This paper presents an in-depth analysis of the application of different techniques for vehicle state and tyre force estimation using the same experimental data and vehicle models, except for the tyre models. Four schemes are demonstrated: (i) an Extended Kalman Filter (EKF) scheme using a linear tyre model with stochastically adapted cornering stiffness, (ii) an EKF scheme using a Neural Network (NN) data-driven linear tyre model, (iii) a tyre model-less Suboptimal-Second Order Sliding Mode (S-SOSM) scheme, and (iv) a Kinematic Model (KM) scheme integrated in an EKF. The estimation accuracy of each method is discussed. Moreover, guidelines for each method provide potential users with valuable insight into key properties and points of attention.

This paper proposes and experimentally validates a two-stage approach for coupled lateral vehicle state and tire model estimation. In a first stage, an extended Kalman filter is employed which provides vehicle slip angles and lateral tire forces from commercial low-cost vehicle sensors. The obtained estimates are exploited in the second stage, where a (quasi-static) tire model is fitted to this data. A major issue in this estimation process is the typical instability of these estimators for situations with (prolonged) straight driving. This issue is traced back to a lack of local observability. The use of a variable model covariance is introduced as a practical method to obtain a stable estimator, irrespective of the unobservability. The developed methodology has a low computational load and the Kalman estimator is able to run in real time, whereas the tire model parameter fitting is cheap enough to run online. The proposed methodology is validated experimentally and provides reliable results in variable driving conditions.