Mixed H2/H∞ control for time‐varying and linear parametrically‐varying systems

This paper provides a solution of the mixed H 2 =H 1 problem with reduced order controller for time-varying systems in terms of solvability of diierential linear matrix inequalities and rank conditions, including a detailed discussion of how to construct a controller. Immediate specialization lead to a solution of the full order problem and the mixed H 2 =H 1 problem for linear systems whose description depend on an unknown but in real-time measurable time-varying parameter. As known for the H 1 problem, we completely resolve the quadratic mixed H 2 =H 1 problem (under certain hypotheses on the parameter dependence) by reducing it to the solution of nitely many algebraic linear matrix inequalities. However, we also point out directions how to overcome the conservatism introduced by using constant solutions of the diierential inequalities. Finally, we clarify that a simple specialization leads to a linear matrix inequality approach to the pure H 2 problem for general LTI systems, and we reveal how to compute the optimal value. 1 Notation R n is equipped with the Euclidean norm, and R nm with the corresponding induced norm, both denoted as k:k. L p denotes the signal space L n p 0; 1) (for an appropriate n) and is equipped with the standard norm k:k p (deened with the Euclidean spatial norm). Functions are tacitly assumed to be continuous and bounded, and smooth functions are, in addition, continuously diierentiable. Time functions are functions deened on 0; 1). For a symmetric valued function X(s) deened on a set S, X is said to be strictly positive (X 0) if there exists an > 0 such that X(s) I for all s 2 S. For the system or input output mapping _ x = Ax + Bu; y = Cx + Du with x(0) = 0 we use the notation " A B C D #. If A B C D ! is a constant matrix, the system is called LTI, if it is a time function, the system is called LTV. The time function A is exponentially stable if the system _ x = Ax has this property.

[1]  J. Willems Least squares stationary optimal control and the algebraic Riccati equation , 1971 .

[2]  W. Haddad,et al.  LQG control with an H∞ performance bound: a riccati equation approach , 1988, 1988 American Control Conference.

[3]  P. Khargonekar,et al.  State-space solutions to standard H2 and H∞ control problems , 1988, 1988 American Control Conference.

[4]  K. Glover,et al.  State-space formulae for all stabilizing controllers that satisfy and H ∞ norm bound and relations to risk sensitivity , 1988 .

[5]  Kemin Zhou,et al.  Optimal control with mixed H 2 and H ∞ performance objectives , 1989 .

[6]  John Doyle,et al.  Mixed H2 and H∞ control , 1990, 1990 American Control Conference.

[7]  Gilead Tadmor,et al.  Worst-case design in the time domain: The maximum principle and the standardH∞ problem , 1990, Math. Control. Signals Syst..

[8]  Carsten W. Scherer,et al.  The Riccati inequality and state-space H∞-optimal control , 1990 .

[9]  Mitsuji Sampei,et al.  An algebraic approach to H ∞ output feedback control problems , 1990 .

[10]  K. Glover,et al.  Minimum entropy H ∞ control , 1990 .

[11]  Okko H. Bosgra,et al.  Necessary conditions for static and fixed order dynamic mixed H2/ H∞ optimal control , 1991, 1991 American Control Conference.

[12]  Pramod P. Khargonekar,et al.  H2-optimal control with an H∞-constraint The state feedback case , 1991, Autom..

[13]  P. Khargonekar,et al.  Multiple objective optimal control of linear systems: the quadratic norm case , 1991 .

[14]  Michael Athans,et al.  Guaranteed properties of gain scheduled control for linear parameter-varying plants , 1991, Autom..

[15]  D. Bernstein,et al.  Mixed-norm H 2 /H ∞ regulation and estimation: the discrete-time case , 1991 .

[16]  D. Bernstein,et al.  Mixed-norm H2/H∞ regulation and estimation: the discrete-time case , 1991, 1991 American Control Conference.

[17]  Carsten Scherer,et al.  H 1 -optimization without Assumptions on Finite or Innnite Zeros , 1992 .

[18]  A. Packard,et al.  Optimal, constant I/O similarity scaling for full-information and state-feedback control problems , 1992 .

[19]  C. Scherer H∞ -control by state-feedback for plants with zeros on the imaginary axis , 1992 .

[20]  Antonie Arij Stoorvogel The singular H2 control problem , 1992, Autom..

[21]  P. Khargonekar,et al.  Exact and Approximate Solutions to a Class of Multiobjective Controller Synthesis Problems , 1993, 1993 American Control Conference.

[22]  A. Packard,et al.  Control of Parametrically-Dependent Linear Systems: A Single Quadratic Lyapunov Approach , 1993, 1993 American Control Conference.

[23]  Andrew Packard,et al.  The complex structured singular value , 1993, Autom..

[24]  A. Packard,et al.  Robust performance of linear parametrically varying systems using parametrically-dependent linear feedback , 1994 .

[25]  B. Anderson,et al.  A Nash game approach to mixed H2/H∞ control , 1994, IEEE Transactions on Automatic Control.

[26]  Stephen P. Boyd,et al.  Linear Matrix Inequalities in Systems and Control Theory , 1994 .

[27]  A. Packard Gain scheduling via linear fractional transformations , 1994 .

[28]  M. Rotea,et al.  An alternative to the D-K iteration? , 1994, Proceedings of 1994 American Control Conference - ACC '94.

[29]  Tetsuya Iwasaki,et al.  All controllers for the general H∞ control problem: LMI existence conditions and state space formulas , 1994, Autom..

[30]  Alan J. Laub,et al.  The LMI control toolbox , 1994, Proceedings of 1994 33rd IEEE Conference on Decision and Control.

[31]  P. Gahinet,et al.  A convex characterization of gain-scheduled H∞ controllers , 1995, IEEE Trans. Autom. Control..

[32]  Fen Wu,et al.  Induced L2‐norm control for LPV systems with bounded parameter variation rates , 1996 .

[33]  Pierre Apkarian,et al.  Self-scheduled H∞ control of linear parameter-varying systems: a design example , 1995, Autom..

[34]  Carsten SchererMechanical,et al.  Mixed H 2 =h 1 Control , 1995 .