Speed Prediction Model For Backward-Curve Section Of Mountain Road

The backward-curve as one of the most common road forms in mountain area,the linear design index is lower than the main road,and the vehicle running state is obviously different from the general road section.The research on the prediction model of the operating speed of the curve section can provide reference for the design and evaluation of linear coordination,and also supplement the prediction model of the two-lane highway operating speed.On the basis of the existing research on the prediction of operating speed at domestic and overseas,the following research is carried out by using the real vehicle test:Firstly,according to the speed factor analysis,through on-the-spot investigation mountainous area highway road vehicle running status on backward-curve,combined with the linear property and the vehicle running characteristic of the backward-curve section,and making real vehicle test plan formulation by the on-board CGI-610 centimeter level integrated navigation system,the mountain road section curve back operating speed,road linear for continuity of data related to the acquisition,and data filtering,screening,repair,etc.Secondly,through the analysis of the measured operating speed data of the backward-curve section of the mountain road,the results show that the overall operating speed of the backward-curve section shows a three-stage change rule of"deceleration-stability-acceleration".Basis for operating speed three phase change law of the backward-curve section is divided into "entrance curve-master curveexport curve" three stages,and defined the backward-curve road operating speed difference value Av,used to expound the transition function of auxiliary curve in the backward-curve sections,studies show that the set of auxiliary curve for transition curve road vehicle running back significantly;Combined with linear indicators,the backward-curve road vehicle speed and acceleration distribution,results show that the initial speed of the curve radius and the bent backward-curve road operating speed affect significantly,in addition,the plane curve length,route Angle,longitudinal slope,export linear indicators such as the impact on the operating speed to some extent.Finally,based on the analysis of operating speed variation characteristics and the selection of linear indicators,the operating speed prediction model of mountain road on backward-curve section was established by using random forest algorithm with Python software,and the reliability of the model was verified by ten-fold cross verification.The establishment of the prediction model of the operating speed of the backward-curve section is great significance to the improvement of the highway safety measures in the mountain area,which can provide reference for the optimization of the highway alignment design index,the decision of speed limit and the improvement of road traffic safety facilities.