State Estimation Of Vehicle Sideslip Angle Based On Reduced-Order Observers

With the popularization of family cars,the problem of vehicle driving safety has been received more and more attention.Generally,vehicle active safety control depends on the accurate and real-time acquisition of the deriving state of the vehicle,especially the vehicle sideslip angle and yaw rate.However,the direct measurement of the sideslip angle is quite expensive.In addition,it also has physical constraints on the instrument installation.Therefore,the general car is mainly through indirect measurement method to obtain the vehicle sideslip angle.Observer design is an important soft-sensing technique,which can reconstruct the vehicle state based on the vehicle dynamics state-space model.Currently,the widely-used observers were the extended full-dimensional Luenberger observer and the Kalman filter,and it was usually assumed that the nonlinear part of the vehicle dynamics model satisfies the classical Lipschitz condition.However,in many case the existing vehicle state observer design was only applicable to nonlinear systems with small Lipschitz constants,so it was more conservative.In this paper,we introduce the so-called one-sided Lipschitz condition for vehicle dynamics systems and propose a reduce-order exponential state observer to reconstruct the vehicle sideslip angle.The main research contents and novelty are as follows:(1)We design a full-order and a reduced-order exponential observers for one-sided Lipschitz nonlinear systems and establish the observer synthesis conditions in term of linear matrix inequalities.The output observation error was mapped into the nonlinear term of observers and the quadratic inner-bounded condition was removed.The freedom of the observer design is enlarged and the design conservativeness is reduced.The new design includes the Luenberger-type observer of the P-one-sided Lipschitz system as a special case.Numerical simulations verify the effectiveness of the proposed design.(2)On the basis of the observer design of the unilateral Lipschitz nonlinear system,an estimation method of the vehicle center-of-mass sideslip angle and yaw rate based on the reduced-order observer is proposed for the simplified two-degree-of-freedom vehicle dynamics model.The proposed design has been validated on Scilab and Matlab software simulation platform.Simulation results show that the proposed method can estimate the vehicle state with the reduced-order observer,and has the advantages of fast convergence speed,simple system structure design and low computational complexity.The real-time performance is expected to improve accuracy of the real-time estimation of vehicle sideslip angle.