Search

Continuous developments in automotive technology brought about the reality of having an increasing number of autonomous vehicles (AV) on roadways soon. However, this advancement comes with the uncertainty of how the interaction between the autonomous and non-autonomous (non-AV) vehicles will play out. Using traffic simulators to model both AVs and non-AVs, and studying the effect of AVs on roadway mobility can help researchers in learning what to expect in real life. In this paper, an open-source simulator called Simulation of Urban Mobility (SUMO) was used to develop AVs by modifying Intelligent Driver Model (IDM) car following and LC2013 lane changing models. Then, three actual roadways in Columbus, Ohio were modeled, and 40 simulations were run for each traffic network. The first set of simulations focused on the impacts that AVs have on vehicle roadway mobility with non-AVs when all the AVs have the same autonomy level. The second set of simulations had mixed AV autonomy levels (SAE Level 2, 3 and 4) with mixed AV and non-AV traffic. The analysis revealed that mobility of AVs and non-AVs increased generally as autonomy level increased. When the AV autonomy levels were mixed, the non-AVs experienced lower average speeds and decreased roadway mobility.

This paper develops a methodology for simultaneously modeling lane-changing and car-following behavior of automated vehicles on freeways. Naturalistic driving data from the Safety Pilot Model Deployment (SPMD) program are used. First, a framework to process the SPMD data is proposed using various data analytics techniques including data fusion, data mining, and machine learning. Second, pairs of automated host vehicle and their corresponding front vehicle are identified along with their lane-change and car-following relationship data. Using these data, a lane-changing-based car-following (LCCF) model, which explicitly considers lane-change and car-following behavior simultaneously, is developed. The LCCF model is based on Gaussian-mixture-based hidden Markov model theory and is disaggregated into two processes: LCCF association and LCCF dissociation. These categories are based on the result of the lane change. The overall goal is to predict a driver’s lane-change intention using the LCCF model. Results show that the model can predict the lane-change event in the order of 0.6 to 1.3 s before the moment of the vehicle body across the lane boundary. In addition, the execution times of lane-change maneuvers average between 0.55 and 0.86 s. The LCCF model allows the intention time and execution time of driver’s lane-change behavior to be forecast, which will help to develop better advanced driver assistance systems for vehicle controls with respect to lane-change and car-following warning functions.

Automatic lane change is one of the most important highway operations. It seriously affects traffic efficiency and safety. It is also an important driving technology for automatic driving. To achieve the best automatic lane-change control, it is necessary to achieve the control from the perspective of multi-objective evaluation. In this paper, to make it applicable for a hybrid condition of car following and lane change, the traditional car-following model is modified by regarding the longitudinal motion during the lane-changing process as a transition of the car-following behavior in the two lanes before and after a certain lane-change behavior. A hyperbolic tangent transition function is introduced to connect the model to achieve a smooth transition of the model output. Then, the discretionary lane-change decision process of highway autonomous vehicles is modeled into a two-vehicle game model, and a comprehensive loss function concerning safety, efficiency, and ride comfort is proposed for the evaluation of the strategies. The optimal strategy is obtained by minimizing the expectation of losses. Finally, to verify the performance of the proposed new model, simulations of different car-following and lane-changing models are carried out, which is for multi-target simulation conditions. The results of the simulation show that the new model exhibits higher traffic efficiency, better homogeneity, and stability.

Before a car-following model can be applied in practice, it must first be validated against real data in a process known as calibration. This paper discusses the formulation of calibration as an optimization problem, and compares different algorithms for its solution. The optimization consists of an arbitrary car following model, posed as either an ordinary or delay differential equation, being calibrated to an arbitrary source of trajectory data which may include lane changes. Typically, the calibration problem is solved using gradient free optimization. In this work, the gradient of the optimization problem is derived analytically using the adjoint method. The computational cost of the adjoint method does not scale with the number of model parameters, which makes it more efficient than evaluating the gradient numerically using finite differences. Numerical results are presented which show that quasi-newton algorithms using the adjoint method are significantly faster than a genetic algorithm, and also achieve slightly better accuracy of the calibrated model., accepted to transportation science

This study tests the hypothesis that an analytically estimated driver reaction time required for asymptotic stability, based on the macroscopic Gazis-Herman-Rothery (GHR) model, serves as an indica...

