摘要: In agricultural context, the principal cause of serious accidents for all-terrain vehicles (ATVs) is rollover. The most important parameters related to this risk is the ground slope. In this paper, we propose a structured observer to estimate the system states and the longitudinal tire forces using only wheel angular velocities measurement. The robust estimation is based on a second order sliding mode observer. This estimation is then used to build up a ground slope estimation. The algorithm is composed by two cascaded estimators. This structured estimation is then applied to the model of an agricultural vehicle G7 (Gregoire^TM) integrated in the driving simulation environment SCANeR^TM-Studio.