论文部分内容阅读
利用随机试验和数学均值化原理,对收集的单车事故数据进行了分析,运用定积分的元素法计算单车事故率面积函数,并对其拟合求导。结合面积函数的线形变化规律,筛选出单车交通事故影响因子和道路几何要素变量,逐一建立了各独立变量对事故率的影响模型,并确定了最终的变量。采用回归分析方法构建不同半径和直线长度的单车事故预测模型,利用三维曲面图对模型进行分析,提出了高速公路设计中适宜采用的半径和直线长度取值区间。研究结果表明:高速公路设计中采用直线-曲线-直线的线形组合时,适宜曲线半径区间应为[2 500,3 500]和直线长度区间应为[1 000,2 000];在曲线半径大于不设超高的最小半径时,与之相连的直线长度宜取1 000 m。
Based on the random test and the mathematical mean value theory, the data of bicycle accident collected are analyzed, and the area function of bicycle accident rate is calculated by the element method of definite integral. Combining with the linear change law of area function, the influence factors of traffic accident and road geometric factors were screened out. The influence models of each independent variable on accident rate were established one by one, and the final variables were determined. The forecasting model of bicycle accident with different radius and straight line length was constructed by regression analysis method. The model was analyzed by three-dimensional surface map, and the radius and straight line length interval suitable for highway design was proposed. The results show that when linear combination of straight line, curve and straight line is adopted in highway design, the radius of suitable curve should be [2 500,3 500] and the length of straight line should be [1 000,2 000]; when the radius of curve is greater than In the absence of an extremely high minimum radius, the length of the straight line to which it is connected shall be 1 000 m.