Car-following Research Articles

A method for long car-following pair extraction and comprehensive data quality assessment: a case study using Zen Traffic Data

ABSTRACT This paper introduces a car-following (CF) extraction algorithm to address challenges in aerial-based trajectory data extraction. The algorithm, comprising four steps – vehicle grouping, elimination of false overtaking behavior, vehicle sorting, and CF pair matching – was applied to Zen Traffic Data, extracting 246 CF pairs. Three datasets were then generated: kilopost-based, geography-based, and velocity-based. A quality analysis revealed significant inconsistencies between data fields, with the geography-based dataset being least affected by high-frequency noise. The extracted CF data also demonstrated a more comprehensive driving regime than NGSIM, with complete driving regimes identified. Furthermore, the impact of data noise on CF model calibration and heterogeneity analysis was thoroughly assessed. This study enhances our understanding of trajectory data quality and highlights the richness of driving behavior information in Zen Traffic Data.

The effects of communication topology on mixed traffic flow: modelling and analysis

This paper aims to evaluate the combined impact of the communication topology (CT) and vehicular spatial distribution (SD) on mixed traffic of human-driven vehicles (HDVs) and connected and autonomous vehicles (CAVs). A generalized car-following (CF) model is presented to capture mixed traffic dynamics, integrating the effects of the CT, human reaction time, information delay and the optimal velocity changes with memory. Then, the linear stable condition of the model is derived using the perturbation method. Finally, extensive simulations are conducted under different CTs (i.e., multiple-predecessor leader following (MPLF) topology, predecessor following (PF) topology, and predecessor-leader following (PLF) topology) and six kinds of SDs. The results reveal that the MPLF topology demonstrates the best steady performance (i.e., stability) and dynamic performance (i.e., smoothness), and that the SD where CAVs are concentrated at the front of the traffic flow shows superior dynamic performance.

MetaFollower: Adaptable personalized autonomous car following

Car-following (CF) modeling, a fundamental component in microscopic traffic simulation, has attracted increasing research interest in recent decades. In this study, we propose an adaptable personalized car-following framework —– MetaFollower, by leveraging the power of meta-learning. Specifically, we first utilize Model-Agnostic Meta-Learning (MAML) to extract common driving knowledge from various CF events. Afterward, the pre-trained model can be fine-tuned on new drivers with only a few CF trajectories to achieve personalized CF adaptation. We additionally combine Long Short-Term Memory (LSTM) and Intelligent Driver Model (IDM) to reflect temporal heterogeneity with high interpretability. MetaFollower can accurately capture and simulate the intricate dynamics of car-following behavior while considering the unique driving styles of individual drivers. We demonstrate the versatility and adaptability of MetaFollower by showcasing its ability to adapt to new drivers with limited training data quickly. To evaluate the performance of MetaFollower, we conduct rigorous experiments comparing it with both data-driven and physics-based models. The results reveal that our proposed framework outperforms baseline models in predicting car-following behavior with higher accuracy and safety. To the best of our knowledge, this is the first car-following model aiming to achieve fast adaptation by considering both driver and temporal heterogeneity based on meta-learning.

A simulation‐based impact assessment of autonomous vehicles in urban networks

AbstractThe behavioural differences between autonomous vehicles (AVs) and human‐driven vehicles (HDVs) can significantly impact traffic efficiency, safety, and emissions. Simulation‐based impact assessments using microscopic traffic models often modify car‐following (CF) and lane‐changing (LC) configurations to differentiate AVs from HDVs. Typically, researchers adjust CF model parameters to replicate AV driving behaviour, but these assumptions can lead to varying conclusions on AV impacts. The scope of each study (e.g., freeways, highways, urban links, intersections) also influences the outcomes. This research conducts an impact assessment utilizing optimized AV driving behavior rather than assumptions on a city network level (Munich) using a simulation‐based platform. The particle swarm optimization (PSO) algorithm is used to calibrate the base model and run simulation experiments under various penetration rates (PRs) and demand scenarios. Results show significant safety improvements throughout the network under higher PRs, while lower PRs might lead to deteriorating safety. At 100% AV PR, the total number of conflicts decreased by around 25% compared to a fully HDV environment. Considering AVs' sensing capabilities, additional safety improvements are found in almost any AV PR. However, AVs might not improve traffic efficiency; in some cases, they may slightly increase average network travel time, though this change is minimal.

