Deploying autonomous vehicles (AVs) in mixed traffic requires a background traffic environment that reproduces the variability and uncertainty of human driving. This paper proposes MIG, a hierarchical highway human driving behavior model combining Maximum Entropy Deep Inverse Reinforcement Learning (MEDIRL) with non-cooperative game theory. Surrounding traffic is organized into three layers: the outer layer replays real trajectory data; the inner layer re-drives vehicles disturbed by the ego vehicle with a MEDIRL model whose nonlinear reward is a deep neural network, avoiding manual feature engineering; and the game layer selects the highest-risk neighbor and resolves the interaction as a Nash equilibrium. Trained and validated on NGSIM US-101 in a purpose-built environment (ngsim_env), MIG converges stably, generates game trajectories close to the real ones, and reproduces the speed and acceleration distributions of real traffic more faithfully than the IDM+MOBIL and IRL baselines.

Qiyue Chang. Driving Behavior Modeling of Human Drivers on Highways Based on Maximum Entropy Deep Inverse Reinforcement Learning and Non-Cooperative Game Theory. Academic Journal of Computing & Information Science (2026), Vol. 9, Issue 9: 31-39. https://doi.org/10.25236/AJCIS.2026.090905.

[1] Wang M, Zhang L, Wang Z, et al. A Real-Time Dynamic Trajectory Planning for Autonomous Driving Vehicles. 2019 3rd Conference on Vehicle Control and Intelligence (CVCI). Hefei, China: IEEE, 2019: 1-6.

[2] Feng S, Yan X, Sun H, et al. Intelligent Driving Intelligence Test for Autonomous Vehicles with Naturalistic and Adversarial Environment. Nature Communications, 2021, 12(1): 748.

[3] Nallaperuma D, De Silva D, Alahakoon D, et al. Intelligent Detection of Driver Behavior Changes for Effective Coordination Between Autonomous and Human Driven Vehicles. IECON 2018 - 44th Annual Conference of the IEEE Industrial Electronics Society. Washington, DC: IEEE, 2018: 3120-3125.

[4] Kalra N, Paddock S M. Driving to Safety: How Many Miles of Driving Would It Take to Demonstrate Autonomous Vehicle Reliability? Transportation Research Part A: Policy and Practice, 2016, 94: 182-193.

[5] Treiber M, Hennecke A, Helbing D. Congested Traffic States in Empirical Observations and Microscopic Simulations. Physical Review E, 2000, 62(2): 1805-1824.

[6] Newell G F. A Simplified Car-Following Theory: A Lower Order Model. Transportation Research Part B: Methodological, 2002, 36(3): 195-205.

[7] Gipps P G. A Behavioural Car-Following Model for Computer Simulation. Transportation Research Part B: Methodological, 1981, 15(2): 105-111.

[8] Kesting A, Treiber M, Helbing D. General Lane-Changing Model MOBIL for Car-Following Models. Transportation Research Record, 2007, 1999(1): 86-94.

[9] Ikotun A M, Ezugwu A E, Abualigah L, et al. K-means Clustering Algorithms: A Comprehensive Review, Variants Analysis, and Advances in the Era of Big Data. Information Sciences, 2023, 622: 178-210.

[10] Siami M, Naderpour M, Lu J. A Mobile Telematics Pattern Recognition Framework for Driving Behavior Extraction. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(3): 1459-1472.

[11] Brown K, Driggs-Campbell K, Kochenderfer M J. A Taxonomy and Review of Algorithms for Modeling and Predicting Human Driver Behavior. arXiv preprint, 2020, arXiv:2006.12188.

[12] Li N, Oyler D W, Zhang M, et al. Game Theoretic Modeling of Driver and Vehicle Interactions for Verification and Validation of Autonomous Vehicle Control Systems. IEEE Transactions on Control Systems Technology, 2018, 26(5): 1782-1797.

[13] Krizhevsky A, Sutskever I, Hinton G E. ImageNet Classification with Deep Convolutional Neural Networks. Communications of the ACM, 2017, 60(6): 84-90.

[14] Hochreiter S, Schmidhuber J. Long Short-Term Memory. Neural Computation, 1997, 9(8): 1735-1780.

[15] Kebria P M, Khosravi A, Salaken S M, et al. Deep Imitation Learning for Autonomous Vehicles Based on Convolutional Neural Networks. IEEE/CAA Journal of Automatica Sinica, 2020, 7(1): 82-95.

[16] Georgakopoulos G F, Kavvadias D J, Sioutis L G. Nash Equilibria in All-Optical Networks. Discrete Mathematics, 2009, 309(13): 4332-4342.

[17] Xu C, Zhao W, Li L, et al. A Nash Q-Learning Based Motion Decision Algorithm with Considering Interaction to Traffic Participants. IEEE Transactions on Vehicular Technology, 2020, 69(11): 12621-12634.

[18] Zhu M, Wang X, Wang Y. Human-Like Autonomous Car-Following Model with Deep Reinforcement Learning. Transportation Research Part C: Emerging Technologies, 2018, 97: 348-368.

[19] Wulfmeier M, Ondruska P, Posner I. Maximum Entropy Deep Inverse Reinforcement Learning. arXiv preprint, 2016, arXiv:1507.04888.

[20] Huang Z, Wu J, Lv C. Driving Behavior Modeling Using Naturalistic Human Driving Data with Inverse Reinforcement Learning. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(8): 10239-10251.

[21] Zhou Y, Fu R, Wang C. Learning the Car-Following Behavior of Drivers Using Maximum Entropy Deep Inverse Reinforcement Learning. Journal of Advanced Transportation, 2020, 2020: 1-13.

[22] Abbeel P, Ng A Y. Apprenticeship Learning via Inverse Reinforcement Learning. Twenty-First International Conference on Machine Learning (ICML '04). Banff, Alberta, Canada: ACM Press, 2004: 1.

[23] Boularias A, Kober J, Peters J. Relative Entropy Inverse Reinforcement Learning. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics (AISTATS). 2011: 182-189.

[24] Ziebart B D, Maas A, Bagnell J A, et al. Maximum Entropy Inverse Reinforcement Learning. Proceedings of the 25th International Conference on Machine Learning (ICML 2008). Helsinki, Finland: ACM Press, 2008: 1433-1438.

[25] Road Traffic Safety Law of the People's Republic of China (2021 Revision). Beijing Traffic Administration. Available online: https://jtgl.beijing.gov.cn/jgj/jgxx/flfg/fl/205308/index.html (accessed on 13 May 2024).

[26] Schafer R. What Is a Savitzky-Golay Filter? [Lecture Notes]. IEEE Signal Processing Magazine, 2011, 28(4): 111-117.