Estimation of Tire Parameters via Second-Order Sliding Mode Observers with Unknown Inputs

This paper considers the identification of unknown parameters related to a vehicle tire model using second-order sliding mode techniques. In a tailored two-stage procedure unknown input state observers recover the characteristic tire parameters. One of them is based on a modified super-twisting algorithm that extends the class of feasible uncertainties. In contrast to other methods for identification of tire parameters, typically requiring extensive sensor equipment, the proposed mechanisms only rely on standard measurement sensors. Furthermore, the algorithms are employable in-vehicle for online identification of model parameters. The overall concept is evaluated by the use of a complex vehicle dynamics simulation software environment.