Technical communiqueRobust control of continuous-time Markov jump linear systems☆
Introduction
In control engineering, there exists a wide class of dynamic systems subject to random abrupt variations in their structures. This is in part due to the fact that very often dynamic systems are inherently vulnerable to component failures or repairs, abrupt variations of the operating point of nonlinear plant, etc. Markovian jump linear systems (MJLSs) are the suitable mathematical models to represent the class of systems. MJLSs belong to the category of hybrid systems with finite operation models. Each operation mode corresponds to some dynamic system, and the model transitions from one to another are governed by a Markov process. The stability analysis, stabilization, optimal control and control of MJLSs have been widely studied, see Fang and Loparo (2002), do Val and Costa (2005) and Zhang, Huang, and Lam (2003) and the references therein. In Rami and Ghaoui (1996), linear matrix inequality (LMI) method is applied to compute the optimal state-feedback control law for MJLSs. Recently, the stability and the stabilizability of a class of nonlinear systems with Markovian jump and Wiener process have been theoretically addressed by Boukas (2005). Moreover, output-feedback control problems are studied in de Farias, Geromel, do Val, and Costa (2000) and do Val, Geromel, and Goncalves (2002), respectively by dynamic output feedback and incomplete observations of the Markov state.
Robust control problem for MJLSs against parameter uncertainties of systems has been one of the most challenging problems and received considerable attention from control engineers and scientists, see Ghaoui and Rami (1996), Shi and Boukas (1997), and the references therein. In Mahmoud and Shi (2002) and Shi, Boukas, and Agarwal (1999) robust state-feedback controller design approaches for discrete-time MJLSs are given, respectively, using Riccati equation approach and LMI technique. For continuous-time MJLSs subject to norm-bounded parameter uncertainties in system matrices, linear quadratic (LQ) control synthesis of MJLSs is formulated by means of LMIs in Bernard, Dufour, and Bertrand (1997) and controller design is studied in de Farias, Geromel, and do Val (2002), whereas MJLSs subject to polytopic-type parameter uncertainty have been treated in, e.g., Boukas and Liu (2000) and Pakshin and Mitrofanov (2005). With respect to the uncertainty on the transition Markov probability matrix of MJLSs, stability analysis and synthesis are respectively considered in Costa and Marques (1998), Costa, do Val, and Geromel (1999) and Karan, Shi, and Kaya (2006). For the parameter uncertainty on both the system matrices and the mode transition rate matrix, a sufficient condition for designing robust state-feedback controllers is proposed in Xiong, Lam, Gao, and Ho (2005). Most of the above-mentioned results are based on a single Lyapunov function, i.e., independent of uncertainty parameters. These methods, referred to as uncertainty-independent, are numerically appealing. However, since a common Lyapunov function is used to ensure stability for all admissible uncertainties, they can be quite conservative. In order to reduce the conservatism, by introducing slack variables to separating system matrix and Lyapunov matrix and based on a parameter-dependent Lyapunov function approach, stability analysis and stabilization of discrete-time MJLSs subject to polytopic-type parameter uncertainty are given in de Souza (2006). For continuous case, de Souza and Coutinho (2006) also give a relaxing condition based on parameter-dependent Lyapunov function for stability analysis. However, by the authors’ knowledge, the control synthesis for continuous-time MJLSs has not been fully addressed based on parameter-dependent Lyapunov function approach. For continuous-time linear systems with polytopic-type parameter uncertainty, a parameter-dependent Lyapunov function approach to stability analysis and control synthesis is given in Apkarian, Tuan, and Bernussou (2001), which is a continuous-time version of the results for discrete-time case in de Oliveira, Bernussou, and Geromel (1999). Differently from the discrete case, one cannot prove that the parameter-dependent Lyapunov function approach generalizes the one based on a single Lyapunov function. But, the parameter-dependent Lyapunov function approach potentially is less conservative because more slacking variables are involved in the LMI formulation. In this paper, a new technique of improving the one given in Apkarian et al. (2001) is developed by adding more slacking variables, and exploited to derive an LMI-based approach to robust control synthesis for continuous MJLSs via parameter-dependent Lyapunov functions. A numerical example is given to illustrate the effectiveness of the proposed approach.
