论文部分内容阅读
针对目前国内外大中城市中普遍存在的无检测器信号交叉口车道交通流信息难于获取的情况,基于信号控制交叉口车道之间的相关性,综合应用聚类分析和逐步回归法预测单点无检测器信号控制交叉口车道流量.首先应用聚类分析将单点无检测器信号控制交叉口的车道与有检测器信号控制交叉口的车道交通流量进行聚类,然后在聚类分析结果的基础上随机选取车道交通流量样本运用逐步回归法预测单点无检测器信号控制交叉口的车道流量,此方法经过南京市的具体车道流量数据验证.此类问题的解决,可广泛应用于城市交通流诱导系统以及交通控制系统.
Aiming at the difficulty in obtaining traffic flow information of signal-free intersections at present in large and medium-sized cities both at home and abroad, based on the correlation between signal-controlled intersection lanes, the cluster analysis and stepwise regression are used to predict the single point No signal is used to control the traffic volume at the intersection.Firstly, cluster analysis is used to cluster the traffic volume between the traffic lane with single-point detectorless signal-controlled intersection and the intersection with the traffic signal with detector signal, Based on the random selection of traffic flow samples, stepwise regression method is used to predict the traffic flow of single-point detectorless signal control intersection, which is validated by Nanjing’s specific lane traffic data.The solution to these problems can be widely applied to urban traffic Flow guidance system and traffic control system.