Research On The Propagation Mechanism Of Braking Deceleration In Expressway Motorcade Aiming To Online Warning Demand

Rear-end collision is the most common type of accidents on expressways,especially the serial rear-end collision often leads to serious consequences.Through the analysis of the rear end collision process,it is found that the correlation between the vehicles in the motorcade is strong when the motorcade follows closely on the highway.When the motorcade is slightly stimulated,the brittleness of the traffic system of the motorcade is stimulated,and gradually intensifies with the spread of the motorcade,which leads to the chain collapse of the traffic system,and eventually leads to the chain rear-end collision accidents.In view of the above characteristics,if the operation status of the whole motorcade is obtained through roadside equipment or Internet of vehicles technology,when there are interference incentives and it is easy to cause rear-end collision accidents,technical means can be taken to carry out active online prevention and control of rear-end collision risk.In this paper,the real road driving experiment and roadside fixed acquisition experiment are carried out,and a large number of vehicle following data are obtained.A total of 355normal following segments of 1679 vehicles and 665 groups of following braking segments are selected.Based on the above motorcade driving data,the propagation mechanism of braking deceleration inside the motorway motorcade is modeled and analyzed,and the braking deceleration prediction of each vehicle in the expressway motorcade is realized in the dangerous following situation,and the early warning scheme is verified by relying on a real case.The main research contents and conclusions of this paper are as follows:1.Based on the above data,the following behavior characteristics are analyzed,and the driving differences of different vehicle numbers are compared.Statistical analysis shows that there are a large number of close following behaviors on expressways,and with the increase of the number of vehicles in the motorcade,The potential risk of following increases.and establishes the following distance risk early warning model and braking response time prediction model.2.Based on SVM（Support Vector Machine）algorithm,the prediction model of braking propagation type is established.By comparing the prediction effects of different parameters,classification standards and parameter optimization methods,the optimal prediction model of braking propagation type is determined as five parameter two classification PSO-SVM model.Based on Time Headway（THW）,the reciprocal of Time to Collision（iTTC）and Risk Perception（RP）parameter,the prediction models of the braking deceleration of the rear vehicle under the dangerous following type are established respectively.After comparing the fitting effect of the three models,the optimal braking deceleration model is a₂vs.a₁,THW model.Based on the three-vehicle brake propagation data,the established brake propagation related models are verified.The verification results show that each model is suitable for multi-vehicle brake propagation,and the multi-vehicle brake deceleration prediction model is optimized.3.Considering the dynamic characteristics of vehicle braking process,the minimum safety braking deceleration model is established,and the parameters are optimized according to the measured data.According to the degree of following danger,the hierarchical warning system is established.A real case is used to verify the early warning algorithm,and the results show that the early warning scheme is reasonable,which provides accurate algorithm support for motorcade rear-end collision risk prevention and control.