The problem of anti-sliding control of wheeled mobile robot is presented. Firstly, lateral vehicle dynamics and
acceleration equations are considered to obtain an equation describing the relationship between the sideslip angle and the cornering stiffness. Then linear Luenberger observer is used to obtain adaptive learning laws for the cornering stiffness, which leads to the convergence of the estimated cornering stiffness to its real value when the persistent excitation condition is satisfied. Therefore, the unknown sideslip angles can be reconstructed precisely. With the results of the sliding-angle reconstruction, a path-following controller with sliding compensation is designed based on the chained system theory. Finally simulation results show that the proposed adaptive learning laws can guarantee the precise reconstruction of the sliding angles, and the anti-sliding controller can moderate negative sliding effects and improve lateral control accuracy in spite of sliding.