Calibrating Car-Following Models Using SUMO-in-the-Loop and Vehicle Trajectories From Roadside Radar

This paper presents an innovative calibration method for car-following (CF) models in the Simulation of Urban MObility (SUMO) using real-world trajectory data from a 1.5 km signalized urban corridor, captured by roadside radars. By applying a sophisticated track-level association and fusion methodology, the study extends trajectory analysis beyond individual radar fields of view. The enhanced data is then utilized to refine the Krauss, IDM, and W99 CF models within SUMO, addressing the literature gap by integrating SUMO into the calibration loop, thereby accommodating the simulator's integration scheme and any model adaptations. The research identifies that default SUMO models tend to exhibit shorter time headways compared to real-world data, with calibration effectively reducing this discrepancy. Moreover, the W99 model, despite its unrealistic acceleration profiles when calibrated without considering acceleration, most accurately captures the higher-end energy consumption distribution. Conversely, the IDM model, with its default parameters, provides the closest approximation to observed acceleration behaviors, highlighting the nuanced performance of CF models in traffic simulation and their implications for energy consumption estimation. Detailed results of optimized parameters for each CF model are provided in appendix in addition to distribution information that may be useful for other modelers to use directly or other datasets to be compared in the future (including expansion of the work to include vehicle classification).

Theory-data dual driven car following model in traffic flow mixed of AVs and HDVs

To model the car following (CF) behavior of mixed traffic flow composed of autonomous vehicles (AVs) and human-driving vehicles (HDVs) in the future, this paper calibrated the lower control algorithm of AVs and proposed a model named theory-data dual driven stochastically generative adversarial networks (TDS-GAN) to describe the CF behavior of HDVs based on experimental data from mixed traffic flow. Firstly, the experimental scenario and collected data were introduced. Secondly, two transfer functions for the lower control algorithm of AVs were compared. Then, to account for the stochasticity of HDVs, the idea of physics-informed deep learning (PIDL) was used to improve generative adversarial networks (GAN) and integrate it with two-dimensional intelligent driver model (2D-IDM). Finally, the effectiveness of the model was verified from both a micro prediction and a macro simulation perspective. The influence of AVs on the stability of mixed traffic flow at different Market Penetration Rate (MPR) was also observed. The results show that TDS-GAN can effectively describe the following behavior and stochasticity of HDVs. When combined with CF model of AVs, it can depict the evolution of mixed traffic flow more accurately. Additionally, AVs can improve traffic flow stability under different penetration rates, and the effect is more significant with higher MPR. However, the introduction of AVs may not necessarily be positive in terms of road capacity.

A unified risk field-based driving behavior model for car-following and lane-changing behaviors simulation

The modeling of driving behavior is pivotal for the accurate simulation of traffic scenarios and for providing human-like decision-making of autonomous driving systems. Car-following (CF) and lane-changing (LC) behaviors are continuous maneuvers within traffic flow, generally modeled separately in the literature. The coherence between these two behaviors may be ignored, leading to unrealistic behavioral simulations. Therefore, this paper establishes a risk field-based driving behavior model for two-dimensional motion, ensuring coherent modeling of CF and LC behaviors under a unified framework. First, a risk quantification method is developed to calculate the risk in two-dimensional scenarios, accounting for risk over the preview time. A cubic polynomial is applied to generate path curves that align with vehicle dynamics. Second, the enhanced behavior model primarily comprises two integral components: path and trajectory planning. These two components aim to identify the path or trajectory that maximizes the benefit while meeting the desired risk. Third, the maximum acceptable risk, representing a higher risk than the desired risk, is defined to facilitate path adjustment and avoid frequent path adjustment. Finally, the proposed model is proved through comparisons with existing models using driving data. Several cases are employed for further analysis to show the model's rationality and potential in various aspects. This study develops the previous risk field-based behavior model from one-dimensional to two-dimensional scenarios, furnishes a unified framework for elucidating driving behavior in various scenarios, and contributes to the progress of behavior modeling.

