An investigation of merging and diverging cars on a multi-lane road using a cellular automation model

In this paper, we have investigated two observed situations in a multi-lane road. The first one concerns a fast merging vehicle. The second situation is related to the case of a fast vehicle leaving the fastest lane back into the slowest lane and targeting a specific way out. We are interested in the relaxation time T, i.e., which is the time that the merging （diverging） vehicle spends before reaching the desired lane. Using analytical treatment and numerical simulations for the NaSch model, we have found two states, namely, the free state in which the merging （diverging） vehicle reaches the desired lane, and the trapped state in which T diverges. We have established phase diagrams for several values of the braking probability. In the second situation, we have shown that diverging from the fast lane targeting a specific way out is not a simple task. Even if the diverging vehicle is in the free phase, two different states can be distinguished. One is the critical state, in which the diverging car can probably reach the desired way out. The other is the safe state, in which the diverging car can surely reach the desired way out. In order to be in the safe state, we have found that the driver of the diverging car must know the critical distance （below which the way out will be out of his reach） in each lane. Furthermore, this critical distance depends on the density of cars, and it follows an exponential law.

A manoeuvre indicator and ensemble learning-based risky driver recognition approach for highway merging areas

Due to the complex traffic characteristics in highway merging areas,drivers tend to exhibit high-risk driving behaviours.To address the characteristics of driving behaviour in highway merging areas,we have developed a real-time identification model for risky drivers by combining a driver risk level labelling method with load balancing-ensemble learning(LB-EL).In this paper,we explore four types of manoeuvre indicator indexes(MIIs)—acute direction,stomp pedal,dangerous following and dangerous lane changing—that can describe the negative behaviours of both individual vehicles and vehicle platoons in highway merging areas.To quantize the label driver risk level,we use the interquartile range(IQR)method and Criteria Importance Though Intercriteria Correlation(CRITIC)while evaluating the reliability of the MII using spatial analysis.Furthermore,we balance the dataset using three load balancing(LB)algorithms and create nine ensemble strategies by pairing adaptive boosting(AdaBoost),extreme gradient boosting(XGBoost)and light gradient boosting machine(LGBM)with the three LB algorithms.Finally,we validate the proposed model using trajectory data extracted from unmanned aerial vehicle(UAV)videos.The results indicate that the distribution laws of risky driving behaviours in the acute direction and stomp pedal show a high degree of similarity and good matching with the distribution laws of traffic conflict points in existing research.Moreover,the synthetic minority over-sampling technique-light gradient boosting machine(SMOTE-LGBM)ensemble model achieves the best performance,reaching an accuracy rate of 93.4%,and a recall rate of 92.1%,which demonstrates the validity of our proposed model.This model can be widely applied to recognize risky drivers in video-based surveillance systems.