Design and comparative study of steering controller for tracked vehicle based on disturbance observation

Abstract

Tracked vehicles are suitable for rough, low-bearing ground or complex environments because of the large contact area and low pressure between tracks and the ground, and the great obstacle crossing and avoidance capabilities of tracked vehicles. For fast and safe turns, an advanced steering controller is required to achieve precise tracking and stability under system uncertainty. In this study, a steering controller is proposed for tracked vehicles by combining a parameter adaptive law and a second order disturbance observer (DO-2). The friction coefficient between track and ground is estimated online as an uncertain parameter. Random disturbance and function uncertainty are resisted by DO-2. Using Lyapunov’s theory, the input-state stabilities of DO-2 and the closed loop system are proved under random disturbance. A sufficient condition is proposed to check the stability of the system under function uncertainty. Field tests are carried out on rough and slippery grounds using a small-scale electric tracked vehicle. Compared with the controllers based on first order disturbance observer (DO-1) or linear extended state observer (LESO), the optimized DO-2-based steering controller has a smaller root-mean-square error of yaw rate tracking. Adaptive law improves tracking accuracy by 22% on slippery ground.