This paper is concerned with the problem of adaptive stabilization for a class of switched linear-parametric nonlinear systems under arbitrary switching. The traditional adaptive backstepping control is successfully extended to switched systems from nonswitched ones where the asymptotic regulation of system state is not destroyed due to rapid or abrupt changes of switching parameters. A new switched adaptive controller is designed by exploiting a common high-orderLyapunov function with a -modification mechanism, which can reflect sufficiently the changes of plant by designing different adaptive laws and control laws for different subsystems. An explicit formula for constructing a continuous and piecewise n−1 virtual control function is given to remove the restriction where some ∞ bound functions have to be constructed blindly by designers in the existing results, which may be somewhat too strict to be applied. A numerical example is provided to validate the proposed approach. /journal/rnc Int J Robust Nonlinear Control. 2018;28:3422-3434.ZHAI ET AL.
3423uncertain switching nonlinear functions via fuzzy logic systems or neural networks. 17,22,23 It is shown that these obtained controllers can semiglobally guarantee the tracking error or system state converges to a small neighborhood of the origin. However, for switched parametric nonlinear systems, there are few results on adaptive controller design to guarantee their globally asymptotic stability. The main difficulties are that some key inequalities such as Young's inequality cannot be applied since they may destroy asymptotic regulation. In the work of Chiang and Fu, 21 based on tuning the estimation of the bounds on switching parameters in a transformed system, the adaptive stabilization controllers are designed for nonlinear systems with unknown switching parameters. However, for practical implementation, the asymptotic regulation has to be sacrificed to remove chatting phenomenon.In the aforementioned literature, the key idea to deal with switching uncertainties in the plant is robust adaptive control, ie, adaptively accommodate the worst effects on system of the uncertainties instead of switching ones, and then design adaptive controllers without explicitly considering the changes of plant. 21 Obviously, these control schemes tend to be unduly conservative, and establishing the mode-dependent adaptive estimation and control may be more reasonable in practical applications. In the works of Long et al, 24,25 switched adaptive control problem was considered for switched nonlinearly parameterized systems with unstable subsystems, where different update laws are designed for different subsystems. However, it is required in the works of the aforementioned authors 24,25 that some key ∞ bound functions should be constructed, not related to switching laws, and satisfy some inequality constraints, which is somewhat too strict in practice to be applied. Moreover, no systematic methods have been proposed to show how to find these bound functions. Hence, the study ...