Validating car-following behaviors of autonomous vehicles by structural assessment with empirical data

Abstract

Many researchers argue that the modular perception–planning–control design of autonomous vehicles (AVs) inevitably introduces delays, yet how such delays influence car-following behaviors and what their magnitudes are remain uncertain due to limited empirical data and ineffective calibration methods. Using large-scale naturalistic autonomous vehicle trajectory data, we develop an optimization-based calibration framework that jointly estimates feedback delay in perception/decision and parasitic lag in actuation, and accommodates both linear constant-time-headway and nonlinear intelligent-driver-model controllers. The method combines Bayesian optimization (BO) for globally searching delay parameters with simultaneous perturbation stochastic approximation (SPSA) for efficiently tuning controller parameters. Results show that incorporating both delays substantially improves model accuracy, reducing the root mean square error (RMSE) by up to an order of magnitude compared with models without delays, while linear controllers reproduce AV trajectories more faithfully and local calibration proves more accurate for short horizons whereas global calibration is more reliable for long horizons.

Keywords

Autonomous vehicles / Car-following / Delay modeling / Parameter calibration

References

[1] Antoniou C, Azevedo CL, Lu L, et al. W-spsa in practice: Approximation of weight matrices and calibration of traffic simulation models. Transp Res Part C Emerg Technol, 2015, 59: 129-146.

[2] Brunner JS, Makridis MA, Kouvelas A. Comparing the observable response times of acc and cacc systems. IEEE Trans Intell Transp Syst, 2022, 23101929919308.

[3] Irshayyid A, Chen J, Xiong G. A review on reinforcement learning-based highway autonomous vehicle control. Green Energy Intell Transp, 2024, 34100156.

[4] Jin IG, Avedisov SS, He CR, et al. Experimental validation of connected automated vehicle design among human-driven vehicles. Transp Res Part C Emerg Technol, 2018, 91: 335-352.

[5] Li L, Chen XM, Zhang L. A global optimization algorithm for trajectory data based car-following model calibration. Transp Res Part C Emerg Technol, 2016, 68: 311-332.

[6] Li T, Chen D, Zhou H, et al. Car-following behavior characteristics of adaptive cruise control vehicles based on empirical experiments. Transp Res Part B Methodol, 2021, 147: 67-91.

[7] Li X. Trade-off between safety, mobility and stability in automated vehicle following control: An analytical method. Transp Res Part B Methodol, 2022, 166: 1-18.

[8] Li X, Ouyang Y. Characterization of traffic oscillation propagation under nonlinear car-following laws. Procedia Soc Behav Sci, 2011, 17: 663-682.

[9] Li X, Wang X, Ouyang Y. Prediction and field validation of traffic oscillation propagation under nonlinear car-following laws. Transp Res Part B Methodol, 2012, 463409423.

[10] Li Y, He H, Chen Y, et al. A cloud-based eco-driving solution for autonomous hybrid electric bus rapid transit in cooperative vehicle-infrastructure systems: A dynamic programming approach. Green Energy Intell Transp, 2023, 26100122.

[11] Ma K, Wang H, Ruan T. Analysis of road capacity and pollutant emissions: Impacts of connected and automated vehicle platoons on traffic flow. Physica A, 2021, 583: 126301.

[12] Ma K, Wang H, Zuo Z, et al. String stability of automated vehicles based on experimental analysis of feedback delay and parasitic lag. Transp Res Part C Emerg Technol, 2022, 145: 103927.

[13] Ma K, Shi H, Ma C, et al. Development, calibration, and validation of a novel nonlinear car-following model: Multivariate piecewise linear approach for adaptive cruise control vehicles. Transp Res Part E Logist Transp Rev, 2025, 199: 104031.

[14] Ma K, Zhou H, Liang Z, et al. Automated vehicle microscopic energy consumption study (av-micro): Data collection and model development. Energy, 2025, 320: 135096.

[15] Makridis M, Mattas K, Ciuffo B. Response time and time headway of an adaptive cruise control: an empirical characterization and potential impacts on road capacity. IEEE Trans Intell Transp Syst, 2019, 21416771686.

[16] Makridis M, Mattas K, Ciuffo B, et al. Empirical study on the properties of adaptive cruise control systems and their impact on traffic flow and string stability. Transp Res Rec, 2020, 26744471484.

[17] Makridis M, Mattas K, Anesiadou A. Openacc. an open database of car-following experiments to study the properties of commercial ACC systems. Transp Res Part C Emerg Technol, 2021, 125: 103047.

[18] Milanés V, Shladover SE. Modeling cooperative and autonomous adaptive cruise control dynamic responses using experimental data. Transp Res Part C Emerg Technol, 2014, 48: 285-300.

[19] Pariota L, Bifulco GN, Brackstone M. A linear dynamic model for driving behavior in car following. Transp Sci, 2016, 50310321042.

[20] Poinsignon F, Chen L, Jiang S, et al. Autonomous vehicle fleets for public transport: scenarios and comparisons. Green Energy Intell Transp, 2022, 13100019.

[21] Punzo V, Simonelli F. Analysis and comparison of microscopic traffic flow models with real traffic microscopic data. Transp Res Rec, 2005, 193415363.

[22] Sha D, Ozbay K, Ding Y. Applying bayesian optimization for calibration of transportation simulation models. Transp Res Rec, 2020, 267410215228.

[23] Shah D, Lee C, Kim YH (2025) Modified gipps model: A collision-free car following model. J Intell Transp Syst 29(1):18–31. https://doi.org/10.1080/15472450.2023.2289149

[24] Shi X, Li X. Empirical study on car-following characteristics of commercial automated vehicles with different headway settings. Transp Res Part C Emerg Technol, 2021, 128: 103134.

[25] Sugiyama Y. Optimal velocity model for traffic flow. Comput Phys Commun, 1999, 121: 399-401.

[26] Treiber M, Kesting A. Microscopic calibration and validation of car-following models—a systematic approach. Procedia Soc Behav Sci, 2013, 80: 922-939.

[27] Wang M. Infrastructure assisted adaptive driving to stabilise heterogeneous vehicle strings. Transp Res Part C Emerg Technol, 2018, 91: 276-295.

[28] Wang Y, Liu C (2021) Dynamic integral sliding mode for vehicle platoon control with constant time headway policy. In: 2021 IEEE International Conference on Robotics and Biomimetics (ROBIO). IEEE, pp 1479–1484. https://doi.org/10.1109/ROBIO54168.2021.9634813

[29] Xiao L, Gao F. Practical string stability of platoon of adaptive cruise control vehicles. IEEE Trans Intell Transp Syst, 2011, 12411841194.

[30] Zhang C, Sun L (2024) Bayesian calibration of the intelligent driver model. IEEE Trans Intell Transp Syst 25(8):9308–9320. https://doi.org/10.1109/TITS.2024.3367820

[31] Zhang C, Osorio C, Flötteröd G. Efficient calibration techniques for large-scale traffic simulators. Transp Res Part B Methodol, 2017, 97: 214-239.

[32] Zhang Y, Ma K, Xu Z, et al. A modeling methodology for car-following behaviors of automated vehicles: Trade-off between stability and mobility. Transp Res Part B Methodol, 2025, 200: 103316.

[33] Zhou H, Ma K, Li X (2024) A review on trajectory datasets on advanced driver assistance system equipped-vehicles. In: 2024 IEEE Intelligent Vehicles Symposium (IV). IEEE, pp 1947–1952. https://doi.org/10.1109/IV55156.2024.10588808

Rights & permissions

The Author(s)