Microscopic simulation models have been widely used as tools to investigate the operation of traffic systems and different intelligent transportation systems applications. The fidelity of microscopic simulation tools depends on the driving behavior models that they implement. However, current models commonly do not consider human-related factors, such as distraction. The potential for distraction while driving has increased rapidly with the availability of smartphones and other connected and infotainment devices. Thus, an understanding of the impact of distraction on driving behavior is essential to improve the realism of microscopic traffic tools and support safety and other applications that are sensitive to it. This study focuses on car-following behavior in the context of distracting activities. The parameters of the well-known GM and intelligent driver models are estimated under various distraction scenarios using data collected with an experiment conducted in a driving simulator. The estimation results show that drivers are less sensitive to their leaders while talking on the phone and especially while texting. The estimated models are implemented in a microscopic traffic simulation model. The average speed, coefficient of variation of speed, acceleration noise and acceleration and deceleration time fractions were used as measures of performance indicating traffic flow and safety implications. The simulation results show deterioration of traffic flow with texting and to some extent talking on the phone: average speeds are lower and the coefficient of variation of speeds are higher. Further experimentation with varying fractions of texting drivers showed similar trends.

There is little research on the degrees of drivers’ short-term behaviours regarding driving safety. To solve this problem, this paper investigated the concept of driving proneness and evaluated the propensities of different drivers to engage in different operations for the following scenarios of urban traffic. From the real driving data of sixteen drivers on a city road, car-following data fragments were extracted and six key parameters were obtained: throttle percentage, change rate of throttle percentage, brake pressure, change rate of brake pressure, absolute value of steering angle and absolute value of steering angle speed. Symbolic Aggregate Approximation was used to reduce the dimensionality of the parameters. The input of the Hidden Markov Model-Viterbi was obtained by the use of statistical methods. The output of the model is the probability of the three proneness states of introversion, neutrality and extroversion, from which the proneness value of each driver was calculated. The weighted proneness value of each driver was obtained by the use of the entropy weight method to assign weights to each parameter. The operating characteristics of the drivers were also analysed and described. The method presented in this paper can provide accurate and real-time warning in network-driven environments.

Automated vehicles (AVs) are expected to improve traffic flow efficiency and safety. The deployment of AVs on motorways is expected to be the first step in their implementation. One of the main concerns is how human drivers will interact with AVs. Dedicating specific lanes to AVs have been suggested as a possible solution. However, there is still a lack of evidence-based research on the consequence of dedicated lanes for AVs on human drivers’ behavior. To bridge this research gap, a driving simulator experiment was conducted to investigate the behavior of human drivers exposed to different road design configurations of dedicated lanes on motorways. The experiment sample consisted of 34 (13 female) licensed drivers in the age range of 20–30. A repeated measures ANOVA was applied, which revealed that the type of separation between the dedicated lane and the other lanes has a significant influence on the behavior of human drivers driving in the proximity of AV platoons. Human drivers maintained a significantly lower time headway (THW) when driving in the proximity of a continuous access dedicated lane as compared to a limited-access dedicated lane with a guardrail separation for AV platoons. A similar result was found for the limited-access dedicated lane in comparison to the limited-access dedicated lane with guardrail separation. Moreover, the results regarding the empirical relationships between THW and sociodemographic variables indicate a significant THW difference between males and females as well as a significant inverse relationship between THW and the years of driving experience.

The platform of a microscopic traffic simulation provides an opportunity to study the driving behavior of vehicles on a roadway system. Compared to traditional conventional cars with human drivers, the car-following behaviors of autonomous vehicles (AVs) and connected autonomous vehicles (CAVs) would be quite different and hence require additional modeling efforts. This paper presents a thorough review of the literature on the car-following models used in prevalent micro-simulation tools for vehicles with both human and robot drivers. Specifically, the car-following logics such as the Wiedemann model and adaptive cruise control technology were reviewed based on the vehicle’s dynamic behavior and driving environments. In addition, some of the more recent “AV-ready (autonomous vehicles ready) tools” in micro-simulation platforms are also discussed in this paper.

