Modeling differential car-following behavior under normal and rainy conditions: A memory-based deep learning method with attention mechanism

In order to analyze and learn the difference in car-following behavior between normal and rainy days, we first collect car-following trajectory data of an urban elevated road on normal and rainy days by microwave radar and analyze the differences in speed, relative speed, acceleration, space headway, and time headway among data through statistics. Secondly, owing to the time-series characteristics of car-following data, we use the long short-term memory(LSTM) neural network optimized by attention mechanism(AM) and sparrow search algorithm(SSA) to learn the different car-following behaviors under different weather conditions and build corresponding models(ASL-Normal, ASL-Rain, where ASL stands for AM-SSA-LSTM), respectively. Finally, the simulation test shows that the mean square error(MSE) and reciprocal of time-to-collision(RTTC) of the ASL model are better than those of LSTM and intelligent diver model(IDM), which is closer to the real data. The ASL model can better learn different driving behaviors on normal and rainy days. However,it has a higher sensitivity to weather conditions from cross test on normal and rainy data-sets which need classification training or sample diversification processing. In the car-following platoon simulation, the stability performances of two models are excellent, which can describe the basic characteristics of traffic flow on normal and rainy days. Comparing with ASL-Rain model, the convergence time of ASL-Normal is shorter, reflecting that cautious driving behavior on rainy days will reduce traffic efficiency to a certain extent. However, ASL-Normal model produces a more severe and frequent traffic oscillation within a shorter period because of aggressive driving behavior on normal days.

A new car-following model considering velocity anticipation

The full velocity difference model proposed by Jiang et al. [2001 Phys. Rev. E 64 017101] has been improved by introducing velocity anticipation. Velocity anticipation means the follower estimates the future velocity of the leader. The stability condition of the new model is obtained by using the linear stability theory. Theoretical results show that the stability region increases when we increase the anticipation time interval. The mKdV equation is derived to describe the kink-antikink soliton wave and obtain the coexisting stability line. The delay time of car motion and kinematic wave speed at jam density are obtained in this modeh Numerical simulations exhibit that when we increase the anticipation time interval enough, the new model could avoid accidents under urgent braking cases. Also, the traffic jam could be suppressed by considering the anticipation velocity. All results demonstrate that this model is an improvement on the full velocity difference model.

Feedback control scheme of traffic jams based on the coupled map car-following model

Based on the pioneering work of Konishi et al. [Phys. Rev. E （1999） 60 4000], a new feedback control scheme is presented to suppress traffic jams based on the coupled map car-following model under the open boundary condition. The effect of the safe headway on the traffic system is considered. According to the control theory, the condition under which traffic jams can be suppressed is analyzed. The results are compared with the previous results concerning congestion control. The simulations show that the suppression performance of our scheme on traffic jams is better than those of the previous schemes, although all the schemes can suppress traffic jams. The simulation results are consistent with theoretical analyses.

A new coupled map car-following model considering drivers' steady desired speed

Based on the pioneering work of Konishi et al., in consideration of the influence of drivers＇ steady desired speed ef/ect on the traffic flow, we develop a new coupled map car-following model in the real world. By use of the control theory, the stability condition of our model is derived. The validity of the present theoretical scheme is verified via numerical simulation, confirming the correctness of our theoretical analysis.

A velocity-difference-separation model for car-following theory

We introduce a velocity-difference-separation model that modifies the previous models in the literature. The improvement of this new model over the previous ones lies in that it not only theoretically retains many strong points of the previous ones, but also performs more realistically than others in the dynamical evolution of congestion. Furthermore, the proposed model is investigated with analytic and numerical methods, with the finding that it can demonstrate some complex physical features observed in real traffic such as the existence of three phases： free flow, synchronized flow, and wide moving jam; sudden flow drop in flow-density plane; and traffic hysteresis in transition between the free and the synchronized flow.

