Estimation of Road Friction Coefficient via the Data Enforced Unscented Kalman Filter

Abstract The tire-road friction coefficient (TRFC) plays a critical role in vehicle safety and dynamic stability, with model-based approaches being the primary method for TRFC estimation. However, the accuracy of these methods is often constrained by the complexity of tire force expressions and uncertainties in tire model parameters, particularly under diverse and complex driving conditions. To address these challenges, this paper proposes a novel data-enforced unscented Kalman filter (DeUKF) approach for precise TRFC estimation in intelligent chassis systems. First, an Unscented Kalman Filter is constructed using a nominal tire model-based vehicle dynamics formulation. Then, leveraging Willems' Fundamental Lemma and historical real-world driving data, the vehicle dynamics model is adaptively corrected within the Unscented Kalman Filter framework. This correction effectively mitigates the adverse effects of tire model uncertainties, thereby enhancing TRFC estimation accuracy. Finally, real vehicle experiments are conducted to validate the effectiveness and superiority of the proposed method.