Anti-windup compensation for systems with sensor saturation: a study of architecture and structure

Abstract

This paper proposes a linear dynamic anti-windup strategy for the alleviation of performance and stability problems in systems which are linear apart from saturating sensors. Unlike anti-windup compensation for systems subject to actuator saturation, there is no agreed architecture for applying anti-windup to systems with sensor saturation and therefore attention is devoted to the discussion of various candidate configurations which can be interpreted as a particular choice of a static nonlinear map. The main results of the paper give existence conditions for compensators which yield global exponential stability and finite L 2 gain of the overall nonlinear closed-loop system. The existence conditions are different to those which have appeared hitherto in the literature.

An anti-windup design to robust H∞ control for singular Markovian jump systems with actuator saturation and general unknown transition rates

“…The framework is based on the following two-step design paradigm: design output feedback controllers to meet the performance without actuator saturation firstly, and then design anti-windup compensators to minimize the adverse effects of actuator saturation on the closed-loop systems [23][24][25][26][27][28][29][30][31][32][33][34][35][36].…”

Coprime factor anti-windup for systems with sensor saturation

“…Email: mct6@le.ac.uk stated in [6]. In particular, these approaches either provide globally stabilising control laws for stable systems [9], [20], or they provide control laws which only guarantee local results [16], [5]. Another interesting observer-based design was proposed in [8], where semi-global stability guarantees were obtained for minimum-phase SISO systems; the results do not appear straightforward to extend to MIMO systems.…”

“…Some possible choices are discussed in [20], but essentially they involve the use of an observer to estimate the output, as initially proposed in [16]. In this paper we shall use an anti-windup architecture which corresponds to the favoured architecture in [20], but which is similar to coprime-factor based residual generation found in fault detection and highperformance control schemes [25], [2]. The resulting architecture is based on fault detection ideas and standard antiwindup applications, in which a coprime factorisation is used to generate a (residual) signal that drives the anti-windup compensator.…”

“…Hence there are a variety of different pseudo-anti-windup architectures which could be used to generate an estimate of the real output in order to generate a signal to drive the compensator. Some possible choices are discussed in [20], but essentially they involve the use of an observer to estimate the output, as initially proposed in [16]. In this paper we shall use an anti-windup architecture which corresponds to the favoured architecture in [20], but which is similar to coprime-factor based residual generation found in fault detection and highperformance control schemes [25], [2].…”

“…However, the control constructions advocated in [12], [6] are of a complex, nonlinear form and are not attractive for implementation. Simpler approaches have been advocated in [9], [1], [16], [5], [20], but they do not provide the generality of the approach Jorge Sofrony is with the Department of Mechatronics, National University of Colombia, Cra. 30 No.…”