Aware of Scene Vehicles—Probabilistic Modeling of Car-Following Behaviors in Real-World Traffic
Donghao Xu,Huijing Zhao,Franck Guillemard,Stephane Geronimi,Francois Aioun
DOI: https://doi.org/10.1109/tits.2018.2863939
IF: 8.5
2018-01-01
IEEE Transactions on Intelligent Transportation Systems
Abstract:Heterogeneity exists in car following behaviors clue to the driver's habit, fatigue, distraction, or surrounding traffic. This research proposes a method of modeling and reasoning heterogeneous car-following behaviors based on a stochastic system, where scene vehicles are involved explicitly in addition to the traditional leader-follower pair in characterizing driving situations, a hidden variable (driver state) is introduced to conjugate driving situations to heterogeneous models in predicting a driver's acceleration control, and the dynamic procedure is described using a dynamic Bayesian network. Experiments are conducted using a large set of naturalistic driving data that were collected by driving an instrumented vehicle on the multi-lane motorways in Beijing, where four distinctive driver states are learnt from data, characterizing the car-following procedure with normal, slow responsive, strong and prompt responsive, and unresponsive behavioral styles. By using the proposed scene-aware multi-state model for acceleration prediction, the error is reduced to 0.19 m/s(2) in average compared with 0.29 m/s(2) of a single-state model. Influence of scene vehicles on a driver state and subsequently on velocity control is verified based on the data. To the best of our knowledge, this is the first work that explicitly incorporates scene vehicles as influential factors in a probabilistic approach for modeling and reasoning heterogeneous car-following behaviors, and the performance is demonstrated on a large set of naturalistic driving data.