Abstract: The longitudinal vehicle speed curve obtained by cross-section acquisition method is not continuous, and the resulting vehicle speed model can not reflect the vehicle speed characteristics in the whole longitudinal slope section. In order to remedy the deficiency of vehicle speed prediction model based on cross-section speed method, a four-lane continuous downgrade section of National Highway G5 in Shanxi province was taken as the research object and 50 groups of continuous speed data of large truck with 14 gradients of the tested section were collected by using vehicle-mounted GPS equipment. The relationship between the vehicle speed V and the parameters such as slope P and slope length L of highway longitudinal slope were analyzed. By introducing the slope ratio r (i.e. the ratio of front to rear longitudinal slopes), a prediction model for the vehicle speed of large trucks on continuous downgrade sections based on P, L and r was established. The vehicle speed prediction model was tested by practical engineering data. The results show that the vehicle speed V is related to P of the longitudinal slope where the vehicle is located, and the relationship between the vehicle speed V and the L is cubic polynomial, which is obviously different from the quadratic polynomial relationship between V and L. The predicted vehicle speed of the proposed model is basically consistent with the actual vehicle speed, which proves the validity of the model.

Key words: highway engineering, continuous downgrade, vehicle speed, prediction model

摘要： 利用断面采集法得到的纵向行车速度曲线是不连续的，由此得到的行车速度模型并不能反映车辆在整个纵坡路段的车速特征。为了弥补基于断面车速法的相关速度预测模型的不足，以国道G5山西省境内的一段四车道连续下坡路段为研究对象，利用车载GPS设备，采集该试验路段14个坡度的50组大货车连续行车速度数据，分析了行车速度V与公路纵坡坡度P、坡长L等参数的关系；引入了坡比r(即车辆经过的前后纵坡的比值)参数，建立了连续下坡路段基于坡度P、坡长L和坡比r的大货车运行速度预测模型；用实际工程对模型进行了检验。结果表明：车速V与车辆所在纵坡的前后坡度P均有关系；车速V与坡长L呈三次多项式关系，这与已有研究的二次多项式关系明显不同；模型预测车速与实测车速基本一致，证明了模型的有效性。

关键词: 道路工程, 长下坡, 行车速度, 预测模型