Implications of stop-and-go traffic on training learning-based car-following control

Learning-based car-following control (LCC) of connected and autonomous vehicles (CAVs) is gaining significant attention with the advancement of computing power and data accessibility. While the flexibility and large model capacity of model-free architecture enable LCC to potentially outperform the model-based car-following (CF) model in improving traffic efficiency and mitigating congestion, the generalizability of LCC for traffic conditions different from the training environment/dataset is not well-understood. This study seeks to explore the impact of stop-and-go traffic in the training dataset on the generalizability of LCC. It uses the characteristics of lead vehicle trajectories to describe stop-and-go traffic, and links the theory of identifiability (i.e., obtaining a unique parameter estimation result using sensor measurements) to the generalizability of behavior cloning (BC) and policy-based deep reinforcement learning (DRL). Correspondingly, the study shows theoretically that: (i) stop-and-go traffic can enable the property of identifiability and enhance the control performance of BC-based LCC in different traffic conditions; (ii) stop-and-go traffic is not necessary for DRL-based LCC to generalize to different traffic conditions; (iii) DRL-based LCC trained with only constant-speed lead vehicle trajectories (not sufficient to ensure identifiability) can be generalized to different traffic conditions; and (iv) stop-and-go traffic increases variance in the training dataset, which improves the convergence of parameter estimation while negatively impacting the convergence of DRL to the optimal control policy. Numerical experiments validate the above findings, illustrating that BC-based LCC entails comprehensive training datasets for generalizing to different traffic conditions, while DRL-based LCC can achieve generalization with simple free-flow traffic training environments. This further suggests DRL as a more promising and cost-effective LCC approach to reduce operational costs, mitigate traffic congestion, and enhance safety and mobility, which can accelerate the deployment and acceptance of CAVs.

Deep Reinforcement Learning Car-Following Control Based on Multivehicle Motion Prediction

Reinforcement learning (RL)–based car-following (CF) control strategies have attracted significant attention in academia, emerging as a prominent research topic in recent years. Most of these control strategies focus solely on the motion status of the immediately preceding vehicle. However, with the development of vehicle-to-vehicle (V2V) communication technologies, intelligent vehicles such as connected autonomous vehicles (CAVs) can gather information about surrounding vehicles. Therefore, this study proposes an RL-based CF control strategy that takes multivehicle scenarios into account. First, the trajectories of two preceding vehicles and one following vehicle relative to the subject vehicle (SV) are extracted from a highD dataset to construct the environment. Then the twin-delayed deep deterministic policy gradient (TD3) algorithm is implemented as the control strategy for the agent. Furthermore, a sequence-to-sequence (seq2seq) module is developed to predict the uncertain motion statuses of surrounding vehicles. Once integrated into the RL framework, this module enables the agent to account for dynamic changes in the traffic environment, enhancing its robustness. Finally, the performance of the CF control strategy is validated both in the highD dataset and in two traffic perturbation scenarios. In the highD dataset, the TD3-based prediction CF control strategy outperforms standard RL algorithms in terms of convergence speed and rewards. Its performance also surpasses that of human drivers in safety, efficiency, comfort, and fuel consumption. In traffic perturbation scenarios, the performance of the proposed CF control strategy is compared with the model predictive controller (MPC). The results show that the TD3-based prediction CF control strategy effectively mitigates undesired traffic waves caused by the perturbations from the head vehicle. Simultaneously, it maintains the desired traffic state and consistently ensures a stable and efficient traffic flow.

On the string stability of neural network-based car-following models: A generic analysis framework

String stability plays a crucial role in regulating traffic flow, as traffic oscillation can be triggered by string instability in the car-following (CF) behavior. Although studies over the past decades have provided various methods for analyzing string stability of analytical CF models, no studies have focused on neural network (NN) based CF models despite the fact that these models have exhibited remarkable performance in learning realistic driving behavior in the recent literature. This paper fills this gap by proposing a generic theoretical framework for analyzing the string stability of NN-based CF models through an Estimation-Approximation-Derivation-Calculation process (referred to as EADC framework). Within the framework, we first estimate the steady states of NN-based CF models by solving the corresponding optimization model and obtain the smooth approximation of the NN-based models for linearization, based on which the transfer function is constructed. We then derive the general stability criteria for two commonly used classes of NN in CF modeling, i.e., feedforward NN-based CF models with basic input and recurrent NN-based CF models with multi-step historical information. The string stability is thus obtained by calculating the partial derivatives through the automatic differentiation method. As two case studies, we apply the proposed stability analysis framework on two typical NN-based CF models, the Mo-MLP model (Mo et al., 2021) and the Huang-LSTM model (Huang et al., 2018), and obtained the complete consistency between the theoretical results and the simulation results for both models, which demonstrates the soundness of the proposed EADC framework. Moreover, we discuss the applicability of the proposed EADC stability analysis framework in the emerging era of connected and autonomous vehicles and artificial intelligence.

