考虑行人因素的交叉口信号最优控制研究

发布时间：2018-03-18 18:47

本文选题：行人预测　切入点：行人专用相位　出处：《重庆交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着社会和经济的发展,机动车保有量不断上升,城市交通拥堵变得越来越严重。这几年城市建设发展非常迅速,为了缓解交通拥堵,道路越修越宽,交叉口也越来越大,这样大大增加了行人通过交叉口的难度。目前多数的城市交通信号控制主要以机动车流畅通的为基础,并没有充分考虑行人过街的需求。随着城市生活水平的提高以及老龄化加速,对行人更加友好的路口交通管理与控制方法引起了人们的广泛重视。本文将研究充分考虑行人因素的交叉口交通信号控制,既保证了对行人过街友好,又保证机动车的畅通。首先,论文介绍了行人过街的交通特性,提出了将激光雷达检测器和摄像机两类检测器联合起来标定并检测行人。并运用高峰时期的行人流量数据,采用小波神经网络的算法对未来行人流量进行短期预测,得到高峰时期行人流的估计,解决了行人交通信息检测难、设备贵和数据传输时差等问题。其次,论文分析了一般情况下交叉口渠化设计的原则,采用了行人友好型渠化设计。分析了一般信号相位的设计方法、原则和相序确定的原则,提出了行人专用相位的方法,解决了交叉口行人和机动车间的冲突问题,满足了行人过街的需求。最后,论文提出了建立多目标优化配时方案的行人友好型交叉口控制模型,模型的多目标函数中加入了行人延误作为控制指标,建立了以行人的延误、交叉口的通行能力、车辆的延误和停车率为性能控制指标的多目标函数。采用了遗传算法对多目标函数的各个性能指标进行求解。并以重庆市某交叉口作为研究对象,采用考虑行人因素的配时方法和不考虑行人因素的配时方法对交叉口进行配时设计,对比了两种方法下的各个性能指标的变化。
[Abstract]:With the development of society and economy, the number of motor vehicles has been increasing, and the traffic jams have become more and more serious. In recent years, urban construction has developed very rapidly. In order to alleviate the traffic congestion, the roads are wider and wider, and the intersections are becoming larger and larger. This greatly increases the difficulty for pedestrians to pass through intersections. At present, most urban traffic signal controls are mainly based on the smooth flow of motor vehicles. The need for pedestrian crossing was not fully taken into account. With the improvement of living standards in cities and the acceleration of aging, People pay more attention to the more friendly traffic management and control method of pedestrian intersection. In this paper, we will study the traffic signal control of intersection which fully considers the pedestrian factor, which ensures the pedestrian crossing to be friendly. First of all, the paper introduces the traffic characteristics of pedestrians crossing the street, and proposes the combination of lidar detector and camera detector to calibrate and detect pedestrians. The wavelet neural network algorithm is used to predict the future passenger flow in the short term, and the estimation of the pedestrian flow during the peak period is obtained, which solves the problems such as difficult detection of pedestrian traffic information, expensive equipment and time difference of data transmission, etc. Secondly, In this paper, the principle of design of intersection is analyzed, and the design of pedestrian friendly channel is adopted. The design method of general signal phase, principle and principle of phase sequence determination are analyzed, and the method of special phase of pedestrian is put forward. The conflict between pedestrians and motor vehicles at intersections is solved, and the needs of pedestrians crossing the street are satisfied. Finally, the control model of pedestrian friendly intersections with multi-objective optimal timing scheme is proposed. In the multi-objective function of the model, the pedestrian delay is added as the control index, and the pedestrian delay and the traffic capacity of the intersection are established. The delay and parking rate of vehicles are multi-objective function of performance control index. Genetic algorithm is used to solve each performance index of multi-objective function, and a intersection in Chongqing is taken as the research object. The time matching method considering pedestrian factor and the time matching method without considering pedestrian factor are adopted to design the intersection. The changes of each performance index under the two methods are compared.
【学位授予单位】：重庆交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.54

【参考文献】