Asynchronous non-fragile control for persistent dwell-time switched singularly perturbed systems with strict dissipativity
Introduction
It is acknowledged that multiple time-scales phenomena often occur in the dynamical industrial systems and electronic power systems [1], [2], [3]. In fact, these intractable phenomena are mainly due to some unavoidable small parasitic parameters, for example, the-small time constant of the actuator in the process of industrial systems and some small inductance in the circuit models. The existence of these small parameters leads to the coexistence of “slow” and “fast” time scales, which may lead to higher-order dynamic equations and even ill-conditioned mathematical problems. It has been widely proven that the singularly perturbed systems [3] are an effective tool to analyze the multiple time-scales phenomena. Compared with the traditional research methods which ignore these small parasitic parameters, SPSs as an effective method of modeling such systems, regard these parameters as singularly perturbed parameters (SPP) and handle them skillfully, which can naturally overcome the “stiffness” problem to a certain extent. Therefore, a large number of research efforts have been made toward the analysis and synthesis of SPSs including stability analysis [4], [5], sliding control [6], [7], filtering analysis [8], and references therein. The theoretical and practical importance of such systems is one of the motivations for the current study.
The theoretical researches of SPSs have achieved rich results, but because of the complexity of the analysis, most of these results are limited to the study of non-switched SPSs. However, due to device damage, environmental mutation, component connection failure, and other similar factors, the parameters and structure changes of the physical systems are often encountered in practical applications. If the switching phenomenon is ignored, the established model will not conform to the reality. Therefore, it is of great significance to explore a suitable switching mechanism to describe the switching characteristics. Nevertheless, in the existing literature, the switching among subsystems of SPSs is generally considered to be controlled by Markov chain [9], [10] or Semi-Markov chain [11], [12], which is very restrictive since it is too complicated or costly to obtain enough samples of the transitions in practice. For much practical application of SPSs, such as the electric power systems, chemical reactions and reactors systems, a typical switching mechanism is that the subsystem switches intermittently between slow switching and fast switching. Therefore, there is an interesting question: Is there a suitable switching mechanism that can be introduced to describe the switching phenomenon in switched SPSs? Fortunately, the persistent dwell-time switching mechanism proposed by Hespanha in [13] can solve this problem well. Therefore, the study of PDT switched discrete-time SPSs is one of the main motivations of this paper.
Dissipativity is another popular research frontier based on the input-output energy consideration, introduced in [14], [15]. It involves various fundamental knowledge including the bounded real lemma, the circle criterion, and the passivity theorem. The dissipativity problem can be turned into the performance [16], [17] or passive performance [18], [19] by setting some parameters. Therefore, it is widely used in control systems such as chemical processes, power systems, and electrical networks. Numerous fundamental and classic works on this subject have been published [20], [21], [22], [23], [24] and references therein. For example, a dissipative asynchronous fuzzy sliding mode control issue has been studied for the T-S fuzzy hidden Markov jump systems model in [22]. As for switched stochastic systems, the event-triggering dissipative control has been developed in [23]. By the dissipative knowledge, [24] has concerned with investigating stability conditions for discrete-time Markov jump systems with mixed time delays. According to our knowledge, the dissipative design problem for discrete-time SPSs with PDT switching has not been analyzed, which is another motivation of this paper.
What’s more, there are two shortcomings in some existing methods about SPSs. On the one hand, they ignore the fragility problem according to an implicit assumption that the controller may be exactly implemented. Such an assumption, sometimes, is unavailable owing to the fact that uncertainties or inaccuracies may occur when a designed controller is implemented [25], [26]. On the other hand, the aforementioned SPSs are required that the controller’s switching is synchronized with the subsystem’s switching. However, due to the time delay of mode identification and activation, such restriction is not satisfied in many practical SPSs, one can refer to [27], [28], [29] for more details. Regrettably, the inclusion of non-fragile control and asynchronous switching in the study of SPSs has not been taken into account so far. As a consequence, there are still some interesting questions requiring further investigation for the asynchronous non-fragile control of PDT switched discrete-time SPSs as follows: (1) How to design an asynchronous non-fragile mode-dependent controller when only applying slow state variables for the switched SPSs? (2) How to ensure that the obtained results are -independent and how to develop a technique to estimate the upper bound of singularly perturbed parameters? (3) Compared with the matrix decoupling method to determine the controller gains in [4], [10], [11], is it possible to propose an improved matrix decoupling method to establish less conservative conditions?
