Car-following model; Neural networks; Optimal state relationships; Traffic flow dynamics; Behavioral assumption; Car-following behavior; Car-following modeling; Network-based; Neural-networks; Optimal state; Optimal state relationship; SciNetS; Symbolic regression; Civil and Structural Engineering; Automotive Engineering; Transportation; Management Science and Operations Research
Résumé :
[en] Mathematical models describing the dynamics of traffic flow have become increasingly popular as tools supporting the analysis and evaluation of traffic systems. This paper focuses on microscopic simulation tools, specifically those employing ordinary differential equations (ODEs). In general, most ODEs-based traffic models (i.e., car-following models or CFMs for short) require prior behavioral assumptions, that is, the optimal traffic state relationships. These assumptions vary widely across traffic scenarios, posing limitations. To overcome this hurdle and enhance CFMs’ practicability, this paper proposes a novel research paradigm—artificial intelligence (AI) for (traffic) physics or AI-driven traffic flow theory, to explore the mechanisms of car-following behaviors. The proposed neural network (SciNet)-based architecture for symbolic regression, called SciNet-CFM, can provide scientific hypotheses for the modeling of car-following behaviors from the AI perspective, thus relaxing the prior behavioral assumptions in current traffic theory. Specifically, symbolic regression is used to generate a tractable mathematical expression for CFM discovery, rather than the unexplained connection structure of traditional neural networks. The numerical and empirical experiments show that the SciNet-CFM has the potential to uncover the hidden properties of the observed microscopic traffic flow dynamics. The comparisons with classical and state-of-the-art models demonstrate a better performance of the proposed SciNet-CFM over traditional physics-based, data-driven, and hybrid models.
Disciplines :
Ingénierie, informatique & technologie: Multidisciplinaire, généralités & autres
Auteur, co-auteur :
Li, Tenglong ; College of Intelligent Manufacturing and Electrical Engineering, Nanyang Normal University, Nanyang, China
Ngoduy, Dong ; Institute of Transport Studies, Monash University, Clayton VIC, Australia
Lee, Seunghyeon ; Department of Transportation Engineering, University of Seoul, Seoul, South Korea
Pu, Ziyuan ; School of Transportation, Southeast University, Nanjing, China ; School of Engineering, Monash University, Selangor, Bandar Sunway, Malaysia
VITI, Francesco ; University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE)
Co-auteurs externes :
yes
Langue du document :
Anglais
Titre :
Discovering the optimal relationship hypothesis of car-following behaviors with neural network-based symbolic regression
Date de publication/diffusion :
2025
Titre du périodique :
Transportation Research. Part C, Emerging Technologies
This work was supported by the Key Scientific Research Project of Colleges and Universities in Henan Province of China (Grant No. 23A580001 ), National Natural Science Foundation of China (Grant No. 52172380 ), Key Scientific and Technological Research Projects in Henan Province (Grant Nos. 222102320369 and 232102211047) , Special Project for High-quality Professionals of Nanyang Normal University (Grant No. 2022ZX001 ), and National Natural Science Foundation Incubation Project of Nanyang Normal University (Grant No. 2023PY025 ).
Bando, M., Hasebe, K., Nakanishi, K., Nakayama, A., Shibata, A., Sugiyama, Y., Phenomenological study of dynamical model of traffic flow. J. Phys. I France 5:11 (1995), 1389–1399, 10.1051/jp1:1995206.
Bando, M., Hasebe, K., Nakayama, A., Shibata, A., Sugiyama, Y., Dynamical model of traffic congestion and numerical simulation. Phys. Rev. E, 51(2), 1995, 1035, 10.1103/PhysRevE.51.1035.
Cui, Z., Wang, X., Ci, Y., Yang, C., Yao, J., Modeling and analysis of car-following models incorporating multiple lead vehicles and acceleration information in heterogeneous traffic flow. Phys. A, 630, 2023, 129259, 10.1016/j.physa.2023.129259.
