adjust the traffic flowdistribution of the city. Under current road condition, it is a efficient and approachto relieve the degree of road congestion and enhance the traffic efficiency byconstruction urban traffic control system.This paper introduces the principle and the implementation of the distributedsignal plan real time selected adaptive Urban Traffic Control System (UTCS) indetail using the current computer, communication and system engineeringtechniques. Particular attention is concentrated on some important aspects ofUTCS such as Phase Model, System Architecture, traffic data model andmanagement object, Control Strategy and Signal Parameters OptimizingAlgorithm, which are associated with the system control efficiency.(1)Research of System Architecture Four different aspects of system architecture, control structure, system principle,device connection network topologic structure and software architecture areresearched in this paper. The system applies distributed system architecture. Thetypical 3-layer controlling structure supports the distribution of the system controlfunction;distributed network structure makes it possible to deploy the distributedhardware device;Components based software architecture enable it to setupcontrol software in different computers, thus it can well balanced the system loadand enhance the flexibility of system control capability. Components of the corecontrol software are integrated by using Microsoft COM technology.(2)Research of Phase ModelPhase model describes the method to define traffic signal phase. Signal phasedefines the classification and combination of the right to pass through the signalcontrol intersection. Phase model researched in this paper introduces the conceptof phase coding, which is convenient to analyze the intersection microcosmiccondition and simplifies the phase implementation command. This paper alsomakes a analysis on the lap-phase, that is aimed to utilize the time and spaceresource efficiently and use the least resource to satisfy the unbalanced traffic flow.From the aspect of traffic control system function, there is no any technologyproblem to realize the lap-phase, the difficulties are to identify the traffic flowstate, select the phase types and innovate. This paper also gives a method to designthe transit priority phase, summarizes the phase design experience reference.Evaluation method of phasing arrangement is presented to decide the goal phasearrangement. It is used to evaluate whether the phase design reaches the practicerequest. The method selects the utmost saturation degree as the evaluation index asto obtain the optimized phase arrangement.(3)Research of Control StrategyThe ordered combination of control mode and control plan is defined as controlstrategy in this paper. The system uses the combination of control strategy drivetime table and intersection control strategy drive time table. In accord with thechange of traffic trend, the appropriate and reasonable control strategy can bedesigned and carried into execution flexibly.(4)Research of Traffic Parameter Detection and Information ManagementTraffic parameters are not only the base of urban traffic signal control, but alsothe reference to realize the other functions of ATMS. It is needed to select thebasic traffic parameters and analyze the feasibility and rationality of datacollecting technology, insuring the reliable data source of the system, the systemso as to make the traffic signal control system practicable and transplantable. Thispaper makes a analysis on the traffic information characteristics and data model,presents the information management goal. To the process of informationmanagement, one need is to process the data, inform and form the informationefficiently;another is to handle the relationship between distribution andconcentration reasonably. The way is to store the highly frequently sharedinformation centrally and lowly one and other subsystems’ distributedly.(5)Research of Signal Plan Parameters Optimizing AlgorithmThe optimizing algorithm in this paper can achieve the cycle, split and offsetadjustment in real time. It can be described as “little pace dynamic adjustmentalgorithm following traffic changing trend” which avoids the fluctuation of timingparameters. continuous one way adjustment can reflect the traffic changing trend,short period accident traffic flow inconsistent with the trend won’t result in thetiming parameters adjustment. According to the need of the algorithm, this paperintroduces one data processing method called four times traffic data processingand method to handle wrong data simply. Layer control is the system’s keyfoundation. To achieve this function, this paper presents two particular aspectscalled cycle control zone and offset/split zone control zone. The control zone isdefined as the smallest unit that needs the same control mode or there won’t besome badness when using same control mode. In the cycle, split, offset threetiming parameters, cycle and split have the optimal value using the singleintersection, but the offset control must make more than two intersections ascontrol object as only certain intersections can manifest the control effect. Controlparameters optimizing algorithm selects control plan according to the control zonereal time traffic condition. Each control plan’s cycle, split, offset has differentcombination according to the different way traffic flow ratio of the keyintersections. The cycle of each control plan can search for the optimal value insmall pace according to the intersection’s traffic flow changing trend in order toadapt traffic need;split control is divided into two control layers as tactic andstrategy so as to meet the request of unitary optimization and partial characteristicswhich enhances the passing efficiency of the intersection more;offset embodiesthe coordinated control effect that firstly takes basic offset plan as referenceaccording to the up-down traffic flow ratio, then decides the relative offsetaccording to the length and average vehicle speed of road section;thirdly, modifiesthe absolute offset according to the executive cycle length of the intersection,decides the final adjustment value according to the least value principle lastly.Finally, we make a short summarization of the distributed adaptive traffic signalcontrol system and indicate the future development of the system.更多还原