Motivated by the above discussions, we shall study the asynchronous non-fragile control issue for PDT switched SPSs under dissipative performance. In view of the switching characteristics, more and more literature have considered the PDT switching mechanism [30], [31], [32], [33], which is more general than dwell-time (DT) [34] or average dwell-time (ADT) switching [35]. A mode-dependent asynchronous non-fragile SSVF controller is constructed via the slow state variables and asynchronous model. We apply the Lyapunov functions to achieve some -independent sufficient conditions, which ensure that EMSS and strict dissipative performance of the closed-loop system. The availability of the obtained results is finally illustrated by applying a numerical example and a modified inverted pendulum model. The contributions of this paper are as follows: (1) To accommodate the model characteristics and reduce design conservatism, a mode-dependent and -dependent non-monotonic Lyapunov function is adopted to obtain controller design algorithms. (2) Unlike some existing results [4], [10], [11], an improved matrix decoupling method is proposed in this paper, which can be served for systems to acquire less conservatism. Moreover, the matrix decoupling methods presented in [4], [10], [11] can be viewed as a special case of our method. (3) The effect of the singularly perturbed parameters upon the system performance is fully addressed, and a technique based on the convex optimization algorithm is developed to estimate the upper bound of singularly perturbed parameters. (4) Compared with the literature [16], [17], [18], [19], the proposed design technique in this paper is more general, which can be transformed into some special cases, including mode-independent, synchronous, , and passive controllers.
The remainder of this work is organized as follows. In Section 2, the system model and some preliminaries are introduced. In Section 3, two main results and two corollaries are given: the exponentially mean-square stable and strict dissipative performance conditions. In Section 4, two examples are provided to demonstrate the effectiveness of the method. Finally, the conclusion of our work is drawn in Section 5.
Notation: Throughout this paper, the mathematic notations are standard. represents the term that is induced by symmetry and is the space of square-summable vector function over . stands for -dimensional real space. means the usual Euclidean norm. The notation diag represents a diagonal matrix. If is a square matrix, stands for the maximum (minimum) eigenvalue of . (or ) is the identity (or zero) matrix. denotes the zero matrix. If , denotes the matrix . If , denotes the matrix . If , denotes the matrix . denotes the expectation of the stochastic variable . The superscribe is matrix transposition. The square matrix indicates that matrix is positive definite.
Section snippets
System description
The discrete-time PDT switched SPSs can be described as followswhere , , , and are the slow state vector, the fast state vector, the control input, the disturbance and the output vector, respectively, . is the singularly perturbed parameters. ,
Main results
The following theorem proposes a sufficient condition to guarantee the exponentially mean-square stable and dissipative performance of the closed-loop system Eq. (8) under the slow state variables feedback controller. Theorem 1 Given scalars , , , , and , matrices , , , then the closed-loop system Eq. (8) is EMSS with a desired strict dissipative performance, if there exist matrices satisfying the following conditions for
Examples
To demonstrate the effectiveness and the applicability of the design method, a numerical example and a modified an inverted pendulum model are provided in this section. Example 1 Before presenting further, let us consider that the PDT switched SPSs with two subsystems has the following parameters Set , , , , , , , ,
Conclusion
In this paper, we have investigated the asynchronous non-fragile dissipativity control problem for a class of PDT switched SPSs. Based on the mode-dependent and -dependent non-monotonic Lyapunov function, we have developed some -independent sufficient conditions, which can ensure EMSS and dissipative performance of the closed-loop system. Furthermore, the asynchronous non-fragile SSVF controller gains have been designed by a new decoupling matrix, which can be served for systems to acquire
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was supported by the Natural Science Foundation of China under Grant 61573013.
References (47)
- et al.
On the composite and reduced observer-based control of discrete two-time-scale systems
J. Frankl. Inst.
(1995) - et al.
Control design for fuzzy discrete-time singularly perturbed systems via slow state variables feedback: an LMI-based approach
Inf. Sci.
(2009) - et al.
A sliding mode approach to stabilization of nonlinear Markovian jump singularly perturbed systems
Automatica
(2018) - et al.
Dissipativity-based asynchronous fuzzy sliding mode control for T-S fuzzy hidden Markov jump systems
IEEE Trans. Cybern.
(2019) - et al.
Dissipativity-based asynchronous control of discrete-time Markov jump systems with mixed time delays
Int. J. Robust Nonlinear Control
(2018) - et al.
Non-fragile control for switched stochastic delay systems with application to water quality process
Int. J. Robust Nonlinear Control
(2014) - et al.
Filtering for switched T-S fuzzy systems with persistent dwell time
IEEE Trans. Cybern.
(2019) - et al.
Stability of switched systems with average dwell-time
Proc. 38th IEEE Conf. Decision and Control, Phoenix, AZ
(1999) - et al.
Stability and stabilization of switched linear systems with mode-dependent average dwell time
IEEE Trans. Autom. Control
(2012) - et al.
Actuator and sensor fault estimation for discrete-time switched t-s fuzzy systems with time delay
J. Frankl. Inst.
(2021)