This paper proposes a risk assessment method based on trajectory data which are used to quantify the risk faced by drivers for application in autonomous vehicles. A risk field is derived from the field theory of traffic flow, based on which the risk repulsion indicator of car-following is determined. By describing the repulsion force perceived by drivers in the process of car-following, the risk faced by drivers is assessed. The validity of the indicator is established from crash trajectory data obtained by simulation, and a binary logit model is employed to predict the crash. The result shows that the risk repulsion indicator based on risk field theory can distinguish crash states and non-crash states significantly. The prediction accuracy of binary logit model based on risk repulsion performs better than that of crash prediction model based on loop detector data.

A large portion of the accidents involving autonomous vehicles (AVs) are not caused by the functionality of AV, but rather because of human intervention, since AVs’ driving behavior was not properly understood by human drivers. Such misunderstanding leads to dangerous situations during interaction between AV and human-driven vehicle (HV). However, few researches considered HV-AV interaction safety in AV safety evaluation processes. One of the solutions is to let AV mimic a normal HV’s driving behavior so as to avoid misunderstanding to the most extent. Therefore, to evaluate the differences of driving behaviors between existing AV and HV is necessary. DRIVABILITY is defined in this study to characterize the similarity between AV’s driving behaviors and expected behaviors by human drivers. A driving behavior spectrum reference model built based on human drivers’ behaviors is proposed to evaluate AVs’ car-following drivability. The indicator of the desired reaction time (DRT) is proposed to characterize the car-following drivability. Relative entropy between the DRT distribution of AV and that of the entire human driver population are used to quantify the differences between driving behaviors. A human driver behavior spectrum was configured based on naturalistic driving data by human drivers collected in Shanghai, China. It is observed in the numerical test that amongst all three types of preset AVs in the well-received simulation package VTD, the brisk AV emulates a normal human driver to the most extent (ranking at 55th percentile), while the default AV and the comfortable AV rank at 35th and 8th percentile, respectively.

Driving simulator (DS) can be introduced as a popular tool in the transport industry for the analysis of driving behavior. Simulators should be validated in term of each specific recorded data, separately. When DSs are the source of data, each part of data which indicates a maneuver or a behavior should be validated separately. This paper is dedicated to the validation of car-following distance in the presence of considerable traffic flow which may increase mental load. In general, the results showed that DSs could be used for determining the behaviors of a driver by focusing on headway and car-following behavior. The distance kept by male drivers was shorter than female participants. It was seen that drivers between the ages of 20–40 were suitable for car-following tests in the DS by considering all real driving tasks in a congested traffic condition. Car-following threshold headway was less than 2 s for 85% of the participants, which is quite similar to the past results from real data in Iran. The average car-following headway threshold in the DS was 0.8 s, which is less than field data (1.2 s).

This paper proposes a new approach of using reinforcement learning (RL) to train an agent to perform the task of vehicle following with human driving characteristics. We refer to the ideal of inverse reinforcement learning to design the reward function of the RL model. The factors that need to be weighed in vehicle following were vectorized into reward vectors, and the reward function was defined as the inner product of the reward vector and weights. Driving data of human drivers was collected and analyzed to obtain the true reward function. The RL model was trained with the deterministic policy gradient algorithm because the state and action spaces are continuous. We adjusted the weight vector of the reward function so that the value vector of the RL model could continuously approach that of a human driver. After dozens of rounds of training, we selected the policy with the nearest value vector to that of a human driver and tested it in the PanoSim simulation environment. The results showed the desired performance for the task of an agent following the preceding vehicle safely and smoothly.

In a parameter study, we systematically varied parameter values, and quantified the resulting traffic flow in each individual lane. We modeled two-, three-, and four-lane freeway sections with the microscopic traffic flow simulation tool PTV Vissim. We compared the results with findings from literature. Simulations using car following model Wiedemann 99 fit better to empirical studies than those using Wiedemann 74. Empirically determinable parameters, that have a relevant influence on lane flow distribution are desired speed distributions (mean for heavy-duty vehicles and standard deviation for cars), heavy-duty vehicle share, and the gradient of the section. Additionally, the driving behavior parameters CC1 (headway time), CC3 (threshold for entering following), and safety distance reduction factor have an influence. As CC1 is one of the most relevant parameters for calibrating capacity, CC3 and the safety distance reduction factor remain for lane flow adjustment.