This paper is organized as follows. Section 2 presents system description and problem statement. In Section 3, an LMI-based robust controller design method is presented. In contrast to the existing results, the effectiveness of the proposed method is illustrated by a numerical example in Section 4. Finally, in Section 5, conclusions are given.
Section snippets
System description and problem statement
Consider a continuous-time Markov jump linear system (MJLS) (1) on a probability space subject to polytopic-type parameter uncertainty described by the following state-space equations: where is the state vector, is the control input, is the disturbance, is the controlled output. is a homogeneous Markov chain on a finite state space
Robust synthesis
This section will deal with robust synthesis problem for continuous-time Markov jump linear systems with polytopic-type parameter uncertainties. At first, the following technical lemma is presented in this development.
Lemma 6 The following statements are equivalent: (i) There exist symmetric matrices,satisfying(6), (7), (8). (ii) There exist symmetric matrices,,and matrices,satisfying the following matrix inequalities
Example
In this section, a numerical example is presented to illustrate the effectiveness of the proposed method.
Example 10 Consider a continuous-time Markov jump linear system (MJLS) (1) with The following transition matrix is considered. The initial distribution is . The control problem of minimizing the norm for
Conclusion
In this paper, the problem of robust controller design for continuous-time Markov jump linear systems (MJLSs) has been investigated. Based on improving the existing technique for linear systems, a new design method via parameter-dependent Lyapunov functions for control synthesis for continuous-time MJLSs is given in terms of solutions to a set of LMIs. A numerical example is given to show the effectiveness of the proposed design method.
Acknowledgments
This work was supported in part by the Program for New Century Excellent Talents in University (NCET-04-0283), the Funds for Creative Research Groups of China (No. 60521003), Program for Changjiang Scholars and Innovative Research Team in University (No. IRT0421), the State Key Program of National Natural Science of China (Grant No. 60534010), the Funds of National Science of China (Grant No. 60674021) and the Funds of Ph.D. program of MOE, China (Grant No. 20060145019).
References (23)
- et al.
Continuous-time state-feedback -control of Markovian jump linear systems via convex analysis
Automatica
(1999) - et al.
A new discrete-time robust stability condition
Systems & Control Letters
(1999) - et al.
The -control for jump linear systems: Cluster observations of the Markov state
Automatica
(2002) - et al.
Transition probability bounds for the stochastic stability robustness of continuous- and discrete-time Markovian jump linear systems
Automatica
(2006) - et al.
On robust stabilization of Markovian jump systems with uncertain switching probabilities
Automatica
(2005) - et al.
model reduction of Markovian jump linear systems
Systems & Control Letters
(2003) - et al.
Continuous-time analysis, eigenstructure assignment, and synthesis with enhanced linear matrix inequalities (LMI) characterizations
IEEE Transactions on Automatic Control
(2001) - et al.
On the JLQ problem with uncertainty
IEEE Transactions on Automatic Control
(1997) Stabilization of stochastic nonlinear hybrid systems
International Journal on Innovative Computing, Information and Control
(2005)- et al.
Robust filtering for polytopic uncertain time-delay systems with Markov jumps
Mixed -control of discrete-time Markovian jump linear systems
IEEE Transactions on Automatic Control
Cited by (87)
H<inf>2</inf> and H<inf>∞</inf> Control with sector bound optimization for Continuous Time Markov Jump Lur'e Systems
2022, IFAC Journal of Systems and ControlPolytopic models for observer and fault-tolerant control designs
2019, New Trends in Observer-Based Control: An Introduction to Design Approaches and Engineering ApplicationsReliable H<inf>∞</inf> control on saturated linear Markov jump system with uncertain transition rates and asynchronous jumped actuator failure
2018, Journal of the Franklin InstituteAdaptive sliding mode controller design of Markov jump systems with time-varying actuator faults and partly unknown transition probabilities
2018, Nonlinear Analysis: Hybrid SystemsEvent-triggered H<inf>∞</inf> filtering of Markov jump systems with general transition probabilities
2017, Information SciencesBase on -Learning Pareto Optimality for Linear Itô Stochastic Systems With Markovian Jumps
2024, IEEE Transactions on Automation Science and Engineering
- ☆
This paper was not presented at any IFAC meeting. This paper was recommended for publication in revised form by Associate Editor Jie Chen under the direction of Editor André Tits.