Davies, A., Veličković, P., Buesing, L., Blackwell, S., Zheng, D., Tomašev, N., Tanburn, R., Battaglia, P., Blundell, C., Juhász, A., et al. Advancing mathematics by guiding human intuition with AI. Nature 600:7887 (2021), 70–74, 10.1038/s41586-021-04086-x.
Geng, M., Li, J., Xia, Y., Chen, X.M., A physics-informed transformer model for vehicle trajectory prediction on highways. Transport. Res. Part C: Emerg. Technol., 154, 2023, 104272, 10.1016/j.trc.2023.104272.
Gunter, G., Gloudemans, D., Stern, R.E., et al. Are commercially implemented adaptive cruise control systems string stable?. IEEE Trans. Intell. Transp. Syst. 22:11 (2020), 6992–7003, 10.1109/TITS.2020.3000682.
Hasani, R., Lechner, M., Amini, A., Liebenwein, L., Ray, A., Tschaikowski, M., Teschl, G., Rus, D., Closed-form continuous-time neural networks. Nat. Mach. Intell. 4:11 (2022), 992–1003.
Hou, L., Pei, Y., He, Q., A car following model in the context of heterogeneous traffic flow involving multilane following behavior. Phys. A, 632, 2023, 129307, 10.1016/j.physa.2023.129307.
Iten, R., Artificial Intelligence for Scientific Discoveries: Extracting Physical Concepts from Experimental Data Using Deep Learning. 2023, Springer Nature, 10.1007/978-3-031-27019-2.
Iten, R., Metger, T., Wilming, H., Del Rio, L., Renner, R., Discovering physical concepts with neural networks. Phys. Rev. Lett., 124(1), 2020, 010508, 10.1103/PhysRevLett.124.010508.
Jiang, R., Wu, Q., Zhu, Z., Full velocity difference model for a car-following theory. Phys. Rev. E, 64(1), 2001, 017101, 10.1103/PhysRevE.64.017101.
Kang, Y., Kim, G., Jeong, S., Sohn, K., Trajectory-based embedding for random coefficients of a theory-based car-following model. Transport. Res. Part C: Emerg. Technol., 152, 2023, 104183, 10.1016/j.trc.2023.104183.
Lee, S., Ngoduy, D., Keyvan-Ekbatani, M., Integrated deep learning and stochastic car-following model for traffic dynamics on multi-lane freeways. Transport. Res. Part C: Emerg. Technol. 106 (2019), 360–377, 10.1016/j.trc.2019.07.023.
Lee, S., Xie, K., Ngoduy, D., Keyvan-Ekbatani, M., An advanced deep learning approach to real-time estimation of lane-based queue lengths at a signalized junction. Transport. Res. Part C: Emerg. Technol. 109 (2019), 117–136.
Lemos, P., Jeffrey, N., Cranmer, M., Ho, S., Battaglia, P., Rediscovering orbital mechanics with machine learning. Mach. Learn.: Sci. Technol., 4(4), 2023, 045002, 10.1088/2632-2153/acfa63.
Li, T., Ngoduy, D., Hui, F., Zhao, X., A car-following model to assess the impact of V2V messages on traffic dynamics. Transport. B: Transp. Dyn. 8:1 (2020), 150–165, 10.1080/21680566.2020.1728591.
Li, T., Ngoduy, D., Lee, S., Pu, Z., Capturing the equilibrium traffic state hypothesis of car-following models with artificial neural network. 2023 8th International Conference on Models and Technologies for Intelligent Transportation Systems, MT-ITS, 2023, 1–6, 10.1109/MT-ITS56129.2023.10241676.
Li, T., Ngoduy, D., Zhao, X., Hopf bifurcation analysis of mixed traffic and its implications for connected and autonomous vehicles. IEEE Trans. Intell. Transp. Syst. 24:6 (2023), 6542–6557, 10.1109/TITS.2023.3242826.
Li, D., Okhrin, O., Modified DDPG car-following model with a real-world human driving experience with CARLA simulator. Transport. Res. Part C: Emerg. Technol., 147, 2023, 103987, 10.1016/j.trc.2022.103987.
