A Probabilistic Framework for Imitating Human Race Driver Behavior
This addresses the challenge of imitating human race driver behavior for advanced vehicle development, representing an incremental improvement with a modular approach.
The paper tackles the problem of modeling human driver behavior in vehicle development by proposing ProMoD, a modular framework that splits the task into multiple modules using Probabilistic Movement Primitives, clothoids, and neural networks. Experiments in simulated car racing show considerable advantages in imitation accuracy and robustness compared to other imitation learning algorithms.
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.