Estimation of longitudinal speed robust to road conditions for ground vehicles

Abstract

Abstract-This article seeks to develop a longitudinal vehicle velocity estimator robust to road conditions by employing a tire model at each corner. Combining the lumped LuGre tire model and the vehicle kinematics, the tires internal deflection state is used to gain an accurate estimation. Conventional kinematicbased velocity estimators use acceleration measurements, without correction with the tire forces. However, this results in inaccurate velocity estimation because of sensor uncertainties which should be handled with another measurement such as tire forces that depend on unknown road friction. The new Kalman-based observer in this paper addresses this issue by considering tire nonlinearities with a minimum number of required tire parameters and the road condition as uncertainty. Longitudinal forces obtained by the unscented Kalman filter on the wheel dynamics is employed as an observation for the Kalman-based velocity estimator at each corner. The stability of the proposed time-varying estimator is investigated and its performance is examined experimentally in several tests and on different road surface frictions. Road experiments and simulation results show the accuracy and robustness of the proposed approach in estimating longitudinal speed for ground vehicles.

A comprehensive study on the stability analysis of vehicle dynamics with pure/combined-slip tyre models

“…In addition, the rubber stiffness is σ 0 = [σ 0x 0; 0 σ 0y ], the rubber damping is σ 1 = [σ 1x 0; 0 σ 1y ], and the relative viscous damping is defined by σ 2 = [σ 2x 0; 0 σ 2y ], in which σ 0j , σ 1j and σ 2j are the rubber stiffness, damping, and relative viscous damping in each direction. These pure and combined-slip models can be used in road-independent state estimation approaches [30], will be incorporated in the lateral dynamics, and are described in the next section.…”

Distributed robust vehicle state estimation

“…The term w 2 shows the deviation of the measured relative acceleration R eω −v xt froṁ v rx because of measurement noises. Combining the modelbased states (1) and kinematic-based states (2), the proposed approach in [15] estimates relative velocities at each corner for ω > 0 by the following linear parameter-varying system with the longitudinal states…”

“…The UKF employs a transformation, which introduces the Sigma vectors Σ ∈ RN ×2N +1 (N is the length of the state vectors) around x, to include the nonlinear and non-Gaussian characteristics of the system that was a challenge for the longitudinal velocity estimation in [15]. Algorithm 1 shows the state estimation steps using the UKF.…”

“…The kinematic approach is combined with the tire's internal states at each corner in [15] to estimate longitudinal speed and to tackle uncertain model parameters and measurement noises. Because of the dynamics in its internal state, the lumped LuGre model [18], [19] has been selected and combined with the kinematics (accelerations at each corner) to estimate longitudinal velocity.…”

“…Such a reliable estimator with corner-based structure is proposed in [15]. However, the developed observer is required to be augmented with a module to address reliability of the estimation setup.…”