A new car-following model with consideration of the traffic interruption probability

In this paper, we present a new car-following model by taking into account the effects of the traffic interruption probability on the car-following behaviour of the following vehicle. The stability condition of the model is obtained by using the linear stability theory. The modified Korteweg-de Vries （KdV） equation is constructed and solved, and three types of traffic flows in the headway sensitivity space-stable, metastable, and unstable--are classified. Both the analytical and simulation results show that the traffic interruption probability indeed has an influence on driving behaviour, and the consideration of traffic interruption probability in the car-following model could stabilize traffic flow.

A control method for congested traffic in the coupled map car-following model

Based on the pioneer work of Konishi et al, a new control method is presented to suppress the traffic congestion in the coupled map （CM） car-following model under an open boundary. A control signal concluding the velocity differences of the two vehicles in front is put forward. The condition under which the traffic jam can be contained is analyzed. The results axe compared with that presented by Konishi et al [Phys. Rev. 1999 E 60 4000-4007]. The simulation results show that the temporal behavior obtained by our method is better than that by the Konishi＇s et al. method, although both the methods could suppress the traffic jam. The simulation results are consistent with the theoretical analysis.

Driving decision-making analysis of car-following for autonomous vehicle under complex urban environment

The decision-making under complex urban environment become one of the key issues that restricts the rapid development of the autonomous vehicles. The difficulty in making timely and accurate decisions like human beings under highly dynamic traffic environment is a major challenge for autonomous driving. Car-following has been regarded as the simplest but essential driving behavior among driving tasks and has received extensive attention from researchers around the world. This work addresses this problem and proposes a novel method RSAN(rough-set artificial neural network) to learn the decisions from excellent human drivers. A virtual urban traffic environment was built by Pre Scan and driving simulation was conducted to obtain a broad set of relevant data such as experienced drivers' behavior data and surrounding vehicles' motion data. Then, rough set was used to preprocess these data to extract the key influential factors on decision and reduce the impact of uncertain data and noise data. And the car-following decision was learned by neural network in which key factor was the input and acceleration was the output. The result shows the better convergence speed and the better decision accuracy of RSAN than ANN. Findings of this work contributes to the empirical understanding of driver's decision-making process and it provides a theoretical basis for the study of car-following decision-making under complex and dynamic environment.

An improved car-following model with consideration of the lateral effect and its feedback control research

A car-following model is presented, in which the effects of non-motor vehicles on adjacent lanes are taken into ac- count. A control signal including the velocity differences between the following vehicle and the target vehicle is introduced according to the feedback control theory. The stability condition for the new model is derived. Numerical simulation is used to demonstrate the advantage of the new model including the control signal; the results are consistent with the analytical ones

A new coupled-map car-following model based on a transportation supernetwork framework

A new car-following model is proposed by considering information from a number of preceding vehicles with intervehicle communication. A supernetwork architecture is first described, which has two layers： a traffic network and a communication network. The two networks interact with and depend on each other. The error dynamic system around the steady state of the model is theoretically analyzed and some nonjam criteria are derived. A simple control signal is added to the model to analyze the criteria of suppressing traffic jams. The corresponding numerical simulations confirm the correctness of the theoretical analysis. Compared with previous studies concerning coupled map models, the controlled model proposed in this paper is more reasonable and also more effective in the sense that it takes into account the formation of traffic congestion.

A control method applied to mixed traffic flow for the coupled-map car-following model

In light of previous work [Phys. Rev. E 60 4000 （1999）], a modified coupled-map car-following model is proposed by considering the headways of two successive vehicles in front of a considered vehicle described by the optimal velocity function. The non-jam conditions are given on the basis of control theory. Through simulation, we find that our model can exhibit a better effect as p = 0.65, which is a parameter in the optimal velocity function. The control scheme, which was proposed by Zhao and Gao, is introduced into the modified model and the feedback gain range is determined. In addition, a modified control method is applied to a mixed traffic system that consists of two types of vehicle. The range of gains is also obtained by theoretical analysis. Comparisons between our method and that of Zhao and Gao are carried out, and the corresponding numerical simulation results demonstrate that the temporal behavior of traffic flow obtained using our method is better than that proposed by Zhao and Gao in mixed traffic systems.