Human drivers can have diverse car-following behaviors when interacting with connected and automated vehicles (CAVs) and other human-driven vehicles in mixed traffic where many human-driven vehicles and a limited number of CAVs frequently interact and share the road. In this study, Inverse Reinforcement Learning (IRL) is used to model unique car-following behaviors of different human drivers when interacting with the CAV and another human-driven vehicle by using their driving demonstrations collected from in-field driving tests. The learned driver behavior model is shown that the personalized driving behaviors accurately and consistently can be characterized when following the different types of preceding vehicles in a variety of traffic situations. Furthermore, the energy efficiency of different human-driven vehicles when interacting with the CAV and the human-driven vehicle is investigated with the heterogeneous characteristics of drivers’ behaviors, considering driving behaviors have significant influences on vehicle fuel economy. A detailed analysis reveals the significant fuel-saving benefits of the CAV to the following human-driven vehicles during the car-following scenario and the extent of such benefits varies among tested human drivers owing to their intrinsic preferences and perception of CAV. These findings suggest that human-CAV interactions can be effectively leveraged to improve the energy efficiency of mixed traffic.

This paper analyzes the car following behavioral stochasticity based on two sets of field experimental trajectory data by measuring the wave travel time series τ ˜ n ( t ) of vehicle n. The analysis shows that (i) No matter the speed of leading vehicle oscillates significantly or slightly, τ ˜ n ( t ) might change significantly; (ii) A follower's τ ˜ n ( t ) can vary from run to run even if the leader travels at the same stable speed; (iii) Sometimes, even if the leader's speed fluctuates significantly, the follower can keep a nearly constant value of τ ˜ n ( t ) . The Augmented Dickey-Fuller test indicates that the time series ξ n ( t ) = d τ ˜ n ( t ) / d t follows a mean reversion process, no matter the oscillations fully developed or not. Based on the finding, a simple stochastic Newell model is proposed. The concave growth pattern of traffic oscillations has been derived analytically. Furthermore, simulation results demonstrate that the new model well captures both macroscopic characteristic of traffic flow evolution and microscopic characteristic of car following.

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.

Car-following theory has received considerable attention as a core component of Intelligent Transportation Systems. However, its application to the emerging autonomous vehicles (AVs) remains an unexplored research area. AVs are designed to provide convenient and safe driving by avoiding accidents caused by human errors. They require advanced levels of recognition of other drivers’ driving-style. With car-following models, AVs can use their built-in technology to understand the environment surrounding them and make real-time decisions to follow other vehicles. In this paper, we design an end-to-end car-following framework for AVs using automated object detection and navigation decision modules. The objective is to allow an AV to follow another vehicle based on Red Green Blue Depth (RGB-D) frames. We propose to employ a joint solution involving the You Look Once version 3 (YOLOv3) object detector to identify the leader vehicle and other obstacles and a reinforcement learning (RL) algorithm to navigate the self-driving vehicle. Two RL algorithms, namely Q-learning and Deep Q-learning have been investigated. Simulation results show the convergence of the developed models and investigate their efficiency in following the leader. It is shown that, with video frames only, promising results are achieved and that AVs can adopt a reasonable car-following behavior.

Traffic interruption is very easy to cause traffic congestion. For the problem of traffic interruption, a novel car-following model in this paper is presented to correct the effect of traffic self-interruption probability with the combination of the anticipation information communication under V2X environment. The stability condition concerning the self-interruption probability correction is acquired via linear analysis. The mKdV equation involving the self-interruption probability correction is deduced from nonlinear analysis. Moreover, numerical simulation demonstrates that the self-interruption probability correction can effectively relieve traffic jam, which agrees with analytical results.

