Adaptive Extended Kalman Filter for the state estimation of Anti-Lock Braking System

An adaptive Extended Kalman Filter (EKF) based estimator for state estimation of an Anti-lock Braking System (ABS) is presented. Though the use of estimators for ABS is widespread, such estimators fail to perform satisfactorily in the presence of sensor noise uncertainties. The proposed adaptive EKF based scheme estimates the vehicle velocity, wheel speed and internal parameter of the tire even when the measurement noise statistics of the system are unknown. The quarter car model with the well-known non-linear LuGre friction model has been considered to model the ABS system. A simulation study has been carried out to evaluate the performance of the adaptive EKF as compared to the non-adaptive filter for the ABS system.