Enhancing ADS and ADAS Under Critical Road Conditions Through Vehicle Sideslip Angle Estimation via Unscented Kalman Filter-Based Interacting Multiple Model Approach

The knowledge of vehicle dynamical states and parameters plays a crucial role in vehicle stability control systems and, specifically, Vehicle Sideslip Angle (VSA) is an essential factor for active safety control systems. However, the demand for real-time knowledge of this parameter is not practical, due to technical and economic reasons. This paper proposes a novel Interacting Multiple Model Unscented Kalman Filter (IMMUF) to estimate VSA, without tire-road friction coefficient information, and integrating three Unscented Kalman Filters (UKF) to estimate vehicle system models in three different driving conditions (dry, wet, and damp asphalt), characterized by a specific coefficient and modeled through a 2-DOFs single-track vehicle model with a Dugoff tire model. A Monte Carlo analysis has been performed on a wide range of non-trivial driving scenarios and vehicle maneuvers, implemented on a 7-DOFs vehicle model. The results of the estimation have been compared to those of a single UKF, in order to validate the effectiveness of the proposed solution and to highlight the worst performances of a single filter solution in hard driving conditions, justifying the specific Multiple Model solution adopted.