【作者】 李捷；

【导师】 王伟智；

【作者基本信息】 福州大学 ， 系统工程， 2006， 硕士

【摘要】 基于对实时交通系统的分析和对前人开发的微观仿真软件的研究,本文开发了一套能够仿真智能交通信号灯控制策略的城市交通微观仿真软件UTMS(Urban Traffic Micro-Simulation),整个软件的设计过程按照软件工程的指导思想进行,采用面向对象的的方法进行建模,开发平台选择微软的Visual C++6.0。软件通过对交通路网的建模,使用户通过交互式输入能够再现交通主要路网模型,为车辆的行驶提供仿真背景,同时对交通活动的重要参与者的运动进行建模,考察及选择国外的主要模型,并在实际的开发过程中改进和提出自己的有效仿真模型,为较真实的反映交通流的运动规律提供了保证,整个仿真过程车辆的运动通过可视化界面采用2D动画的形式实时输出,使用户能够十分直观地观测到整个仿真过程车流的运动情况。为了模拟实际交通参数的检测手段,软件通过建立虚拟检测线圈,模拟现实中的检测线圈的工作原理,检测仿真环境中的交通状况,从而为用户开发设计智能交通信号控制策略提供交通流数据。为了灵活的支持用户自定的智能信号灯控制策略,UTMS提出一种采用动态链接库的方式对外部用户自定交通信号控制策略进行仿真的方法,同时在软件中通过建立外部抽象接口提供一整套的API函数,提供给外部包括检测器数据,交通信号灯数据,道路拓扑数据等实时数据以及进行数据库操作,这样就能为用户开发自己的智能交通控制策略动态链接库插件提供仿真中实时交通检测数据,并使动态链接库插件在主程序中实现来获得交通信号灯的实时显示状态,从而指挥路口车流的通行。同时仿真结束后能够通过可视化界面输出系统对智能交通控制策略的分析和评价,这样就能使用户通过编写动态链接库插件来验证许多较复杂的控制策略如模糊控制、神经网络控制等智能控制在交通信号控制系统中的作用,并能够很直观的比较各种控制策略的有效性。文章在最后开发了基于类饱和度的模糊单路口交通信号灯控制、基于排队长度的模糊单路口交通信号灯控制、基于排队长度的模糊神经网络单路口交通信号灯控制和多路口的递阶协调交通信号灯控制等多个复杂的智能交通控制策略动态连接库插件进行实例仿真,通过各种方案仿真比较,结果评价分析,证明智能控制策略在交通信号灯控制系统中的有效性,同时也说明了本文开发的以仿真智能交通信号控制策略为核心的城市微观交通仿真系统实用性。更多还原

【Abstract】 Based on the analysis of the real traffic system and the research on the former simulation software,The thesis introduces the development of a urban traffic micro-simulation sofeware of intelligent traffic signal control system.The development of the sofeware is guided by the thinking of Software Engineering,and uses Object-oriented Technology.The tool of development is Visual C++. Many functions of the sofeware are presented in the thesis.Frist, the sofeware makes modeling from the true road network and provides users to employ interactive interfaces to enter data and to represent the condition of ubran road network. Second,based on the foreign model of traffic participants,the sofeware makes more available modeling of traffic participants so that it can factually reflect the law of action of traffic flow.And the movement of traffic participants is exported in real time on visual interface through the 2D cartoon during whole process of simulation. In order to imitate the detection means of the real traffic parameter, the software uses the vitual coil whose principle of the measuring is as same as the real coil to measure the traffic parameters in the artificial environment, thus users can develop intellectual traffic signal control strategy by these traffic parameters.Third,to flexiblely support users’ intellectual traffic signal control strategys, UTMS puts forward a new way in which users can develop their Dynamic Link Librarys of intellectual traffic signal control strategy.And UTMS set up a number of API functions through establishing the abstract interface outside which can provide detection data , the traffic signal lamp data, the topological data of the road and can make users operate the database,so that uses can employ the real-time date to develop their intellectual traffic signal control strategys.UTMS transfers the Dynamic Link Library plug-in to obtain real-time condition of traffic signal in real-time during the process of simulation which can direct the traffic participants at the crossing.And UTMS can output the analysis and appraisal of the control strategy of intellectual traffic signal through the visual interface after simulation,thus uses can verify a lot of complicated control strategys such as fuzzy control, neural network control and can find what the role these control straegys play in intelligent traffic signal control system,and can directly compare which is more available. Finally the author develops many complicated control strategys included fuzzy traffic signal control based saturation,fuzzy traffic signalcontrol based the length of queue, fuzzy neural network control of traffic signal and the hierarchical coordinated control of the traffic artery.Comparing and analyzing the result of these stategys,we can find that it is efficient to use intelligent traffic signal control stategy in traffic signal control system.And it proves that urban traffic micro-simulation sofeware of intelligent traffic signal control system is good at simulating.更多还原