This paper discusses the design of a human-aware cooperative adaptive cruise control (CACC) system that (i) takes into account driver comfort in autonomous cruise control, and (ii) provides assistive corrections to avoid driver errors. To incorporate driver characteristics into system controller design, two self-learning algorithms are used to estimate driver’s preferred time headway. We then develop a human-like blending control for CACC based on a model predictive control (MPC)-type method, which integrates the driver comfort, traffic efficiency, and fuel economy criteria. Furthermore, a driving assistance controller is developed to help human driver to maintain string stability in platoon. Simulation results show that (i) the human-like CACC design can significantly improve driving experience, and (ii) with the help of the assistive controller, string stability is satisfied for both exclusively autonomous CACC and when the CACC switches to manual driving in a platoon.

This content is only available via PDF.
You do not currently have access to this content.