Backstepping-based decentralized tracking control for a class of interconnected stochastic nonlinear systems coupled via a directed graph (Q2201675)

For single stochastic non-linear systems tracking control has been investigated using backstepping via quartic Lyapunov functions, see \textit{H.-B. Ji} and \textit{H.-S. Xi} [IEEE Trans. Autom. Control 51, No. 2, 355--360 (2006; Zbl 1366.93620)]. More recently, systems consisting of a set of stochastic nonlinear state-dependently interconnected systems have been studied. An adaptive neural decentralized controller, where all signals are semiglobally bounded, was designed by \textit{H. Q. Wang}, \textit{X. P. Liu} and \textit{K. F. Liu} [IEEE Trans. Neural Netw. Learn Syst. 27, No. 3, 519--523 (2016)]. The problem of global decentralized tracking control for such systems has so far remained open. It is treated here, at least for a certain class of such systems. To achieve the desired tracking performance, the authors present a new strictly decentralized control and a modified compensation mechanism addressing the unknown state-dependent interconnections. By introducing a quartic Lyapunov function for each subsystem and using graph theory, a global Lyapunov function is provided, so that global boundedness in probability of all signals of the resulting closed-loop system is secured. It is shown that the tracking errors of each subsystem converge to an adjustable compact set with a radius related to the controller design parameters chosen. Two examples are given to show the effectiveness of the method.