Traffic chaos and its prediction based on a nonlinear car-following model

This paper discusses the dynamic behavior and its predictions for a simulated traffic flow based on the nonlinear response of a vehicle to the leading car＇s movement in a single lane. Traffic chaos is a promising field, and chaos theory has been applied to identify and predict its chaotic movement. A simulated traffic flow is generated using a car-following model（ GM model）, and the distance between two cars is investigated for its dynamic properties. A positive Lyapunov exponent confirms the existence of chaotic behavior in the GM model. A new algorithm using a RBF NN （radial basis function neural network） is proposed to predict this traffic chaos. The experiment shows that the chaotic degree and predictable degree are determined by the first Lyapunov exponent. The algorithm proposed in this paper can be generalized to recognize and predict the chaos of short-time traffic flow series

Numerical Simulation of Motor Vehicles’ Courtesy Behavior on Crosswalk Based on Car-Following Model

In observing driver courtesy towards pedestrians at unsignalized crosswalks, a behavioral model was adopted in a simulation based on the GM Car-Following Model. The SIMI Motion Software was used to extract the vehicle operation data from Wenyi South Road and Hanyuan Road in Xi'an City. The parameters of the GM Car-Following Model were calibrated by genetic algorithm. The road simulation environment based on the Car-Following Model was constructed by MATLAB. In the case of no stopping, uniform deceleration avoidance with advance notice, emergency brake avoidance without advance notice, changes such as the displacement of the Car-Following queue, headway, speed, acceleration, and deceleration were analyzed by numerical simulation. The results show that when there is advance notice before the crosswalk, the minimum headway distances of Car1-Car2, Car2-Car3, Car3-Car4 and Car4-Car5 are 7.09 m, 7.38 m, 7.65 m, 7.91 m, and the average rates of change of the headway during deceleration are 0.78 m/s, 0.74 m/s, 0.71 m/s, 0.68 m/s respectively;in the absence of advance notice before the crosswalk, the minimum headway distances of Car1-Car2, Car2-Car3, Car3-Car4 and Car4-Car5 are 7.28 m, 7.75 m, 8.19 m, 8.59 m, and the average rates of change of the headway during deceleration are 1.57 m/s, 1.25 m/s, 1.04 m/s, 0.96 m/s, respectively. Therefore, in order to effectively prevent the occurrence of vehicle rear-end events, it's necessary to set traffic signs and markings on the preceding section of the intersection or road exhibiting behavioral comity.

Modeling and analysis of car-following behavior considering backward-looking effect

The car-following behavior can be influenced by its driver’s backward-looking effect.Especially in traffic congestion,if vehicles adjust the headway by considering backward-looking effect,the stability of traffic flow can be enhanced.A model of car-following behavior considering backward-looking effect was built using visual information as a stimulus.The critical stability conditions were derived by linear and nonlinear stability analyses.The results of parameter sensitivity analysis indicate that the stability of traffic flow was enhanced by considering the backward-looking effect.The spatiotemporal evolution of traffic flow of different truck ratios and varying degrees of backward-looking effect was determined by numerical simulation.This study lays a foundation for exploring the complex feature of car-following behavior and making the intelligent network vehicles control rules more consistent with human driver habits.

A car-following model with the anticipation effect of potential lane changing

In this paper, a new car-following model is presented, taking into account the anticipation of potential lane changing by the leading vehicle. The stability condition of the model is obtained by using the linear stability theory. The modified Korteweg-de Vries （KdV） equation is constructed and solved, and three types of traffic flow in the headway-sensitivity space, namely stable, metastable and unstable ones, are classified. Both the analytical and simu- lation results show that anxiety about lane changing does indeed have an influence on driving behavior and that a consideration of lane changing probability in the car-following model could stabilize traffic flows. The quantitative relationship between stability improvement and lane changing probability is also investigated.

