摘 要

我国大部分城市城市交通道路拥堵现象普遍存在，为解决这一问题，减轻道路压力，一方面，政府可以采取增加道路资源建设、改善交通基本设施、道路结构优化等措施，但往往需要投入巨大的成本；另一方面，可通过智能化的手段来对交通进行协调控制，提高道路出行效率，例如以新型智能交通信号灯为核心的智能交通控制系统正在不断成熟、推广，但智能信号灯是否能够时刻保持最优效率，智能信号灯是否与道路中存在的车辆数存在某种关系引人深思。本文基于交通智能信号灯数学模型，采用Netlogo仿真软件进行相应的代码编写，在定量记录智能系统下车辆平均等待时间与数量的基础上，对比传统定时分配交通信号灯和智能信号灯的效率，同时进一步探究在智能信号灯下，道路车辆数是否存在某一阈值，使智能信号灯效率开始下降。交通系统的主体为“人”，而人是能自主控制，自主适应的智能体，为了最优化道路资源，使各个智能体能彼此协调，本文同时利用博弈论理论来对相关联的交通路口进行控制，在整体上提高车辆的行驶速率，减少道路资源的浪费。

关键词 多智能体 智能信号灯 Netlogo 博弈论

Traffic Lights Simulation Research Based on Multi-agent

ABSTRACT

In order to solve this problem and alleviate road pressure, the government can take measures to increase road resource construction, improve traffic infrastructure and optimize road structure, but it often needs to invest huge costs. On the other hand, it can coordinate traffic control and improve road structure through intelligent means. Road travel efficiency, such as the intelligent traffic control system with the new intelligent traffic lights as the core, is constantly maturing and popularizing. But whether the intelligent traffic lights can keep the optimal efficiency at all times and whether there is a certain relationship between the intelligent traffic lights and the number of vehicles in the road is thought-provoking. Based on the mathematical model of traffic intelligent signal lamp, the corresponding code is compiled with Netlogo simulation software. On the basis of quantitatively recording the average waiting time and number of vehicles in intelligent system, the efficiency of traditional timing traffic signal lamp and intelligent signal lamp is compared. At the same time, whether there is a threshold for the number of road vehicles under intelligent signal lamp is further explored to make intelligent signal. The efficiency of the lights began to decline. The main body of traffic system is "human", and human is an agent that can control and adapt independently. In order to optimize road resources and make each agent coordinate with each other, in order to reduce the waste of road resources and improve the speed of vehicles, this paper uses Game Theory to control the related traffic intersections.

Key words: multi-agent; Intelligent Signal Lamp; Netlogo; Game Theory

目录

摘要 I

ABSTRACT II

第一章 导论 1

1.1引言 1

1.2 论文研究的目的和意义 2

1.3 国内外对基于多智能体的交通系统的研究综述 2

第二章 基于多智能体的智能信号灯控制仿真研究探讨 6

2.1多智能体和Netlogo 6

2.2基于多智能体的智能信号灯设计 10

2.3 基于多智能体的Netlogo代码编写和运行原理 11

2.4 NetLogo的仿真结果分析 14

第三章 基于多智能体的信号灯控制仿真系统应用案例 21

3.1 应用场地介绍—南京新街口的某十字路口 21

3.2 某十字路口实测交通数据 21

3.3实测数据和仿真结果分析 23

3.4 NetLogo仿真探究影响智能信号灯的阈值 28

第四章 多智能体的协调机制 32

4.1 博弈论 32

4.2 多十字路口的协调方法步骤 33

第五章 总结与展望 35

参考文献 36

致谢 39

第一章 导论

1.1引言

多智能体系统，顾名思义，着重于“系统”，因为各个智能体的集合构成了整体，而它不单单只是各个分散，毫无关联的智能体的组合，正是由于它们之间紧密联系，能彼此交互通信、协调控制、冲突消解才使之称为“系统”—系统的效益大于各个效益之和，正所谓众人拾柴火焰高。交通系统中行驶的车辆涉及人类活动，也是多智能体的研究对象之一，它们不仅分布范围广、各个主体间能相互协调，而且还拥有自我组织、学习的能力。