High speed driving with the Affine in the Force Input model

Abstract

The AFI model is an effective representation of vehicle lateral dynamics, particularly suitable to design linear model-based trajectory tracking controllers. As already observed in the literature, however, controllers designed on the AFI model are affected at high speed by poorly damped yaw rate oscillations, that severely hamper passenger riding comfort and safety. This paper proposes a system-theoretical analysis that explains the cause of these oscillations, and opens the way to the design of an advanced gain-scheduling controller that keeps a constant desired damping ratio independently of vehicle velocity. This controller can be used as an ADAS to increase the safety and comfort of a human driven vehicle, or as the inner loop of a cascaded control architecture in an autonomous vehicle. Simulations demonstrate the effectiveness and robustness of the proposal in damping yaw rate oscillations during a critical manoeuvre, like double-lane-change, performed at different velocities.