Journal of Applied Mathematics
VolumeΒ 2009Β (2009), Article IDΒ 169790, 16 pages
doi:10.1155/2009/169790
Research Article

Generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -Pairs for Uncertain Linear Infinite-Dimensional Systems

1Division of Science, School of Science and Engineering, Tokyo Denki University, Hatoyama-Machi, Hiki-Gun, Saitama 350-0394, Japan
2Department of Management and Information Sciences, Nagasaki Institute of Applied Science, Aba-Machi 536, Nagasaki 851-0193, Japan

Received 8 June 2009; Accepted 15 September 2009

Academic Editor: M. A.Β Petersen

Copyright Β© 2009 Naohisa Otsuka and Haruo Hinata. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

We introduce the concept of generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pairs which is related to generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant subspaces and generalized 𝑆 ( 𝐢 , 𝐴 ) -invariant subspaces for infinite-dimensional systems. As an application the parameter-insensitive disturbance-rejection problem with dynamic compensator is formulated and its solvability conditions are presented. Further, an illustrative example is also examined.

1. Introduction

In the framework of the so-called geometric approach, many control problems with state feedback and/or incomplete-state feedback (e.g., controllability and observability problems, decoupling problems, and disturbance-rejection problems, etc.) have been studied for finite-dimensional systems (see, e.g., [1, 2]). Further, the concept of ( 𝐢 , 𝐴 , 𝐡 ) -pairs was first introduced by Schumacher [3], and this concept has been used successfully to design dynamic compensators. After that Curtain extended the geometric concepts to infinite-dimensional systems and various control problems have been studied (see, e.g., [411]). On the other hand, from the practical viewpoint, Ghosh [12] and Otsuka [13] studied the concepts of simultaneous ( 𝐢 , 𝐴 , 𝐡 ) -pairs and of generalized ( 𝐢 , 𝐴 , 𝐡 ) -pairs, respectively, for finite-dimensional systems, and the parameter-insensitive disturbance-rejection problems for uncertain linear systems were studied. Then, Otsuka and Inaba [1416] extended the concepts of simultaneous invariant subspaces and simultaneous ( 𝐢 , 𝐴 , 𝐡 ) -pairs to infinite-dimensional systems. Further, Otsuka and Hinata [17] studied the concept of generalized invariant subspaces for infinite-dimensional systems.

The objective of this paper is to investigate the concept of generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pairs for infinite-dimensional systems and to study the parameter-insensitive disturbance-rejection problem with dynamic compensator.

The paper is organized as follows. Section 2 gives the concept of generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pairs and its properties. In Section 3, the parameter-insensitive disturbance-rejection problem with dynamic compensator is formulated and its solvability conditions are presented. Section 4 gives an example to illustrate our results. Finally, some concluding remarks are given in Section 5.

2. Generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pairs

First, we give some notations used throughout this investigation. Let 𝐁 ( 𝒳 ; 𝒴 ) denote the set of all bounded linear operators from a Hilbert space 𝒳 into another Hilbert space 𝒴 ; for notational simplicity, we write 𝐁 ( 𝒳 ) for 𝐁 ( 𝒳 ; 𝒳 ) . For a linear operator 𝐴 the domain, the image, the kernel, and the 𝐢 0 -semigroup generated by 𝐴 are denoted by D ( 𝐴 ) , Im 𝐴 , K e r 𝐴 , and { 𝑆 𝐴 ( 𝑑 ) ; 𝑑 β‰₯ 0 } , respectively. Further, the dimension and the orthogonal complement of a closed subspace 𝒱 are denoted by dim( 𝒱 ) and ( 𝒱 ) βŸ‚ , respectively.

Next, consider the following linear systems defined in a Hilbert space 𝒳 :

