Cost-Effective Estimation of Vehicle Lateral Tire-Road Forces and Sideslip Angle via Nonlinear Sampled-Data Observers: Theory and Experiments
Anh‐Tu Nguyen, Luciano Frezzatto, Thierry‐Marie Guerra, S. Delprat
Abstract
This article proposes a cost-effective method to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">jointly</i> estimate the vehicle sideslip angle and lateral tire-road forces, which are crucial to improve the stability and performance of vehicle control systems. This method only requires the information from onboard sensors, readily available on mass-production vehicles. In particular, we consider the case of sampled <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">asynchronous</i> measurements, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"/> i.e. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"/> , the vehicle sensor signals used for observer design are transmitted at <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">arbitrary</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">distinct</i> times in a certain window bound over the vehicle networked control system. To this end, we propose a new data-sampled observer design, where the asynchronous phenomenon caused by the sampling process is explicitly taken into account via a linear parameter-varying framework. Based on an augmented Lyapunov--Krasovskii functional and specific relaxation techniques, the observer design conditions are derived to guarantee an <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathcal {L}_{2}-$</tex-math></inline-formula> gain performance for the discrete-continuous estimation error dynamics and a maximum allowable sampling period. The observer design is recast as a convex optimization problem, subject to linear matrix inequality (LMI) constraints, which can be efficiently resolved through conventional numerical solvers. The proposed sampled-data observer is experimentally evaluated with an autonomous vehicle under several dynamic driving scenarios, performed on a real test track.