基于流量预测的交通信号控制技术研究

发布时间：2018-03-01 12:14

本文关键词： 智能交通流量采集模糊神经网络流量预测信号控制　出处：《浙江工业大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着社会经济的高速发展,汽车的普及率也在不断地提高。汽车数量的日益增多,导致城市道路资源变得十分匮乏,道路拥堵正成为近年来各大中城市面临的最大挑战之一。因此,提出一种高效的交通流量预测模型,并对路口信号灯做出智能控制,以实现道路资源利用率的最大化和高效化,对于缓解交通拥堵问题将起到十分重要的作用。本文设计了单交叉多相位路口交通信号的控制模型,以交叉路口的交通流量预测和交通信号控制为研究对象,以车辆延误时间最少为优化目标,实现对交通信号配时方案的调整,最终合理控制交通信号灯。主要研究了以下几方面的内容：(1)分析了智能交通控制系统关键技术的工作原理、研究现状与应用。在此基础上提出了基于流量预测的单交叉口短时交通预测控制系统的整体结构设计。(2)研究了一种实时流量采集的方法——以背景帧差法为前提的基于YUV色彩模型的目标车辆检测方法。并以杭州市某路口为例实地验证其检测的高效性和低错误率。(3)提出了短时交通流量预测的改进方案以及基于模糊神经网络的交通流量预测模型并利用粒子群(PSO)蚁群(ACO)混合算法对模型进行学习算法的改进。(4)建立了单交叉多相位路口的交通信号模糊神经网络控制模型,并采用粒子群算法对模型进行训练,获得最佳的信号配时方案。从而实现路口各相位信号灯的协调控制,确保系统的性能指标最优。(5)进行了针对所设计的单交叉多相位路口的基于短时流量预测的交通信号控制模型系统的仿真验证。
[Abstract]:With the rapid development of social economy, the popularization rate of automobile is also increasing. The number of cars is increasing day by day, which leads to the scarcity of urban road resources. Traffic congestion is becoming one of the biggest challenges faced by large and medium-sized cities in recent years. Therefore, an efficient traffic flow forecasting model and intelligent control of intersection signal lights are proposed to maximize the utilization of road resources and achieve high efficiency. It will play an important role in alleviating traffic congestion. In this paper, the traffic signal control model of single intersection and multi-phase intersection is designed, and the traffic flow prediction and traffic signal control of intersection are taken as the research object. Taking the minimum vehicle delay time as the optimization goal, the traffic signal timing scheme can be adjusted and the traffic signal light can be controlled reasonably. The main contents of this paper are as follows: 1) the working principle of the key technology of the intelligent traffic control system is analyzed. On the basis of this, the whole structure design of short-time traffic forecasting control system for single intersection based on flow prediction is proposed. (2) A real-time traffic acquisition method is studied, which is based on background frame difference method. The method of target vehicle detection based on YUV color model is presented. The high efficiency and low error rate of the detection are verified by an example of a intersection in Hangzhou. An improved method of short-term traffic flow prediction and fuzzy neural network are proposed. The traffic flow prediction model and the improved model learning algorithm by particle swarm optimization (PSO) ant colony algorithm (ACO) are used to establish the traffic signal fuzzy neural network control model of single intersection and multi-phase intersection. Particle swarm optimization (PSO) algorithm is used to train the model to obtain the best signal timing scheme, so as to realize the coordinated control of each phase signal light at the intersection. To ensure the optimal performance of the system, the simulation verification of the traffic signal control model system based on short-term flow prediction for the single intersection and multi-phase intersection is carried out.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.51;TP273

【参考文献】