ξƒ― 𝑑 𝑆 ( 𝛼 , 𝛽 , 𝛾 ) ∢ 𝑑 𝑑 π‘₯ ( 𝑑 ) = 𝐴 ( 𝛼 ) π‘₯ ( 𝑑 ) + 𝐡 ( 𝛽 ) 𝑒 ( 𝑑 ) , 𝑦 ( 𝑑 ) = 𝐢 ( 𝛾 ) π‘₯ ( 𝑑 ) , ( 2 . 1 ) where π‘₯ ( 𝑑 ) ∈ 𝒳 , 𝑒 ( 𝑑 ) ∈ 𝒰 ∢ = 𝐑 π‘š , 𝑦 ( 𝑑 ) ∈ 𝒴 ∢ = 𝐑 β„“ are the state, the input, and the measurement output, respectively. Operators 𝐴 ( 𝛼 ) , 𝐡 ( 𝛽 ) , and 𝐢 ( 𝛾 ) are unknown in the sense that they are represented as the forms: 𝐴 ( 𝛼 ) = 𝐴 0 + 𝛼 1 𝐴 1 + β‹― + 𝛼 𝑝 𝐴 𝑝 ∢ = 𝐴 0 + Ξ” 𝐴 ( 𝛼 ) , 𝐡 ( 𝛽 ) = 𝐡 0 + 𝛽 1 𝐡 1 + β‹― + 𝛽 π‘ž 𝐡 π‘ž ∢ = 𝐡 0 + Ξ” 𝐡 ( 𝛽 ) , 𝐢 ( 𝛾 ) = 𝐢 0 + 𝛾 1 𝐢 1 + β‹― + 𝛾 π‘Ÿ 𝐢 π‘Ÿ ∢ = 𝐢 0 + Ξ” 𝐢 ( 𝛾 ) , ( 2 . 2 ) where 𝛼 := ( 𝛼 1 , … , 𝛼 𝑝 ) ∈ 𝐑 𝑝 , 𝛽 := ( 𝛽 1 , … , 𝛽 π‘ž ) ∈ 𝐑 π‘ž , 𝛾 := ( 𝛾 1 , … , 𝛾 π‘Ÿ ) ∈ 𝐑 π‘Ÿ , 𝐴 0 is the infinitesimal generator of a 𝐢 0 -semigroup { 𝑆 𝐴 0 ( 𝑑 ) ; 𝑑 β‰₯ 0 } on 𝒳 , 𝐴 𝑖 ∈ 𝐁 ( 𝒳 ) ( 𝑖 = 1 , … , 𝑝 ) , 𝐡 𝑖 ∈ 𝐁 ( 𝐑 π‘š ; 𝒳 ) ( 𝑖 = 0 , … , π‘ž ) , and 𝐢 𝑖 ∈ 𝐁 ( 𝒳 ; 𝐑 β„“ ) ( 𝑖 = 0 , … , π‘Ÿ ) . Here, in the system 𝑆 ( 𝛼 , 𝛽 , 𝛾 ) ( 𝐴 0 , 𝐡 0 , 𝐢 0 ) and ( Ξ” 𝐴 ( 𝛼 ) , Ξ” 𝐡 ( 𝛽 ) , Ξ” 𝐢 ( 𝛾 ) ) mean the nominal system model and a specific uncertain perturbation, respectively.

Since 𝐴 𝑖 ( 𝑖 = 1 , … , 𝑝 ) are bounded linear operators, we remark that 𝐴 ( 𝛼 ) always generates a 𝐢 0 -semigroup and has the domain D ( 𝐴 ( 𝛼 ) ) = D ( 𝐴 0 ) for all 𝛼 ∈ 𝐑 𝑝 . Further, from the practical viewpoint it is assumed that the dimensions of input and output are finite.

Now, introduce a compensator ( 𝐾 , 𝐿 , 𝑀 , 𝑁 ) defined in a Hilbert space 𝒲 of the form : ξƒ― 𝑑 Ξ£ ∢ 𝑑 𝑑 𝑀 ( 𝑑 ) = 𝑁 𝑀 ( 𝑑 ) + 𝑀 𝑦 ( 𝑑 ) , 𝑒 ( 𝑑 ) = 𝐿 𝑀 ( 𝑑 ) + 𝐾 𝑦 ( 𝑑 ) , ( 2 . 3 ) where 𝑁 is the infinitesimal generator of a 𝐢 0 -semigroup { 𝑆 𝑁 ( 𝑑 ) ; 𝑑 β‰₯ 0 } on a Hilbert space 𝒲 with the domain 𝐷 ( 𝑁 ) = 𝒲 , 𝑀 ∈ 𝐁 ( 𝐑 β„“ ; 𝒲 ) , 𝐿 ∈ 𝐁 ( 𝒲 ; 𝐑 π‘š ) , and 𝐾 ∈ 𝐁 ( 𝐑 β„“ ; 𝐑 π‘š ) .

If a compensator of the form Ξ£ is applied to the system 𝑆 ( 𝛼 , 𝛽 , 𝛾 ) , the resulting closed-loop system 𝑆 c l ( 𝛼 , 𝛽 , 𝛾 ) with the extended state space 𝒳 𝑒 ∢ = 𝒳 βŠ• 𝒲 is easily seen to be 𝑑 ξ‚Έ ξ‚Ή = ξ‚Έ ξ‚Ή , 𝑑 𝑑 π‘₯ ( 𝑑 ) 𝑀 ( 𝑑 ) 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐾 𝐢 ( 𝛾 ) 𝐡 ( 𝛽 ) 𝐿 𝑀 𝐢 ( 𝛾 ) 𝑁 ξ‚Ή ξ‚Έ π‘₯ ( 𝑑 ) 𝑀 ( 𝑑 ) ( 2 . 4 ) where 𝒳 βŠ• 𝒲 means the direct sum of 𝒳 and 𝒲 . For the closed-loop system 𝑆 c l ( 𝛼 , 𝛽 , 𝛾 ) , define π‘₯ 𝑒 ξ‚Έ ξ‚Ή ( 𝑑 ) ∢ = π‘₯ ( 𝑑 ) 𝑀 ( 𝑑 ) , 𝐴 𝑒 𝛼 𝛽 𝛾 ξ‚Έ ξ‚Ή ∢ = 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐾 𝐢 ( 𝛾 ) 𝐡 ( 𝛽 ) 𝐿 𝑀 𝐢 ( 𝛾 ) 𝑁 ( 2 . 5 ) with domain D ( 𝐴 𝑒 𝛼 𝛽 𝛾 ) ( = D ( 𝐴 0 ) βŠ• 𝒲 ) .