A Car-Following Model Considering Missing Data Based on TransGAN Networks

Car-following behavior is closely related to the longitudinal control of the vehicle, affecting the safety of the vehicle and traffic flow stability. In order to interact with the preceding vehicle, the target vehicle usually collects the driving data of the preceding vehicle. However, data acquisition devices often face malfunctions caused by various unpredictable disruptions, resulting in missing value problems. This may cause the target vehicle to make wrong control decisions. Given this situation, a new car-following(CF) model considering missing data based on Transformer-Generative Adversarial Networks (TransGAN) is proposed. Firstly, Transformer Network with multi-head attention is used to deeply extract the potential features from incomplete vehicle state data, which can filter important information from the input and focus on these, while capturing long distance dependencies. Secondly, a Generative Adversarial Network is constructed. The Generator generates the future multi-step control states of the target vehicle based on the features extracted by Transformer Network. The Discriminator with a fully connected network is applied to simultaneously ensure the generation accuracy. Finally, our proposed model was trained and tested on a publicly available NGSIM I-80 dataset. Compared with other existing advanced works, our model can fit the actual control states of the target vehicle with higher accuracy under different data missing rates of the preceding vehicle, which demonstrates that the proposed method effectively improves the robustness of vehicle longitudinal car-following control under missing data.

Applicability of Transfer Function to Disturbance Growth Estimation and Traffic Stability

In linear stability analysis, some unrealistic assumptions (e.g., small and long wavelength disturbance) are made for the derivation of the transfer function, which may result in an inaccurate estimation of disturbance growth rate and incorrect judgment of traffic stability. To gain more insights into the relaxation of assumptions on disturbance, this paper studies the effect of disturbance magnitude and frequency on the accuracy of disturbance growth rate estimation and the applicability of transfer function to traffic stability. The disturbance growth rates of several car-following (CF) models, including both linear and nonlinear CF models, are examined through two-vehicle simulations. The calibrated growth rate is then verified with a simulation involving a long vehicle platoon. The simulation results demonstrate that the calibrated growth rate from a two-vehicle CF pair is applicable to predict the disturbance magnitude of a long platoon. In addition, a concave growth pattern is observed, for which a novel explanation is provided based on the traditional deterministic CF model for the first time.

Integrated self-consistent macro-micro traffic flow modeling and calibration framework based on trajectory data

Calibrating microscopic car-following (CF) models is crucial in traffic flow theory as it allows for accurate reproduction and investigation of traffic behavior and phenomena. Typically, the calibration procedure is a complicated, non-convex optimization issue. When the traffic state is in equilibrium, the macroscopic flow model can be derived analytically from the corresponding CF model. In contrast to the microscopic CF model, calibrated based on trajectory data, the macroscopic representation of the fundamental diagram (FD) primarily adopts loop detector data for calibration. The different calibration approaches at the macro- and microscopic levels may lead to misaligned parameters with identical practical meanings in both macro- and micro-traffic models. This inconsistency arises from the difference between the parameter calibration processes used in macro- and microscopic traffic flow models. Hence, this study proposes an integrated multiresolution traffic flow modeling framework using the same trajectory data for parameter calibration based on the self-consistency concept. This framework incorporates multiple objective functions in the macro- and micro-dimensions. To expeditiously execute the proposed framework, an improved metaheuristic multi-objective optimization algorithm is presented that employs multiple enhancement strategies. Additionally, a deep learning technique based on attention mechanisms was used to extract stationary-state traffic data for the macroscopic calibration process, instead of directly using the entire aggregated data. We conducted experiments using real-world and synthetic trajectory data to validate our self-consistent calibration framework.

Modified Gipps model: a collision-free car following model

Car following (CF) models are used in microscopic traffic simulation tools to help assess the effects of a new road design or to assess the effect of change in traffic flow. In 1981, Gipps developed a collision avoidance CF model using the Newtonian laws of motion to describe the motion of each vehicle in a stream of traffic. It is one of the most widely used CF models in both research and practice. Although it is claimed that the Gipps model produces collision-free results, the model produces a collision when the intention of the following vehicle is to brake harder than the perceived deceleration of lead vehicle. For the ease of simulations, a traffic simulation tool is expected to not show unrealistic crashes. This study was carried out to make the Gipps model collision-free in all conditions. It first highlights the conditions where the original Gipps model produces a collision. Then the study derives an equation for a collision-free Gipps CF model. This modified Gipps CF model produces collision-free results that always maintain a safe spacing with the lead vehicle.