Liu, Z., He, K., A decade's battle on dataset bias: are we there yet?. 2024 arXiv preprint arXiv:2403.08632.
Mo, Z., Shi, R., Di, X., A physics-informed deep learning paradigm for car-following models. Transport. Res. Part C: Emerg. Technol., 130, 2021, 103240, 10.1016/j.trc.2021.103240.
Newell, G.F., Nonlinear effects in the dynamics of car following. Oper. Res. 9:2 (1961), 209–229, 10.1126/10.1287/opre.9.2.209.
Ngoduy, D., Noise-induced instability of a class of stochastic higher order continuum traffic models. Transp. Res. Part B: Methodol. 150 (2021), 260–278, 10.1016/j.trb.2021.06.013.
Ngoduy, D., Lee, S., Treiber, M., Keyvan-Ekbatani, M., Vu, H., Langevin method for a continuous stochastic car-following model and its stability conditions. Transport. Res. Part C: Emerg. Technol. 105 (2019), 599–610, 10.1016/j.trc.2019.06.005.
Pan, W., Zhang, J., Tian, J., Cui, F., Wang, T., Analysis of car-following behaviors based on data-driven and theory-driven car-following models: Heterogeneity and asymmetry. Phys. A, 632, 2023, 129324, 10.1016/j.physa.2023.129324.
Punzo, V., Zheng, Z., Montanino, M., About calibration of car-following dynamics of automated and human-driven vehicles: Methodology, guidelines and codes. Transport. Res. Part C: Emerg. Technol., 128, 2021, 103165, 10.1016/j.trc.2021.103165.
Schmidt, M., Lipson, H., Distilling free-form natural laws from experimental data. Science 324 (2009), 81–85, 10.1126/science.1165893.
Shiomi, Y., Li, G., Knoop, V.L., Analysis of stochasticity and heterogeneity of car-following behavior based on data-driven modeling. Transport. Res. Rec. 2677:12 (2023), 604–619, 10.1177/03611981231169279.
Sun, J., Zheng, Z., Sharma, A., Sun, J., Stability and extension of a car-following model for human-driven connected vehicles. Transport. Res. Part C: Emerg. Technol., 155, 2023, 104317, 10.1016/j.trc.2023.104317.
Treiber, M., Hennecke, A., Helbing, D., Congested traffic states in empirical observations and microscopic simulations. Phys. Rev. E, 62(2), 2000, 1805, 10.1103/PhysRevE.62.1805.
Treiber, M., Kesting, A., Traffic Flow Dynamics: Data, Models and Simulation. 2013, Springer, Berlin, 10.1007/978-3-642-32460-4.
Udrescu, S.-M., Tegmark, M., AI feynman: A physics-inspired method for symbolic regression. Sci. Adv., 6(16), 2020, 2631, 10.1126/sciadv.aay2631.
Wang, H., Fu, T., Du, Y., Gao, W., Huang, K., Liu, Z., Chandak, P., Liu, S., Van Katwyk, P., Deac, A., et al. Scientific discovery in the age of artificial intelligence. Nature 620:7972 (2023), 47–60, 10.1038/s41586-023-06221-2.
Wang, G., Li, W., Zhang, L., Sun, L., Chen, P., Yu, L., Ning, X., Encoder-x: solving unknown coefficients automatically in polynomial fitting by using an autoencoder. IEEE Trans. Neural Netw. Learn. Syst. 33:8 (2021), 3264–3276, 10.1109/TNNLS.2021.3051430.
Wang, T., Ngoduy, D., Li, Y., Lyu, H., Zou, G., Dantsuji, T., Koopman theory meets graph convolutional network: Learning the complex dynamics of non-stationary highway traffic flow for spatiotemporal prediction. Chaos Solitons Fractals, 187, 2024, 115437, 10.1016/j.chaos.2024.115437.
Wang, T., Ngoduy, D., Zou, G., Dantsuji, T., Zongshi, L., Li, Y., PI-STGnet: Physics-integrated spatiotemporal graph neural network with fundamental diagram learner for highway traffic flow prediction. Expert Syst. Appl., 258, 2024, 125144, 10.1016/j.eswa.2024.125144.