For the system 𝑆 ( 𝛼 , 𝛽 , 𝛾 ) , we give the following invariant subspaces.

Definition 2.1. Let 𝒱 be a closed subspace of 𝒳 .
(i) 𝒱 is said to be a generalized ( 𝐴 , 𝐡 ) -invariant if there exists an 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that ξ€· ( 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ) 𝒱 ∩ D ξ€· 𝐴 0 ξ€Έ ξ€Έ βŠ‚ 𝒱 , βˆ€ 𝛼 , 𝛽 . ( 2 . 6 ) Also 𝐅 ( 𝒱 ) ∢ = { 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) ∣ ( 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ) ( 𝒱 ∩ D ( 𝐴 0 ) ) βŠ‚ 𝒱 for all 𝛼 , 𝛽 } .
(ii) 𝒱 is said to be a generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant if there exists an 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , βˆ€ 𝑑 β‰₯ 0 a n d a l l 𝛼 , 𝛽 . ( 2 . 7 ) Also 𝒱 ( 𝐴 , 𝐡 ; Ξ› ) ∢ = { 𝒱 ∣ 𝒱 is a generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant and is contained in a given closed subspace Ξ› . } . 𝐅 𝑠 ( 𝒱 ) ∢ = { 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) ∣ 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 f o r a l l 𝑑 β‰₯ 0 a n d a l l 𝛼 , 𝛽 } .
(iii) 𝒱 is said to be a generalized ( 𝐢 , 𝐴 ) -invariant if there exists a 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that ξ€· ( 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ) 𝒱 ∩ D ξ€· 𝐴 0 ξ€Έ ξ€Έ βŠ‚ 𝒱 , βˆ€ 𝛼 , 𝛾 . ( 2 . 8 ) Also 𝐆 ( 𝒱 ) ∢ = { 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) ∣ ( 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ) ( 𝒱 ∩ D ( 𝐴 0 ) ) βŠ‚ 𝒱 for all 𝛼 , 𝛾 } .
(iv) 𝒱 is said to be a generalized 𝑆 ( 𝐢 , 𝐴 ) -invariant if there exists a 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , βˆ€ 𝑑 β‰₯ 0 a n d a l l 𝛼 , 𝛾 . ( 2 . 9 ) Also 𝒱 ( πœ€ ; 𝐢 , 𝐴 ) ∢ = { 𝒱 ∣ 𝒱 is a generalized 𝑆 ( 𝐢 , 𝐴 ) -invariant and contains a given closed subspace πœ€ } . 𝐆 𝑠 ( 𝒱 ) ∢ = { 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) ∣ 𝑆 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 and all 𝛼 , 𝛾 } .

Remark 2.2. (i) For the system 𝑆 ( 𝛼 , 𝛽 , 𝛾 ) a generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant subspace 𝒱 has the property that if an arbitrary initial state π‘₯ ( 0 ) ∈ 𝒱 , then there exists a state feedback 𝑒 ( 𝑑 ) = 𝐹 π‘₯ ( 𝑑 ) which is independent of 𝛼 and 𝛽 such that the state trajectory π‘₯ ( 𝑑 ) ∈ 𝒱 for all 𝑑 β‰₯ 0 .
(ii) If 𝐴 0 is a bounded linear operator on 𝒳 (i.e., 𝐴 0 ∈ 𝐁 ( 𝒳 ) ), then the statements (i), (ii) and (iii), (iv) in Definition 2.1 are equivalent, respectively. Further, in this case 𝐅 𝑠 ( 𝒱 ) = 𝐅 ( 𝒱 ) and 𝐆 𝑠 ( 𝒱 ) = 𝐆 ( 𝒱 ) .

