Approximation-based decentralized adaptive tracking for a class of arbitrarily switched interconnected non-affine nonlinear systems

This paper investigates the decentralized adaptive tracking problem for a class of uncertain switched interconnected non-affine nonlinear systems under arbitrary switching. Based on dynamic surface design and common Lyapunov function method, an adaptive approximation design methodology for the decentralized adaptive tracking is proposed to ensure that the total controlled closed-loop system is asymptotically bounded and the local output tracking performance is maintained within preselected time-varying bounds independent of switched interactions. For this decentralized tracker design, an auxiliary nonlinear function is introduced to overcome the interaction problem of unknown switched nonlinearities. Contrary to the related results in the literature, the proposed new design approach firstly deals with the switched interactions among switched subsystems in the non-affine form, ensures the transient performance at switching instants within preselected bounds, and does not require the information on the signs of control gain functions. A simulation example is provided to show the effectiveness of the resulting decentralized control system.