Improved Finite-Time Disturbance Observer-Based Control of Networked Nonholonomic High-Order Chained-Form Systems

Abstract

This article investigates the finite-time leader–follower tracking problem for consensus control of nonholonomic MASs with chained-form dynamics under unknown time-varying disturbances. First, a finite-time distributed observer, consisting of the tangent hyperbolic function with an induced high gain effect, is constructed to estimate the leader’s information both quickly and accurately for each follower leading to communication loop avoidance. Besides, to cope with the adverse impact resulting from external disturbances, a finite-time disturbance observer is developed to provide accurate estimations of unknown terms within a finite time. A finite-time backstepping control scheme with a fast convergence rate is then proposed for each follower based on estimated disturbances to track the estimated states of the leader. Regardless of how close or how far away from the equilibrium point the follower states are, this method accelerates the convergence rate. An approximation technique using piecewise functions is also employed to bring the upper estimate of the convergence time closer to its real value. Finally, the efficiency of the presented control protocol is verified by some simulations on a number of connected wheeled mobile robots under external disturbances.