Theorem 2.3 (see [17, 18]). Suppose that 𝑝 𝑖 , π‘ž 𝑖 ( 𝑖 = 1 , … , 𝑝 ) , π‘Ÿ 𝑖 , 𝑠 𝑖 ( 𝑖 = 1 , … , π‘ž ) are arbitrary fixed real numbers such that 𝑝 𝑖 < π‘ž 𝑖 ( 𝑖 = 1 , … , 𝑝 ) and π‘Ÿ 𝑖 < 𝑠 𝑖 ( 𝑖 = 1 , … , π‘ž ) . Then, the following three statements are equivalent. (i) 𝒱 is a generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant.(ii)There exists an 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝑆 𝐴 0 + 𝐡 0 𝐹 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 ( 𝑑 β‰₯ 0 ) and 𝐡 𝑖 𝐹 𝒱 βŠ‚ 𝒱 ( 𝑖 = 1 , … , π‘ž ) , and 𝐴 𝑖 𝒱 βŠ‚ 𝒱 ( 𝑖 = 1 , … , 𝑝 ) . (iii)There exists an 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , ( 𝑑 β‰₯ 0 ) ( 2 . 1 0 ) for all 𝛼 𝑖 ∈ [ 𝑝 𝑖 , π‘ž 𝑖 ] ( 𝑖 = 1 , … , 𝑝 ) and 𝛽 𝑖 ∈ [ π‘Ÿ 𝑖 , 𝑠 𝑖 ] ( 𝑖 = 1 , … , π‘ž ) . (iv)There exists an 𝐹 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , ( 𝑑 β‰₯ 0 ) ( 2 . 1 1 ) for all 𝛼 𝑖 ∈ { 𝑝 𝑖 , π‘ž 𝑖 } ( 𝑖 = 1 , … , 𝑝 ) and 𝛽 𝑖 ∈ { π‘Ÿ 𝑖 , 𝑠 𝑖 } ( 𝑖 = 1 , … , π‘ž ) .

The following theorem is the dual version of Theorem 2.3.

Theorem 2.4 (see [17, 18]). Suppose that 𝑝 𝑖 , π‘ž 𝑖 ( 𝑖 = 1 , … , 𝑝 ) , 𝑑 𝑖 , 𝑒 𝑖 ( 𝑖 = 1 , … , π‘Ÿ ) are arbitrary fixed real numbers such that 𝑝 𝑖 < π‘ž 𝑖 ( 𝑖 = 1 , … , 𝑝 ) and 𝑑 𝑖 < 𝑒 𝑖 ( 𝑖 = 1 , … , π‘Ÿ ) . Then, the following three statements are equivalent. (i) 𝒱 is a generalized 𝑆 ( 𝐢 , 𝐴 ) -invariant.(ii)There exists a 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that 𝑆 𝐴 0 + 𝐺 𝐢 0 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 ( 𝑑 β‰₯ 0 ) and 𝐺 𝐢 𝑖 𝒱 βŠ‚ 𝒱 ( 𝑖 = 1 , … , π‘Ÿ ) , and 𝐴 𝑖 𝒱 βŠ‚ 𝒱 ( 𝑖 = 1 , … , 𝑝 ) . (iii)There exists a 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , ( 𝑑 β‰₯ 0 ) ( 2 . 1 2 ) for all 𝛼 𝑖 ∈ [ 𝑝 𝑖 , π‘ž 𝑖 ] ( 𝑖 = 1 , … , 𝑝 ) and 𝛾 𝑖 ∈ [ 𝑑 𝑖 , 𝑒 𝑖 ] ( 𝑖 = 1 , … , π‘Ÿ ) . (iv)There exists a 𝐺 ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that 𝑆 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ( 𝑑 ) 𝒱 βŠ‚ 𝒱 , ( 𝑑 β‰₯ 0 ) ( 2 . 1 3 ) for all 𝛼 𝑖 ∈ { 𝑝 𝑖 , π‘ž 𝑖 } ( 𝑖 = 1 , … , 𝑝 ) and 𝛾 𝑖 ∈ { 𝑑 𝑖 , 𝑒 𝑖 } ( 𝑖 = 1 , … , π‘Ÿ ) .

For finite-dimensional systems, Schumacher [3] first introduced the concept of ( 𝐢 , 𝐴 , 𝐡 ) -pair. The following definition is a generalized and infinite-dimensional version of ( 𝐢 , 𝐴 , 𝐡 ) -pair.

Definition 2.5. Let 𝒱 1 and 𝒱 2 be closed subspaces of 𝒳 . A pair ( 𝒱 1 , 𝒱 2 ) of subspaces is said to be a generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pair if the following three conditions hold.(i) 𝒱 1 is a generalized 𝑆 ( 𝐢 , 𝐴 ) -invariant.(ii) 𝒱 2 is a generalized 𝑆 ( 𝐴 , 𝐡 ) -invariant.(iii) 𝒱 1 βŠ‚ 𝒱 2 .

For closed-loop system 𝑆 c l ( 𝛼 , 𝛽 , 𝛾 ) , we give the following definition.

