Adaptive Estimation of Vehicle Velocity From Updated Dynamic Model for Control of Anti-Lock Braking System

Abstract

The main challenge for designing the automotive control systems like the anti-lock braking system (ABS) is to access a reliable model because of uncertainties and variations in vehicle dynamics and tire forces. In this article, the initial dynamic model of the vehicle braking system is updated at each instant by estimation of model compensatory terms using a novel prediction approach. The proposed estimation method uses the global navigation satellite system (GNSS) as a reference to adaptively estimate the vehicle velocity with high frequency in the presence of variable bias of acceleration sensor. The proposed estimation algorithm is mathematically analyzed and experimentally evaluated to show its high accuracy obtained by the adaptive tuning of parameters. Accordingly, a nonlinear controller based on the estimated dynamic model is designed to prevent the tire from being locked by tracking the desired slip. The designed control system is tested within the CarSim software as a virtual vehicle experiment platform. The obtained results reveal that the designed model-based control system remarkably improves the braking performance compared with the conventional sensor-based ABS controllers.