Improved Car-Following Model Considering Modified Backward Optimal Velocity and Velocity Difference with Backward-Looking Effect

In this paper, a new traffic flow model called the forward-backward velocity difference (FBVD) model based on the full velocity difference model is proposed to investigate the backward-looking effect by applying a modified backward optimal velocity using generalized backward maximum speed. The FBVD model belongs to the family of microscopic models that consider spatiotemporally continuous formulations. Neutral stability conditions of the discrete car-following model are derived using the linear stability theory. The stability analysis results prove that the modified backward optimal velocity has a significant positive effect in stabilizing the traffic flow. Through nonlinear analysis, a kink-antikink solution is derived from the modified Korteweg-de Vries equation of the FBVD model to explain traffic congestion of the model. The validity of this theoretical model is checked using numerical results, according to which traffic jams were found to have been significantly diminished by the introduction of the modified backward optimal velocity.

A new car-following model yielding log-normal type headways distributions

Modeling time headways between vehicles has attracted increasing interest in the traffic flow research field recently, because the corresponding statistics help to reveal the intrinsic interactions governing the vehicle dynamics. However, most previous micro-simulation models cannot yield the observed log-normal distributed headways. This paper designs a new car-following model inspired by the Galton board to reproduce the observed time-headway distributions as well as the complex traffic phenomena. The consistency between the empirical data and the simulation results indicates that this new car-following model provides a reasonable description of the car-following behaviours.

ONE-DIMENSIONAL CELLULAR AUTOMATON MODEL OFTRAFFIC FLOW BASED ON CAR-FOLLOWING IDEA

An improved one-dimensional CA ( Cellular Automaton) traffic model was proposed to describe the highway traffic under the periodic boundary conditions. This model was based on the idea of the car-following model, which claims that the motion of a vehicle at one time step depends on both its headway and the synchronous motion of the front vehicle, thus including indirectly the influence of its sub-neighboring vehicle. It? addition, the so-called safety distance was introduced to consider the deceleration behavior of vehicles and the stochastic factor was taken into account by introducing the deceleration probability. Meanwhile, the conditional deceleration in the model gives a better description of the phenomena observed on highways. It is found that there exists the metastability and hysteresis effect of traffic flow in the neighborhood of critical density under different initial conditions. Since this model gives a reasonable depiction of the motion of a single vehicle, it is easy to be extended to the case of traffic flow tinder the control of traffic lights in cities.

Stabilization strategy of a car-following model with multiple time delays of the drivers

An extended car-following model with multiple delays is constructed to describe driver's driving behavior.Through stability analysis,the stability condition of this uncontrolled model is given.To dampen the negative impact of the driver's multiple delays(i.e.,stability condition is not satisfied),a novel control strategy is proposed to assist the driver in adjusting vehicle operation.The control strategy consists of two parts:the design of control term as well as the design of the parameters in the term.Bifurcation analysis is performed to illustrate the necessity of the design of parameters in control terms.In the course of the design of parameters in the control term,we improve the definite integral stability method to reduce the iterations by incorporating the characteristics of bifurcation,which can determine the appropriate parameters in the control terms more quickly.Finally,in the case study,we validate the control strategy by utilizing measured data and configuring scenario,which is closer to the actual traffic.The results of validation show that the control strategy can effectively stabilize the unstable traffic flow caused by driver's delays.

Simulating train movement in railway traffic using a car-following model

Based on a car-following model, in this paper, we propose a new traffic model for simulating train movement in railway traffic. In the proposed model, some realistic characteristics of train movement are considered, such as the distance headway and the safety stopping distance. Using the proposed traffic model, we analyse the space-time diagram of traffic flow, the trajectory of train movement, etc. Simulation results demonstrate that the proposed model can be successfully used for simulating the train movement. Some complex phenomena can be reproduced, such as the complex acceleration and deceleration of trains and the propagation of train delay.