Definition 2.6. Let 𝒱 𝑒 be a closed subspace of 𝒳 𝑒 .(i) 𝒱 𝑒 is said to be a generalized 𝐴 𝑒 -invariant if 𝐴 𝑒 𝛼 , 𝛽 , 𝛾 ( 𝒱 𝑒 ∩ D ( 𝐴 𝑒 𝛼 , 𝛽 , 𝛾 ) ) βŠ‚ 𝒱 𝑒 for all 𝛼 , 𝛽 , 𝛾 . (ii) 𝒱 𝑒 is said to be a generalized 𝑆 𝐴 𝑒 ( 𝑑 ) -invariant if 𝑆 𝐴 𝑒 𝛼 , 𝛽 , 𝛾 ( 𝑑 ) 𝒱 𝑒 βŠ‚ 𝒱 𝑒 for all 𝑑 β‰₯ 0 and all 𝛼 , 𝛽 , 𝛾 .

The following lemma was shown by Zwart.

Lemma 2.7 (see [11]). Let 𝒱 𝑒 be a closed subspace of 𝒳 𝑒 and the following three subspaces are introduced: 𝑆 o r t h ξƒ― | | | | ξ‚Έ π‘₯ 𝑀 ξ‚Ή ∈ ξ€Ί 𝒱 ∢ = π‘₯ ∈ 𝒳 𝑒 ξ€» βŸ‚ f o r s o m e ξƒ° 𝑀 ∈ 𝒲 = 𝑃 𝒳 ξ‚€ ξ€Ί 𝒱 𝑒 ξ€» βŸ‚  , 𝑆 1 ξ€Ί 𝑆 ∢ = o r t h ξ€» βŸ‚ , 𝑆 2 ξƒ― | | | | ξ‚Έ π‘₯ 𝑀 ξ‚Ή ∈ ξ€Ί 𝒱 ∢ = π‘₯ ∈ 𝒳 𝑒 ξ€» f o r s o m e ξƒ° 𝑀 ∈ 𝒲 = 𝑃 𝒳 𝒱 ξ€· ξ€Ί 𝑒 , ξ€» ξ€Έ ( 2 . 1 4 ) where 𝑃 𝒳 is the projection operator from 𝒳 𝑒 onto 𝒳 along 𝒲 . Then, the following statements hold. (i) 𝑆 1 = { π‘₯ ∈ 𝒳 | [ π‘₯ 0 ] ∈ 𝒱 𝑒 } . (ii) 𝑆 1 βŠ‚ 𝑆 2 . (iii)If d i m ( 𝒲 ) < ∞ , then 𝑆 2 is a closed subspace of 𝒳 and d i m ( 𝑆 2 ∩ 𝑆 βŸ‚ 1 ) < ∞ .

Lemma 2.8 (see [4, 6, 11]). Suppose that 𝐴 is the infinitesimal generator of a 𝐢 0 -semigroup { 𝑆 𝐴 ( 𝑑 ) ; 𝑑 β‰₯ 0 } on 𝒳 , and 𝒱 is a closed subspace of 𝒳 and 𝑄 1 ∈ 𝐁 ( 𝒳 ) . Then, the following statements hold. (i)If 𝑆 𝐴 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 , then 𝐴 ( 𝒱 ∩ D ( 𝐴 ) ) βŠ‚ 𝒱 . (ii)If 𝒱 βŠ‚ D( 𝐴 ) and 𝐴 𝒱 βŠ‚ 𝒱 , then 𝑆 𝐴 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 . (iii)If 𝑆 𝐴 + 𝑄 1 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 , then 𝒱 ∩ D ( 𝐴 ) = 𝒱 . (iv)If there exists a 𝑄 2 ∈ 𝐁 ( 𝒳 ) such that 𝑆 𝐴 + 𝑄 2 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 and ( 𝑄 1 βˆ’ 𝑄 2 ) ( 𝒱 ∩ D ( 𝐴 ) ) βŠ‚ 𝒱 , then 𝑆 𝐴 + 𝑄 1 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 . (v)If there exists a 𝑄 2 ∈ 𝐁 ( 𝒳 ) such that 𝑆 𝐴 + 𝑄 2 ( 𝑑 ) 𝒱 βŠ‚ 𝒱 for all 𝑑 β‰₯ 0 and ( 𝑄 1 βˆ’ 𝑄 2 ) ( 𝒱 ∩ D ( 𝐴 ) ) = { 𝟎 } , then 𝑆 𝐴 + 𝑄 1 ( 𝑑 ) π‘₯ = 𝑆 𝐴 + 𝑄 2 ( 𝑑 ) π‘₯ for all 𝑑 β‰₯ 0 and all π‘₯ ∈ 𝒱 .

The following two lemmas are extensions of the results of Otsuka [13] to infinite-dimensional systems.