The paper investigates the capability of modeling the car following behavior by training shallow and deep recurrent neural networks to reproduce observed driving profiles, collected in several experiments with pairs of GPS-equipped vehicles running in typical urban traffic conditions. The input variables are relative speed, spacing, and vehicle speed. In the model, we assume that the reaction is not instantaneous. However, it may occur during a time interval of the order of a few tenth of seconds because of both the psychophysical driver’s reaction process and the mechanical activation of braking or dispensing the traction power to the wheels. Experimental results confirm the reliability of this assumption and highlight that the deep recurrent neural network outperforms the simpler feed-forward neural network.

Model predictive control (MPC) has been widely adopted for cooperative adaptive cruise control (CACC) due to its superior performance in achieving fuel-efficient driving while satisfying constraints such as inter-vehicle distance. The core of an MPC-based algorithm is to predict the vehicle’s behavior using a dynamic model, and the space-domain vehicle dynamic model is frequently implemented in recent research along with the time-domain vehicle dynamic model. This paper presents a comparative performance analysis between the space-domain and the time-domain models in the MPC framework for the car-following problem. An MPC design process and analysis method for the high-speed car-following scenario is suggested and presented for equivalent performance comparison between the two approaches. In order to analyze trends between speed tracking and fuel-saving performance, which are conflicting objectives as car-following performance, a bi-objective cost function is proposed and manipulated by various weightings. It is observed that the space-domain model presents stable tracking performance, and the time-domain model shows better fuel efficiency. However, the space-domain model with road information is superior in tracking and fuel efficiency compared to the time-domain model with limited road information. Pareto analysis was implemented to visualize and describe performance differences in various situations regarding tracking error, fuel efficiency, and road grade information levels.

This paper presents a behavioral car following model, named the chained asymmetric behavior model, that improves on the asymmetric behavior model. This model is inspired by the empirical observation that vehicles react proportionately to the magnitude of disturbance experienced when traversing through a stop-and-go oscillation, deviating from a constant following behavior observed in equilibrium conditions. Findings from simulation experiments suggest that this “second-order” effect significantly affects traffic throughput and evolution under disturbances. Knowledge obtained from the model is leveraged toward designing control for connected automated vehicles in mixed traffic streams.

Connected and Automated Vehicles (CAVs) can potentially improve the performance of the transportation system by reducing human errors. This paper investigates the safety impact of CAVs in a mixed t...

Incorporation of Human Factors (HF) into the mathematical car-following (CF) models has always been the research hotspot. Ignorance of such inclusion would inevitably hinder us from acquiring a com...

Eco-driving strategies have been applied to smooth traffic flow and reduce greenhouse gas emissions along with air pollution. In this paper, we propose an eco-driving strategy to reduce traffic oscillation and smooth trajectories for connected vehicles following a moving bottleneck on a single-lane road. The eco-driving strategy, which leverages vehicle-to-vehicle (V2V) communications, designs advisory speed limits for each following vehicle through a control algorithm. The algorithm is based on the prediction of the following vehicle trajectories dictated by a moving bottleneck. The following vehicle trajectories are obtained by analytically solving the moving bottleneck problem in which the moving bottleneck speeds vary over time. In addition, the bounded acceleration rate is imposed in car-following behavior. The benefits of this strategy are demonstrated by applying it to four scenarios with different bottleneck movements. By simulating the scenarios with Newell’s car-following model with bounded acceleration and VT-Micro emission model, we find that both speed fluctuations and emissions are reduced with the algorithm in the scenarios in which the moving bottleneck has a constant speed, accelerates, decelerates and stops-and-goes. The results indicate that the proposed eco-driving algorithm can smooth traffic flow behind a moving bottleneck.

Comprehensive analysis of traffic behavior requires continuous studies to develop traffic theories explaining that behavior at the microscopic level. The study aims to develop a microsimulation program to evaluate the freeway weaving performance depending on the observed data. FWASIM represents a microscopic analysis of the freeway traffic features. It scans events periodically. The developed FWASIM involves the formulation of driver and vehicle behavior at freeway link, on-ramp, off-ramp, and combine them to produce a flexible, friendly use simulation model. Its concept is mainly depending on the car following and lane change theories. Analytical model validation was conducted based on a comparison of FWASIM output with the VISSIM output. Tests consider the important factors that may affect the traffic behavior for a given segment configuration. The obtained results show agreement between FWASIM and VISSIM outputs. Besides, the field data were used to validate FWASIM. Graphical and t-test methods were used to examine the results. The results are statistically significant which implies that the model provides reasonably accurate measures of effectiveness for the validated range of input data.

