Research On The Influence Of Highway Alignment Continuous Degradation In 3-dimensional Space On Vehicle Trajectory Deviation

At this stage,my country’s road network is formed.By the end of 2020,the total mileage of roads will exceed 5.01.25 million kilometers.It is one of the countries with the most road mileage in the world.The subsequent road traffic accidents are still high.Whether the alignment design is reasonable or not is a fundamental issue related to road safety.The lack of consistency between the road alignment and the vehicle trajectory will increase the probability of accidents.Therefore,the alignment design should meet the requirements of the vehicle trajectory and driving law.Since the driving trajectory is a continuous spatial curve,the highway is required to provide a continuous line to match it.Early studies have shown that in the traditional horizontal and vertical combined design method,where the two-dimensional alignment elements change,there are varying degrees of jumps in curvature and torsion,which cause the continuous degradation of the spatial curve.This continuous degradation will inevitably cause changes in the trajectory of the vehicle,thereby affecting driving safety.Therefore,studying the characteristics of driving trajectory deviation has great theoretical significance for optimizing highway alignment safety design.To deeply explore the impact of the continuous degradation of the three-dimensional space curve on the deviation of the driving trajectory,this paper firstly focuses on some complicated two-way four-lane highways in some areas of Guangdong and uses a combination of fixedpoint video recording and natural driving tests to consider the road-vehicle coupling,collecting data on vehicle trajectory and actual driving state of the vehicle;Then,the relationship between different road conditions and vehicle trajectories in road traffic system is discussed,dividing road sections into 9 different alignment shapes according to different slopes and steering Combination conditions,under different alignment combination conditions,respectively,explore its trajectory offset characteristics,distribution law,normality and difference,and the results show that each section obeys the normal distribution;when the vehicle is downhill,the vehicle has a tendency to shift to the left;When the vehicle is turning to the left,most of the vehicles will behave severely to the left before entering the turn,and return to normal after entering the turn;when turning to the right,most of the vehicles will behave severely to the right after entering the turn,and behave normally before entering the turn;On the uphill section,the vehicle tends to drive on the right side of the lane most prominently.As a result,27 characteristic parameter indexes that affect the deviation of the driving trajectory are preliminarily selected.Finally,the correlation and principal component analysis method is used to establish the characterization index set,and the three machine learning algorithms of SVM,GF,and GBDT are used to construct the driver’s trajectory deviation model of the spatial curve continuous degradation condition.Through comparison and inspection,the GBDT model has the best applicability.It is recommended as a driving trajectory law analysis model.The GBDT model is used to verify and optimize the indicators based on the example road sections.The indicator threshold range and road traffic safety facilities improvement suggestions are proposed,which is based on a three-dimensional alignment optimization design method for the study of spatial alignment mechanism.The research results can provide a theoretical basis for evaluating the safety of highway geometric alignment,and can also provide a new perspective for road safety research based on driving trajectory.