Stability and extension of a car-following model for human-driven connected vehicles

Despite the prosperous development of connected vehicle (CV) and its car-following (CF) models, human-driven CV and its CF properties are rarely investigated. This paper studies the stability characteristics of and then extends a recently-developed CF model of human-driven CV which incorporates human factors (CV-CF hereafter) by considering two levels of driver compliance, i.e., low compliance and high compliance. First, we investigate the stability of the CV-CF model, and validate the results with the simulation experiments. We then assess CV’s impact on the mixed traffic flow by deriving a stability criterion of heterogeneous traffic with the Laplace transform based method and analysing the influence of different levels of connectivity and their penetration rates. Furthermore, we extend the CV-CF model by considering two important additional human factors, i.e., time delay and estimation error, and evaluate different human factors’ impact on the stability and oscillation characteristics of the CV-CF model. The results reveal that the connected environment indeed promotes the CF stability and alleviates traffic congestion, and that higher compliance to the information provided is generally more beneficial to the stability of traffic flow, except the situation with a large time delay.

A Method of Identifying Personalized Car-Following Characteristics for Adaptive Cruise Control System

Although it’s very common nowadays to experience onboard adaptive cruise control systems (ACCS), most of them are initialized with fixed parameters and little consideration is taken to adapting to different drivers. In order to improve the ride comfort and operation familiarity of drivers when using the ACCS, driver’s personalized car-following (CF) characteristics should be considered in ACCS developments. In this paper, an ACCS framework with CF characteristics identification and application is designed, where the desired safety margin (DSM) model is utilized to describe driver’s CF behaviors. In view of the limited performance of on-board computers, a simple method for CF characteristics identification from the sight of driver’s physiology and psychology mechanism is proposed, which tries to estimate DSM model parameters on the basis of CF data. With the proposed method, the CF data under manual driving are collected and filtered for the identification of drivers’ response time, upper and lower limits of DSM and sensitivity factors for acceleration and deceleration iteratively. And the latest iteration results of DSM model parameters would then be applied into the ACCS when it’s enabled. Several naturalistic driving tests are adopted for the verification whose results support that, the proposed method could perform well in identifying drivers’ CF characteristics online by collecting and analyzing manual driving data with a much smaller resource consumption when compared with genetic algorithm (GA). Besides, with the increase of CF data volume, the identified results are also becoming more and more stable and robust. Therefore, the proposed ACCS framework is able to imitate different drivers.

Comparing Measured Driver Behavior Distributions to Results from Car-Following Models using SUMO and Real-World Vehicle Trajectories from Radar

In this study, the physical principles governing car-following (CF) behavior and their impact on traffic flow at signalized intersections are investigated. High temporal-resolution radar data is used to provide valuable insights into actual CF behavior, including acceleration, deceleration, and time headway distribution. Demand-calibrated SUMO simulations are run using empirical CF parameter distributions, and three CF models are evaluated: IDM, EIDM, and Krauss. By emulating radar data in SUMO and processing simulated vehicle traces, discrepancies between empirical and simulated parameter distributions are identified. Further analysis includes comparisons with default SUMO CF model parameters. The findings reveal that measured accelerations differ from CF model parameter accelerations and using the empirical value ($\mu = 0.89m/s^2$) leads to unrealistic simulations that fail volume-based calibration. Default parameters for all three models reasonably approximate the mean and median of measured parameters, but fail to capture the true distribution shape, partly due to homogeneity when using default parameters. The results show that it is more effective to simulate with the default parameters provided by SUMO rather than using measurements of real-world distributions without additional calibration. Future work will investigate closing the loop between the measured real-world and SUMO distributions using traditional calibration tactics, as well as assess the impact of calibrated vs. default CF parameters on simulation outputs like fuel consumption.