Lemma 2.9. If a pair ( 𝒱 1 , 𝒱 2 ) of subspaces of 𝒳 is a generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pair such that π‘ž  𝑖 = 1 I m 𝐡 𝑖 βŠ‚ 𝒱 1 βŠ‚ 𝒱 2 βŠ‚ π‘Ÿ  𝑖 = 1 K e r 𝐢 𝑖 , 𝐴 𝑖 𝒱 2 βŠ‚ 𝒱 1 , ( 𝑖 = 1 , … , 𝑝 ) , ( 2 . 1 5 ) then there exist 𝐺 ∈ 𝐆 𝑠 ( 𝒱 1 ) , 𝐺 ( 𝛽 ) ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) , 𝐹 ( 𝛾 ) ∈ 𝐅 𝑠 ( 𝒱 2 ) , 𝐹 0 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) and 𝐾 ∈ 𝐁 ( 𝐑 β„“ ; 𝐑 π‘š ) such that 𝐺 = 𝐡 ( 𝛽 ) 𝐾 + 𝐺 ( 𝛽 ) , I m 𝐺 ( 𝛽 ) βŠ‚ 𝒱 2 , 𝐹 ( 𝛾 ) = 𝐾 𝐢 ( 𝛾 ) + 𝐹 0 , K e r 𝐹 0 βŠƒ 𝒱 1 ( 2 . 1 6 ) for all ( 𝛽 , 𝛾 ) ∈ 𝐑 π‘ž Γ— 𝐑 π‘Ÿ .

Proof. Suppose that a pair ( 𝒱 1 , 𝒱 2 ) is a generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pair satisfying the stated above conditions. Since βˆ‘ π‘ž 𝑖 = 1 I m 𝐡 𝑖 βŠ‚ 𝒱 2 , we remark that 𝒱 2 + I m 𝐡 ( 𝛽 ) = 𝒱 2 + I m 𝐡 0 .
Claim 1.  𝐺 𝐢 ( 𝛾 ) 𝒱 1 βŠ‚ 𝒱 2 + I m 𝐡 0 for all  𝐺 ∈ 𝐆 𝑠 ( 𝒱 1 ) and 𝛾 ∈ 𝐑 π‘Ÿ .
To prove Claim 1, choose an arbitrary element π‘₯ ∈ 𝒱 1 . Then, by Lemma 2.8(iii) there exists an π‘₯ 𝑛 ∈ 𝒱 1 ∩ 𝐷 ( 𝐴 π‘œ ) such that l i m 𝑛 β†’ ∞ π‘₯ 𝑛 = π‘₯ . ( 2 . 1 7 ) Now, noticing that  ( 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ) ( 𝒱 1 ∩ 𝐷 ( 𝐴 0 ) ) βŠ‚ 𝒱 1 and 𝒱 1 βŠ‚ 𝒱 2 we have  𝐺 𝐢 ( 𝛾 ) π‘₯ 𝑛 = ξ‚€   π‘₯ 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) 𝑛 βˆ’ 𝐴 ( 𝛼 ) π‘₯ 𝑛 ∈ 𝒱 1 + 𝒱 2 + I m 𝐡 ( 𝛽 ) = 𝒱 2 + I m 𝐡 ( 𝛽 ) , = 𝒱 2 + I m 𝐡 0 . ( 2 . 1 8 ) Since dim 𝐡 ( 𝛽 ) < ∞ , that 𝒱 2 + I m 𝐡 0 is a closed subspace. Further, noticing that  𝐺 𝐢 ( 𝛾 ) are bounded operators,  𝐺 𝐢 ( 𝛾 ) π‘₯ = l i m 𝑛 β†’ ∞  𝐺 𝐢 ( 𝛾 ) π‘₯ 𝑛 ∈ 𝒱 2 + I m 𝐡 0 , ( 2 . 1 9 ) which proves Claim 1.

Next, Claims 2 and 3 hold as follows.
Claim 2. There exists a 𝐺 ∈ 𝐆 𝑠 ( 𝒱 1 ) such that I m 𝐺 βŠ‚ 𝒱 2 + I m 𝐡 0 .
To prove Claim 2, choose a  𝐺 ∈ 𝐆 𝐬 ( 𝒱 1 ) and π‘₯ ( = 𝑦 + 𝑧 ) ∈ 𝐑 β„“ such that βˆ‘ 𝑦 ∈ π‘Ÿ 𝑖 = 0 𝐢 𝑖 𝒱 1 and 𝑧 ∈ πœ™ with βˆ‘ π‘Ÿ 𝑖 = 0 𝐢 𝑖 𝒱 1 βŠ• πœ™ = 𝐑 β„“ . Define a linear map 𝐺 ∈ 𝐑 𝑛 Γ— β„“ by  𝐺 π‘₯ ∢ = 𝐺 𝑦 . Then, for some π‘₯ 𝑖 ∈ 𝒱 1  𝐺 π‘₯ = 𝐺 𝑦 = π‘Ÿ  𝑖 = 0  𝐺 𝐢 𝑖 π‘₯ 𝑖 ∈ 𝒱 2 + I m 𝐡 0 , ( b y C l a i m 1 ) , ( 2 . 2 0 ) which proves Claim 2.

