Abstract: Based on the measured traffic loading data, the reliability theory was employed to analyze the traffic flow characteristics. The probability distribution of the given type heavy truck was a double Weibull distribution, which was put forward according to the measured traffic volume and vehicle weight information. On the basis of the measured wheelbase, the wheelbase representative values were obtained by using kernel density estimation method. In accordance with the measured vehicle velocity, the distribution type of vehicle velocity was also obtained. Moreover, the distribution of 0.05 tantile was used as the representative vehicle velocity. A relational expression among the maximum loading (Wmax), mean (μ), and standard deviation (σ) of vehicle weight could be fitted by using regression analysis. Furthermore, the variation coefficient could be obtained by using statistics method based on vehicle weight’s samples. Finally, the multi-lane transverse reduction factors suitable for the current traffic condition were proposed by the probabilistic algorithms.

Key words: bridge engineering, measured traffic data, reliability theory, probabilistic algorithms, multi-lane transverse reduction factor

摘要： 采用可靠度理论并针对实测交通数据对交通流特性进行统计分析。根据实测交通量和车重信息提出给定车型的重车概率分布类型为双重威布尔分布，依据轴距数据运用核密度估计分析得出轴距的代表值，通过速度信息得出车速的分布类型，并取分布的0.05分位值作为车速的代表值。采用回归分析拟合最大荷载Wmax与均值μ及标准差σ之间的关系式并获取车重样本的变异系数变化情况，进一步采用概率算法得出了适用于当前交通状况的多车道荷载横向折减系数。

关键词: 桥梁工程, 实测交通数据, 可靠度理论, 概率算法, 多车道横向折减系数