Robust tube-based MPC for automotive adaptive cruise control design

Abstract

This paper proposes a robust model predictive control (MPC) approach for the design of Adaptive Cruise Control (ACC) systems. ACC enables a vehicle to follow a preceding vehicle autonomously and without any additional input form the driver. A reliable ACC must be able to handle driving constraints especially due to safety requirements of the car-following. In practice, constraint handling can be achieved by solving a constrained moving horizon control problem with optimizing a cost function using a prediction of the preceding vehicle's motion. However, uncertainty in the measured data and modeling errors can result in constraint violation and harsh undesired accelerations. The proposed Tube-based MPC ACC uses a tube resulted from bounded additive uncertainties in the system, to achieve robust control with guaranteed constraint handling. Simulations performed on a high-fidelity vehicle model in car-following scenario shows that the designed robust ACC is able to handle defined constraints while following a preceding vehicle in the presence of radar delay, modeling errors and disturbances.