Claim 3. There exists a 𝐾 ∈ 𝐁 ( 𝐑 β„“ ; 𝐑 π‘š ) and 𝐺 ( 𝛽 ) ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) such that 𝐺 = 𝐡 ( 𝛽 ) 𝐾 + 𝐺 ( 𝛽 ) , I m 𝐺 ( 𝛽 ) βŠ‚ 𝒱 2 for all 𝛽 ∈ 𝐑 π‘ž .
To prove Claim 3, let { 𝑦 1 , … , 𝑦 β„“ } be a basis of 𝐑 β„“ . Then, it follows from Claim 2 that there exists an π‘₯ 𝑖 ∈ 𝒱 2 and 𝑒 𝑖 ∈ 𝐑 π‘š such that 𝐺 𝑦 𝑖 = π‘₯ 𝑖 + 𝐡 0 𝑒 𝑖 = π‘₯ 𝑖 βˆ’ π‘ž  𝑖 = 1 𝛽 𝑖 𝐡 𝑖 𝑒 𝑖 + 𝐡 0 𝑒 𝑖 + π‘ž  𝑖 = 1 𝛽 𝑖 𝐡 𝑖 𝑒 𝑖 =  π‘₯ 𝑖 βˆ’ π‘ž  𝑖 = 1 𝛽 𝑖 𝐡 𝑖 𝑒 𝑖 ξƒͺ + 𝐡 ( 𝛽 ) 𝑒 𝑖 . ( 2 . 2 1 ) Define linear maps 𝐾 ∈ 𝐁 ( 𝐑 β„“ ; 𝐑 π‘š ) and 𝐺 ( 𝛽 ) ∈ 𝐁 ( 𝐑 β„“ ; 𝒳 ) by 𝐾 𝑦 𝑖 ∢ = 𝑒 𝑖 , 𝐺 ( 𝛽 ) 𝑦 𝑖 ∢ = π‘₯ 𝑖 βˆ’ π‘ž  𝑖 = 1 𝛽 𝑖 𝐡 𝑖 𝑒 𝑖 , r e s p e c t i v e l y . ( 2 . 2 2 ) Then, 𝐺 𝑦 𝑖 = 𝐺 ( 𝛽 ) 𝑦 𝑖 + 𝐡 ( 𝛽 ) 𝐾 𝑦 𝑖 , 𝐺 ( 𝛽 ) 𝑦 𝑖 ∈ 𝒱 2 , ( 2 . 2 3 ) which proves Claim 3.

Claim 4. There exists an 𝐹 † ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that Im ( 𝐺 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 † ) | 𝒱 2 βŠ‚ 𝒱 2 .
In fact, it follows from Claim 2 and hypotheses of this lemma that there exists a 𝐺 ∈ 𝐆 𝑠 ( 𝒱 1 ) such that
I m 𝐺 𝐢 0 | 𝒱 2 βŠ‚ I m 𝐺 βŠ‚ 𝒱 2 + I m 𝐡 0 . ( 2 . 2 4 ) Let 𝑦 ∈ 𝒱 2 be an arbitrary element. Then, there exist π‘₯ ∈ 𝒱 2 and 𝑒 ∈ 𝐑 π‘š such that 𝐺 𝐢 0 𝑦 = π‘₯ + 𝐡 0 𝑒 . Define 𝐹 † ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝐹 † 𝑦 = βˆ’ 𝑒 , ( 2 . 2 5 ) Hence, 𝐺 𝐢 0 𝑦 = π‘₯ + 𝐡 0 ( βˆ’ 𝐹 † 𝑦 ) which implies ( 𝐺 𝐢 0 + 𝐡 0 𝐹 † ) 𝑦 = π‘₯ ∈ 𝒱 2 . Then, we can easily obtain I m ξ€· 𝐺 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 † ξ€Έ | 𝒱 2 βŠ‚ 𝒱 2 , ( 2 . 2 6 ) which proves Claim 4.

