摘要
信号交叉口排队车辆静态间距对交叉口的通行效率和交通安全均有重要影响。以哈尔滨市26辆出租车的行车记录仪视频为数据基础,利用计算机单目视觉技术,构建了基于前车车牌高度的信号交叉口排队车辆静态间距测算模型,进而结合统计分析方法,探讨了静态间距的分布特性,分析了驾驶员因素(年龄和出租车驾龄)和道路交通条件因素(车型、是否高峰时段以及车道类型)对静态间距的影响。研究结果表明:所构建的静态间距测算模型具有较高的测算精度;对测算得到的1 896个静态间距样本的分析发现,信号交叉口排队车辆静态间距服从Loglogistic分布;静态间距与驾驶员年龄和出租车驾龄存在一定的负相关趋势,并且随着年龄或驾龄的增加将趋向于临界安全静态间距;高平峰时段和前车车型对静态间距有显著影响,且非高峰时段、前车为大中型车等因素会导致静态间距的增加,而当前车辆所在车道类型对静态间距没有显著影响。上述研究成果对全面掌握和理解信号交叉口的交通运行规律具有重要意义。
Static spacing of queuing vehicles at signalized intersection influences the operating efficiency as well as the traffic safety significantly. Based on the video collected by driving recorder from 26 taxis in Harbin City, and by using the computer monocular vision technique, the static spacing measuring model of queuing vehicles at signalized intersection on the basis of the height of vehicle licence plate is established. Then, combining the statistical analysis approaches, the distribution character of static spacings is discussed, and the taxi driver factors (age, taxi driving years) and road traffic conditions (vehicle type, peak hours or not, and lane type ) that might affect the static spacings are analyzed. The research result shows that the established static spacing measuring model has a high measuring accuracy. Based on the analysis of 1 896 static spacing samples, it is found that ( 1 ) the static spacing of queuing vehicles at signalized intersections generally follow the Log-logistic distribution; (2) the static spacings and the ages of drivers as well as the taxi driving years are negatively correlated in general, and they will finally tend to the critical safe static spacingwith their increase; (3) the static spacings are significantly affected by the peak-flat hours and the front vehicle type, and the static spacing would increase when they are in non-peak hours and the front vehicle is large/middle sized vehicle, while the type of the lane occupied by the subject vehicle has little effect on the static spacing. The research result is important for fully understanding the operation rules of the vehicles at signalized intersections.
出处
《公路交通科技》
CAS
CSCD
北大核心
2018年第2期95-100,108,共7页
Journal of Highway and Transportation Research and Development
基金
中国博士后科学基金项目(2016M590285)
黑龙江省博士后基金项目(LBH-Z15092)
关键词
交通工程
测距模型
计算机视觉
静态间距
信号交叉口
分布特性
traffic engineering
distance measuring model
computer vision
static spacing
signalizedintersection
distribution character