This paper addresses the challenges of low lateral control accuracy and complex disturbances encountered by autonomous tractors operating in unstructured agricultural environments by proposing an active disturbance rejection control strategy. The tractor is modelled as a rear-wheel-drive, front-wheel-steering vehicle, where factors such as variations in tyre-ground friction and the non-linear characteristics of the steering mechanism are treated as external disturbances acting upon the electric steering motor. A controller based on an disturbance observer is designed to achieve precise motor position control, utilising the observer to estimate and compensate for lumped disturbances in real time. This approach, which does not require an exact tractor model, significantly enhances lateral accuracy and disturbance suppression performance in complex agricultural terrain, providing an innovative control framework for autonomous agricultural machinery.
Li et al. (2026) studied this question.