Adaptive decentralized sensor failure compensation control for nonlinear switched interconnected systems with average dwell time

Abstract

In this article, we investigate the adaptive decentralized output feedback sensor failure compensation control issue of nonlinear switched interconnected systems. A new switched state observer is designed based on the false state information for estimating unmeasured states. To address the challenges incurred by the partial loss of the effectiveness of sensors, a novel adaptive failure compensation mechanism containing error signals is adopted. Mode‐dependent adaptive laws that can explicitly reflect switchings of the system are designed. A decentralized adaptive failure‐tolerant controller is constructed by using backstepping technique. It is proved that the presented scheme can guarantee that the system outputs track the reference signals and all signals in the closed‐loop system are bounded with average dwell time switchings. A practical simulation example is provided to illustrate the effectiveness of the presented scheme.

Non‐approximation‐based outputs stabilization for switched large‐scale nonlinear systems with quantized inputs and sensor uncertainties

“…• The controller for each subsystem of the SLSNS is linear-like and the observers and controllers of the subsystems share the same dynamic gain, which greatly reduces the complexity of the controllers, and in the meanwhile, can effectively compensate for the effects of the sensor uncertainties, external disturbances, quantized inputs, and parameter uncertainties. • Compared with the current research on large-scale systems with sensor uncertainties [27], the requirement that the uncertainties are differentiable is unnecessary. Although the differentiability restriction is removed in [25], instead of one dynamic gain of our paper, two gains are needed to construct the observers and controllers.…”

“… • To the best of the authors' knowledge, this paper makes the first attempt on the control of SLSNSs by taking into account simultaneously arbitrary switching, quantized inputs and sensor uncertainties. • With the gain scaling technique and under the polynomial growth assumption on system nonlinear functions, our controller design needs neither backstepping technique nor fuzzy logic systems to approximate the system nonlinear functions. Moreover, the nonlinear interaction for each subsystem includes not only the outputs of the subsystems but also their internal states, which is different from [8–15], where all the interactions are just the functions of the outputs of the subsystems. • The controller for each subsystem of the SLSNS is linear‐like and the observers and controllers of the subsystems share the same dynamic gain, which greatly reduces the complexity of the controllers, and in the meanwhile, can effectively compensate for the effects of the sensor uncertainties, external disturbances, quantized inputs, and parameter uncertainties. • Compared with the current research on large‐scale systems with sensor uncertainties [27], the requirement that the uncertainties are differentiable is unnecessary. Although the differentiability restriction is removed in [25], instead of one dynamic gain of our paper, two gains are needed to construct the observers and controllers. …”