This study proposes a two-stage framework for real-time estimation of tire–road friction coefficient of a vehicle on the basis of lateral dynamics of the vehicle. The estimation framework employs a new cascade structure consisting of an extended Kalman filter and two unscented Kalman filters to reduce the computational burden. In the first stage, extended Kalman filter is utilized to estimate lateral velocity of the vehicle and thereby both the front and rear tires’ side-slip angles. In the second stage, a two–unscented Kalman filters sub-framework is formulated in sequence to observe both the front- and rear-axle tire forces, and to subsequently identify their respective tire–road friction coefficient, regarded as two unknown states. All the measured signals required in the study could be realized from the conventional on-board sensors. Typical double-lane change and single-lane change maneuvers were designed and the developed algorithm was verified through CarSim–MATLAB/Simulink software platform considering high-, mid-, and low-friction road conditions. The simulation results show that the proposed method can yield accurate and rapid estimations of the tire–road friction coefficient for mid- and low-friction road conditions even under a single-lane change maneuver, although double-lane change maneuver is needed to accurately estimate the tire–road friction coefficient for high-friction road condition.
Independent Wheel Effects in Real Time Estimation of Tire-Road Friction Coefficient from Steering Torque