This paper establishes a general bifurcation structure of a car-following model with multiple time delays. By applying the Hopf theorem, analytical calculation and numerical simulation are carried ...

Safety is an important challenge in the development of autonomous vehicles (AVs). To ensure the safety of AVs, Intel and Mobileye have proposed a model called Responsibility-Sensitive Safety (RSS). Previous studies have shown that RSS has the potential to improve the safety performance of AVs, especially for partial autonomous driving algorithms. However, it is been shown that RSS leads to a considerable car-following distance, which has a negative effect on traffic efficiency. To improve the efficiency of RSS when applied to adaptive cruise control (ACC) systems, this paper proposes an improved strategy that involves triggering conditions of RSS. Two triggers of safety distance are defined according to different car-following assumptions. To test the performance of RSS models, original and improved RSS models are embedded in ACC based on model predictive control (MPC) algorithms. Car-following scenarios with a sudden deceleration of the lead vehicle (LV) at various time headways are simulated to evaluate the performance of improved RSS models. Results show that triggering RSS at the boundary of the safety distance calculated by considering the vehicle’s intentions is a better RSS model. This improved RSS model has a similar safety improvement effect to the ACC system as the original RSS in most risk scenarios and performs better in car-following efficiency. As the improved RSS model achieves a better trade-off between safety and efficiency, it can be used to improve the safety performance of partial autonomous driving algorithms like ACC on autonomous car-following maneuvers on expressways.

Understanding car-following behavior in different countries is essential for the design and development of autonomous driving and further development of active safety systems that can function well worldwide, in particular, in mixed traffic conditions. However, very few studies exist that compare car-following behaviors in different countries based on real driving data. This paper analyses the similarities and differences of drivers’ car-following behavior, in terms of time gap, gap distance, and time to collision (TTC), using both China and Sweden datasets from real road driving studies, in a bid to identify how these indicators affect drivers’ speed control in car-following situations. The results indicate that the highest frequency of gap distance is observed in the same value range in both datasets, while the highest frequency of time gap in the Sweden dataset is found at a lower value range than the corresponding value range in the China dataset. For both datasets, time gap is observed to be a more reliable indicator for car-following analysis than gap distance since it is less sensitive to speed variations. Furthermore, TTC in the low travel speed ranges ( v v > 90 km/h). Therefore, time gap is recommended as the safety indicator for car-following analysis in high-speed conditions, while a combination of time gap and TTC is recommended for low-speed conditions, especially on urban roads.

This paper analyzes the impact of uphill grades on the acceleration drivers choose to impose on their vehicles. Statistical inference is made based on the maximum likelihood estimation of a two-regime stochastic car-following model using Next Generation SIMulation (NGSIM) data. Previous models assume that the loss in acceleration on uphill grades is given by the effects of gravity. We find evidence that this is not the case for car drivers, who tend to overcome half of the gravitational effects by using more engine power. Truck drivers only compensate for 5% of the loss, possibly because of limited engine power. This indicates not only that current models are severely overestimating the operational impacts that uphill grades have on regular vehicles, but also underestimating their environmental impacts. We also find that car-following model parameters are significantly different among shoulder, median and middle lanes but more data is needed to understand clearly why this happens.

The expectation from automated vehicles has been increasing in recent years. Potential advantages for automated vehicles include driving safety, traffic efficiency, and positive environmental outcomes. To evaluate the extent to which automated vehicles are advantageous against conventional vehicles, this paper presents a simple car-following model, named the linked vehicle model, which reflects the capabilities of automated vehicles and the interaction between adjacent automated vehicles. The idea behind linked vehicle model is inspired by the collective motion of self-propelled particles in nature and the study of vehicle platooning from the Safe Road Trains for the Environment project. Moreover, the model is integrated into a traffic simulator in order to evaluate the performance of automated vehicles in an ideal circumstance. Comparisons with other representative car-following models are also provided, and they show that linked vehicle model achieves both optimal microscopic and macroscopic performance. Simulated results also show that linked vehicle model attains the best fuel economy and the lowest pollutant emissions.