Microscopic Modelling of Car-Following Behaviour: Developments and Future Directions

Review Microscopic Modelling of Car-Following Behaviour: Developments and Future Directions Yinglong He 1, * , Quan Zhou 2, * , Chongming Wang 3, Ji Li 2, Bin Shuai 2, Lei Lei 4, and Hongming Xu 2 1 School of Mechanical Engineering Sciences, University of Surrey, Guildford GU2 7XH, UK 2 Department of Mechanical Engineering, University of Birmingham, Birmingham B15 2TT, UK 3 Institutes for Future Transport and Cities, Coventry University, Coventry CV1 5FB, UK 4 College of Information Science and Engineering, Jiaxing University, Jiaxing 314001, China * Correspondence: yinglong.he@surrey.ac.uk (Y.H.); q.zhou@bham.ac.uk (Q.Z.) Received: 17 April 2023 Accepted: 21 June 2023 Published: 27 June 2023 Abstract: The study of driving behaviour has become increasingly important in the development of transport and vehicle technologies. Microscopic traffic models simulate individual driver behaviour to understand and predict traffic flow. One of the key components in microscopic simulation is the car-following (CF) model, which describes the behaviour of vehicles in terms of how they follow the vehicle in front of them. Some excellent reviews of CF models are available, however, to the best of the authors’ knowledge, none of them provides a comprehensive analysis that covers and compares different model categories including kinematics-based, dynamics-based, psychological-based, and learning-based. This paper, therefore, provides an overview of the developments and future directions of CF models, encompassing all the previously mentioned categories. It first introduces the fundamental concepts of traffic models, in particular CF models. It then reviews the progress of CF models, which are classified into the above four categories. The advantages and limitations of existing CF models are discussed. The paper further identifies several research directions for future work, including the integration of emerging vehicle technologies, the incorporation of real-world traffic data, and the calibration and validation of model parameters. It concludes by emphasizing the importance of interdisciplinary collaboration and the need for further research to improve the accuracy and practicality of CF models.

Data-driven analysis for disturbance amplification in car-following behavior of automated vehicles

This paper presents a data-driven framework to quantitatively analyze the disturbance amplification behavior of automated vehicles in car-following (CF). The data-driven framework can be applied to unknown CF controllers based on the concept of empirical frequency response function (FRF). Specifically, a well-known signal processing method, Welch's method, together with a short time Fourier transformation is developed to extract the empirical transfer functions from vehicle trajectories. The method is first developed assuming a generic linear controller with time-invariant CF control features (e.g., control gains) and later extended to capture time-variant features. The proposed methods are evaluated for estimation consistencies via synthetic data-based simulations. The evaluation includes the performances of the linear approximation accuracy for a linear time-invariant controller, a nonlinear controller, and a linear time-variant controller. Results indicate that our framework can provide reasonably consistent results as theoretical ones in terms of disturbance amplification. Further it can perform better than a linear theoretical analysis of disturbance amplification, particularly when nonlinearity in CF behavior is present. The methods are applied to existing field data collected from vehicles with adaptive cruise control (ACC) on the market. Findings reveal that all tested vehicles tend to amplify disturbances, particularly in low frequency (< 0.5 Hz). Further, the results demonstrate that these ACC vehicles exhibit time-varying features in terms of disturbance amplification ratio depending on the leading vehicle trajectories.

Autonomous Vehicle’s Impact on Traffic: Empirical Evidence From Waymo Open Dataset and Implications From Modelling

Previous empirical behavior analysis on Autonomous Vehicles (AV) mainly focused on vehicles with Adaptive Cruise Control (ACC) system due to the lack of high-level AV dataset. Recently released SAE Level-4 AV datasets such as the Waymo Open Dataset provide great opportunities to evaluate their behavioral impact on traffic flow. In this study, we aim to characterize the empirical Car Following (CF) behaviors of the Waymo autonomous vehicle and compare its feature with human-driven Vehicles (HV), and capture such behavioral differences using the IDM CF model. Our main findings include: (a) AV is much safer than HV, based on our analysis using surrogate safety measures, as time headways and jam spacings of the AV are significantly larger than HV; (b) the response time of AV is also significantly larger than that of HV in response to various types of stimuli; (c) despite the short length of trajectories in the Waymo Open Dataset, we have confirmed that these trajectories are suitable for calibrating some of the IDM parameters; and the calibration results of IDM are consistent with our empirical analysis. Moreover, the modelling results, reveal that the proportion of string unstable behavior of AV is less than that of HV; and (d) for HV, there is generally no significant difference between following AV and following HV except a smaller jam spacing when following AV. Overall, we conclude that currently AV behaves in a conservative way to ensure its safety at the cost of traffic efficiency.