Vehicle sideslip angle estimation under tire nonlinearity: A nonlinear disturbance estimator with experimental validation

Accurate estimation of the vehicle sideslip angle at nonlinear driving conditions is important for the vehicle stability control systems. A nonlinear disturbance estimator (NDE) that is designed for real-time implementation on production vehicle electronic control units is presented in this paper. A seven-degree-of-freedom vehicle model is first developed with an explicit consideration of vertical tire force variations to capture load‑transfer‑induced nonlinear tire characteristics, which is an important component affecting the vehicle stability during extreme maneuvers. The NDE is constructed using an adaptive square‑root cubature Kalman filter (ASRCKF) which recursively updates noise covariances to keep robustness under different road conditions. The estimator is experimentally validated by double‑lane change tests on dry asphalt and slalom tests on low‑friction icy surfaces, with results quantitatively compared against a conventional linear disturbance estimator (LDE). The NDE decreases root mean square error by 42% on dry asphalt and 66% on ice, while the maximum absolute error—critical for safety applications—is decreased by 42% and 49%, respectively. The algorithm is proven by the computational analysis to be adequate to implement in real‑time with 2.8 ms per iteration. These results show that the proposed NDE provides reliable performance in extreme conditions, which is a practical solution to improving vehicle stability control.