Numerous car-following models have been developed since the 1950s. However, there still exist many traffic phenomena that cannot be demonstrated using the existing models. Therefore, this research proposed a new car-following model, the Asymmetric car-following (ACF) model based on the understanding of driver’s asymmetric behavior, which can explain complex traffic phenomena. We established the asymmetric car-following (ACF) rule under the vehicle’s safety constraints using eight parameters that indicate the driver’s characteristics and vehicle’s performances. To evaluate the ACF model, we performed the simulation for car-following pairs and conducted a comparison analysis with the existing models: Newell, Gipps, GM, and IDM. As a result, the proposed ACF model showed good fitness with the empirical trajectory and the apparent asymmetric behavior compared to others. For further investigation in the congested traffic stream, we simulated a group of vehicles by adding an error term to represent the driver’s unexpected behavior. The simulation showed growth, propagation, and dissipation of the stop-and-go traffic. These results proved that the ACF model has the strength to elaborate on various traffic phenomena, such as traffic hysteresis and stop-and-go traffic.

A large-scale naturalistic vehicle trajectory dataset from German highways, highD, was used to investigate the car-following behavior of individual drivers. These data include trajectories of 110,000 vehicles recorded for a duration of 16.5 h. Solving a nonlinear optimization problem, the intelligent driver model and the optimal velocity model with two leaders in interaction were calibrated by minimizing the deviations between the observed and simulated gaps when following the prescribed leading vehicle. The obtained calibration errors ranged between 5.2% and 6.9%, which were slightly lower than previous findings. This was explained by the shorter highD trajectories, predominantly free-flow traffic, and the good precision metrics of this dataset. The optimal velocity model with multivehicle anticipation resulted in lower calibration errors. This confirmed that natural drivers take into account several leading vehicles ahead. The ratio between interdriver and intradriver variability was investigated by performing global and platoon calibrations. Intradriver variation accounted for a larger portion of the calibration errors than interdriver variation. We analyzed the acceleration time-series of the natural highD and artificial drivers using simulations of two car-following models. A new cumulative measure, proportional to the energy of the follower’s position time-series curve, was calculated both for natural and modeled drivers. Human drivers had higher energy and demonstrated more acceleration fluctuations, sometimes behaving irrationally. In contrast, artificial drivers followed the logical rules incorporated in the model, resulting in a smoother acceleration profile. This led to less fuel consumption and gas emissions.

The increasing number of vehicles in cities brings new challenges to urban traffic management. Analyzing and modeling traffic is of great practical significance to urban intelligent traffic management. In this paper, the existing traffic simulation research is reviewed and summarized. Firstly, the crowd modeling and crowd animation are analyzed by referring to the idea of crowd simulation. Secondly, it compares and analyzes various existing car following technologies, and points out that animated traffic simulation is a hotspot in traffic simulation research. And then the concept of affective computing is integrated into the traffic simulation, considering the impacts of drivers’ emotion on vehicle driving and it is pointed out that the emotion-driven traffic flow is more authentic. Finally, combined with the status quo, the existing research drawbacks are analyzed, and the direction of future traffic simulation is pointed out.

Introduction: Aggressive driving has been associated as one of the causes for crashes, sometimes with very serious consequences. The objective of this study is to investigate the possibility of identifying aggressive driving in car-following situations on motorways by simple jerk metrics derived from naturalistic data. Method: We investigate two jerk metrics, one for large positive jerk and the other for large negative jerk, when drivers are operating the gas and brake pedal, respectively. Results: The results obtained from naturalistic data from five countries in Europe show that the drivers from different countries have a significantly different number of large positive and large negative jerks. Male drivers operate the vehicle with significantly larger number of negative jerks compared to female drivers. The validation of the jerk metrics in identifying aggressive driving is performed by tailgating (following a leading vehicle in a close proximity) and by a violator/non-violator categorization derived from self-reported questionnaires. Our study shows that the identification of aggressive driving could be reinforced by the number of large negative jerks, given that the drivers are tailgating, or by the number of large positive jerks, given that the drivers are categorized as violators. Practical applications: The possibility of understanding, classifying, and quantifying aggressive driving behavior and driving styles with higher risk for accidents can be used for the development of driver support and coaching programs that promote driver safety and are enabled by the vast collection of driving data from modern in-vehicle monitoring and smartphone technology.