Wang, X., Zhang, J., Li, H., He, Z., A mixed traffic car-following behavior model. Phys. A, 632, 2023, 129299, 10.1016/j.physa.2023.129299.
Wang, S., Zhu, Y., Li, Z., Wang, Y., Li, L., He, Z., ChatGPT as your vehicle co-pilot: An initial attempt. IEEE Trans. Intell. Veh. 8:12 (2023), 4706–4721, 10.1109/TIV.2023.3325300.
Wen, X., Jian, S., He, D., Modeling the effects of autonomous vehicles on human driver car-following behaviors using inverse reinforcement learning. IEEE Trans. Intell. Transp. Syst., 2023, 10.1109/TITS.2023.3298150.
Wen, Y., Xu, P., Li, Z., Xu, W., Wang, X., RPConvformer: A novel transformer-based deep neural networks for traffic flow prediction. Expert Syst. Appl., 218, 2023, 119587, 10.1016/j.eswa.2023.119587.
Wu, F., Stern, R.E., Cui, S., et al. Tracking vehicle trajectories and fuel rates in phantom traffic jams: methodology and data. Transport. Res. Part C: Emerg. Technol. 99 (2019), 82–109, 10.1016/j.trc.2018.12.012.
Xu, D., Gao, G., Qiu, Q., Li, H., A car-following model considering missing data based on TransGAN networks. IEEE Trans. Intell. Veh., 2023, 10.1109/TIV.2023.3270336.
Yang, X., Zou, Y., Zhang, H., Qu, X., Chen, L., Improved deep reinforcement learning for car-following decision-making. Phys. A, 624, 2023, 128912, 10.1016/j.physa.2023.128912.
Yao, H., Li, X., Yang, X., Physics-aware learning-based vehicle trajectory prediction of congested traffic in a connected vehicle environment. IEEE Trans. Veh. Technol. 72:1 (2023), 102–112, 10.1109/TVT.2022.3203906.
Yao, Z., Wu, Y., Jiang, Y., Ran, B., Modeling the fundamental diagram of mixed traffic flow with dedicated lanes for connected automated vehicles. IEEE Trans. Intell. Transp. Syst. 24:6 (2023), 6517–6529, 10.1109/TITS.2022.3219836.
Yin, J., Li, Z., Cao, P., Li, L., Ju, Y., Car-following modeling based on morse model with consideration of road slope in connected vehicles environment. Phys. A, 622, 2023, 128827, 10.1016/j.physa.2023.128827.
Yu, W., Ngoduy, D., Hua, X., Wang, W., On the stability of a heterogeneous platoon-based traffic system with multiple anticipations in the presence of connected and automated vehicles. Transport. Res. Part C: Emerg. Technol., 157, 2023, 104389, 10.1016/j.trc.2023.104389.
Zhang, Y., Talebpour, A., Characterizing human–automated vehicle interactions: An investigation into car-following behavior. Transport. Res. Rec., Online First, 2023, 03611981231192999, 10.1177/03611981231192999.
Zhang, Z., Yuan, Y., Li, M., Lu, P., Yang, X.T., Empirical study of the effects of physics-guided machine learning on freeway traffic flow modelling: model comparisons using field data. Transport. A: Transp. Sci., Online, 2023, 10.1080/23249935.2023.2264949.
Zheng, S.-T., Jiang, R., Jia, B., Tian, J., Bouadi, M., Makridis, M.A., Kouvelas, A., A parsimonious enhanced Newell's model for accurate reproduction of driver and traffic dynamics. Transport. Res. Part C: Emerg. Technol., 154, 2023, 104276, 10.1016/j.trc.2023.104276.
Zhou, Y., Zhong, X., Chen, Q., Ahn, S., Jiang, J., Jafarsalehi, G., Data-driven analysis for disturbance amplification in car-following behavior of automated vehicles. Transport. Res. Part B: Methodol., 174, 2023, 102768, 10.1016/j.trb.2023.05.005.
