A Probabilistic Framework for Imitating Human Race Driver Behavior

Understanding and modeling human driver behavior is crucial for advanced vehicle development. However, unique driving styles, inconsistent behavior, and complex decision processes render it a challenging task, and existing approaches often lack variability or robustness. To approach this problem, we propose Probabilistic Modeling of Driver behavior (ProMoD), a modular framework which splits the task of driver behavior modeling into multiple modules. A global target trajectory distribution is learned with Probabilistic Movement Primitives, clothoids are utilized for local path generation, and the corresponding choice of actions is performed by a neural network. Experiments in a simulated car racing setting show considerable advantages in imitation accuracy and robustness compared to other imitation learning algorithms. The modular architecture of the proposed framework facilitates straightforward extensibility in driving line adaptation and sequencing of multiple movement primitives for future research.

[1]  Wolfram Burgard,et al.  Learning driving styles for autonomous vehicles from demonstration , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[2]  Charles C. MacAdam,et al.  Understanding and Modeling the Human Driver , 2003 .

[3]  Pawel Cichosz,et al.  Imitation learning of car driving skills with decision trees and random forests , 2014, Int. J. Appl. Math. Comput. Sci..

[4]  A. Modjtahedzadeh,et al.  A control theoretic model of driver steering behavior , 1990, IEEE Control Systems Magazine.

[5]  Riender Happee,et al.  Differences between racing and non-racing drivers: A simulator study using eye-tracking , 2017, PloS one.

[6]  Markus Maurer,et al.  Model predictive control based trajectory generation for autonomous vehicles — An architectural approach , 2017, 2017 IEEE Intelligent Vehicles Symposium (IV).

[7]  Geoffrey J. Gordon,et al.  A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning , 2010, AISTATS.

[8]  Enrico Bertolazzi,et al.  Fast and accurate $G^1$ fitting of clothoid curves , 2013, 1305.6644.

[9]  Stefano Ermon,et al.  Generative Adversarial Imitation Learning , 2016, NIPS.

[10]  Jan Peters,et al.  Probabilistic Movement Primitives , 2013, NIPS.

[11]  William F. Milliken,et al.  Race Car Vehicle Dynamics , 1994 .

[12]  E. Bertolazzi,et al.  G1 fitting with clothoids , 2015 .

[13]  Mayank Bansal,et al.  ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst , 2018, Robotics: Science and Systems.

[14]  Silvia Ferrari,et al.  Modeling of human driver behavior via receding horizon and artificial neural network controllers , 2013, 52nd IEEE Conference on Decision and Control.

[15]  Jan Peters,et al.  Using probabilistic movement primitives in robotics , 2018, Auton. Robots.

[16]  Miriam Ruf Geometrie und Topologie von Trajektorienoptimierung für vollautomatisches Fahren , 2018 .

[17]  Tommaso Novi A control system framework for autonomous vehicles at the limits of handling , 2019 .

[18]  David González,et al.  A Review of Motion Planning Techniques for Automated Vehicles , 2016, IEEE Transactions on Intelligent Transportation Systems.

[19]  Florian Strecker,et al.  Simulation-based Development Process for Motor Sport and Series Production , 2017 .

[20]  Kyunghyun Cho,et al.  Query-Efficient Imitation Learning for End-to-End Simulated Driving , 2017, AAAI.

[21]  Dieter Schramm,et al.  Modellbildung und Simulation der Dynamik von Kraftfahrzeugen , 2010 .

[22]  A. M. Turing,et al.  Computing Machinery and Intelligence , 1950, The Philosophy of Artificial Intelligence.

[23]  Hermann Winner,et al.  Autonomous Driving: Technical, Legal and Social Aspects , 2016 .

[24]  John A. Michon,et al.  A critical view of driver behavior models: What do we know , 1985 .

[25]  Mykel J. Kochenderfer,et al.  Imitating driver behavior with generative adversarial networks , 2017, 2017 IEEE Intelligent Vehicles Symposium (IV).

[26]  Daniele Loiacono,et al.  Learning drivers for TORCS through imitation using supervised methods , 2009, 2009 IEEE Symposium on Computational Intelligence and Games.