This paper proposes a two-dimensional car-following model to tackle traffic flow problems where considering continuum lateral distances enables a simpler or more natural mathematical formulation compared to traditional car-following models. These problems include (i) the effects of lateral friction often observed in HOV lanes and diverge bottlenecks, (ii) the relaxation phenomenon at merge bottlenecks, (iii) the occurrence of accidents due to lane changing, and (iv) traffic models for autonomous vehicles (AVs). We conjecture that traditional car-following models, where the lateral dimension is discretized into lanes, struggle with these problems and one has to resort to ad-hoc rules conceived to directly achieve the desired effect, and that are difficult to validate. We argue that the distance maintained by drivers in order to avoid collisions in all directions plays a fundamental role in all these problems. To test this hypothesis, we propose a simple two-dimensional microscopic car-following model based on the social force paradigm, and build simulation experiments that reproduce these phenomena. These phenomena are reproduced as an indirect consequence of the model’s formulation, as opposed to ad-hoc rules, thus shedding light on their causes. A better understanding of the behavior of human drivers in the lateral dimension can be translated to improving autonomous driving algorithms so that they are human-friendly. In addition, since AV technology is proprietary, we argue that the proposed model should provide a good starting point for building AV traffic flow models when real data becomes available, as these data come from sensors that cover two-dimensional regions.

This study investigates the car-following behavior during braking at intersections and segments. Car-following events were extracted from a naturalistic driving dataset, mapped using ArcGIS, and analyzed to differentiate between the intersection- and segment-related events. The intersection-related events were identified according to an intersection influence area, which was estimated based on the stopping sight distance and the speed limit. Five behavioral measures were quantified based on exploring the probability density functions (PDF) for intersection- and segment-related events. The results showed that there were significant differences between the PDFs of the measures for both cases. Moreover, it was indicated that drivers tend to be more aggressive at intersections compared with segments. Thus, it is crucial to consider the driver’s location when investigating driver behavior. The quantified behavioral measures are a rich data source that can be used for car-following microscopic modeling, surrogate safety analysis, and driver assistance systems development.

The present research work provides a methodology of calibration of the vehicle-following behavior in VISSIM using vehicle trajectory data obtained under non-lane-based mixed traffic conditions. In ...

Autonomous vehicle (AV) technology is widely studied in both industrial and academic communities since it is regarded as a promising means for improving transportation safety and efficienc...

Forward collision warning (FCW) systems function by alerting drivers to upcoming hazards ahead and have been shown to help drivers respond more quickly under emergency situations. As FCW directly affects how vehicles interact longitudinally with one another, it may also influence car-following behavior such as reaction time, which has been little researched. To investigate these effects, driving data were collected from the Shanghai Naturalistic Driving Study. Five data-collecting vehicles were equipped with Mobileye® systems, which included an FCW function with headway display and warning system. Participants drove the vehicles for two months, with the Mobileye® system activated the second month only. From the 161,055 km of naturalistic driving data collected from 60 drivers, 3,000 car-following events were selected, and the effects of FCW on car-following headway and reaction time, and on the parameter values of the Gazis-Herman-Rothery (GHR) model were examined. Results showed that (1) drivers tended to maintain shorter headway with FCW enabled, while the proportion of time in short headways did not increase; (2) FCW reduced car-following reaction time when the lead vehicle was accelerating and when the relative speed between the lead and following vehicle was large; and (3) a reduction in the space headway exponent of the GHR was observed when FCW was enabled, indicating that drivers follow more closely with FCW because the system increases drivers’ sensitivity to changes in following gaps. Results of this study suggest that an FCW system with a headway monitoring function may increase traffic efficiency and stability without degrading safety.