Vehicle Sideslip Angle Estimation Based on Linearization Method

Abstract

A comparison of two kinds of nonlinear observers based on different tire models is proposed in this paper. A 2 DOF bicycle vehicle model and two kinds of tire models are adopted for sideslip angle estimation. One sideslip angle observer using Atan tire model utilizes Jacobi linearization method while the other one using Dugoff tire model utilizes local linearization method. Jacobi linearization method is suitable to models with simpler expressions and makes it easy to calculate feedback gains of an observer. On the other hand, local linearization method has higher commonality and makes it possible to employ tire models with higher accuracy and more complicated expressions while results in relatively greater calculated amount. The effectiveness and parameter variation robustness of the proposed algorithms are evaluated in a high-fidelity vehicle dynamics simulation environment. Simulation results show that on the whole the observer using Dugoff tire model has higher estimation accuracy than that using Atan tire model due to more accurate models and proper feedback gains.