Now, choose 𝐹 βˆ— ∈ 𝐅 𝑠 ( 𝒱 2 ) and define 𝐹 0 ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that
𝐹 0 = 𝐹 βˆ— + 𝐹 † , o n 𝒱 βŸ‚ 1 , 𝐹 0 = 0 , o n 𝒱 1 . ( 2 . 2 7 ) Then, the following claim holds.
Claim 5. One has ( 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 0 ) ( 𝒱 2 ∩ 𝐷 ( 𝐴 0 ) ) βŠ‚ 𝒱 2 for all ( 𝛼 , 𝛽 , 𝛾 ) ∈ 𝐑 𝑝 Γ— 𝐑 π‘ž Γ— 𝐑 π‘Ÿ and 𝒱 1 βŠ‚ K e r 𝐹 0 .
In fact, at first, 𝒱 1 βŠ‚ K e r 𝐹 0 is obvious. Therefore, we prove the first one. Since 𝐹 βˆ— ∈ 𝐅 𝑠 ( 𝒱 2 ) implies 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 βˆ— ( 𝑑 ) 𝒱 2 βŠ‚ 𝒱 2 , it follows from Lemma 2.8(v) that it suffices to show ( 𝐡 ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ) 𝒱 2 βŠ‚ 𝒱 2 in order to prove 𝑆 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) ( 𝑑 ) 𝒱 2 βŠ‚ 𝒱 2 .
Now, let π‘₯ ∈ 𝒱 2 be an arbitrary element. Then, there exists 𝑦 ∈ 𝒱 1 and 𝑧 ∈ 𝒱 βŸ‚ 1 ∩ 𝒱 2 such that π‘₯ = 𝑦 + 𝑧 . Then, we have ξ€· 𝐡 ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ ξ€· 𝐡 π‘₯ = ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ 𝑦 + ξ€· 𝐡 ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ 𝑧 = ξ€· 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ ξ€· 𝑦 + 𝐡 ( 𝛽 ) 𝐹 † ξ€Έ + 𝐺 𝐢 ( 𝛾 ) 𝑧 . ( 2 . 2 8 ) Now, it follows from Lemma 2.8(iii) that there exists a 𝑦 𝑛 such that 𝑦 𝑛 β†’ 𝑦 .
Hence, ξ€· 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ 𝑦 𝑛 = ( 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) ) 𝑦 𝑛 βˆ’ ξ€· 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ 𝑦 𝑛 ∈ 𝒱 2 . ( 2 . 2 9 ) Noticing that ( 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ) is bounded linear operator, it follows from (2.28) and Claim 4 that ξ€· 𝐡 ( 𝛽 ) 𝐹 0 + 𝐺 𝐢 ( 𝛾 ) βˆ’ 𝐡 ( 𝛽 ) 𝐹 βˆ— ξ€Έ π‘₯ ∈ 𝒱 2 , ( 2 . 3 0 ) which proves Claim 5.

Finally, define a bounded linear operator 𝐹 ( 𝛾 ) ∈ 𝐁 ( 𝒳 ; 𝐑 π‘š ) such that 𝐹 ( 𝛾 ) ∢ = 𝐾 𝐢 ( 𝛾 ) + 𝐹 0 . Then, the following claim holds.
Claim 6 ( 𝐹 ( 𝛾 ) ∈ 𝐅 ( 𝒱 2 ) ). In fact, ξ€· 𝒱 ( 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐹 ( 𝛾 ) ) 2 ξ€· 𝐴 ∩ 𝐷 0 = ξ€· ξ€Έ ξ€Έ 𝐴 ( 𝛼 ) + 𝐡 ( 𝛽 ) 𝐾 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 0 𝒱 ξ€Έ ξ€· 2 ξ€· 𝐴 ∩ 𝐷 0 = ξ€½ 𝐴 ξ€Έ ξ€Έ ( 𝛼 ) + ( 𝐡 ( 𝛽 ) 𝐾 + 𝐺 ( 𝛽 ) ) 𝐢 ( 𝛾 ) βˆ’ 𝐺 ( 𝛽 ) 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 0 𝒱 ξ€Ύ ξ€· 2 ξ€· 𝐴 ∩ 𝐷 0 = ξ€½ ξ€Έ ξ€Έ 𝐴 ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 0 𝒱 βˆ’ 𝐺 ( 𝛽 ) 𝐢 ( 𝛾 ) ξ€Ύ ξ€· 2 ξ€· 𝐴 ∩ 𝐷 0 ξ€Έ ξ€Έ , ( b y C l a i m βŠ‚ ξ€· 𝐴 3 ) ( 𝛼 ) + 𝐺 𝐢 ( 𝛾 ) + 𝐡 ( 𝛽 ) 𝐹 0 𝒱 ξ€Έ ξ€· 2 ξ€· 𝐴 ∩ 𝐷 0 + ξ€Έ ξ€Έ I m 𝐺 ( 𝛽 ) βŠ‚ 𝒱 2 , ( b y C l a i m 3 a n d C l a i m 5 ) , ( 2 . 3 1 ) which proves Claim 6. This completes the proof of Lemma 2.9.

Lemma 2.10. If a pair ( 𝒱 1 , 𝒱 2 ) of subspaces of 𝒳 is a generalized 𝑆 ( 𝐢 , 𝐴 , 𝐡 ) -pair such that π‘ž  𝑖 = 1 I m 𝐡 𝑖 βŠ‚ 𝒱 1 βŠ‚ 𝒱 2 βŠ‚ π‘Ÿ  𝑖 = 1 K e r 𝐢 𝑖 , 𝐴 𝑖 𝒱 2 βŠ‚ 𝒱 1 , ( 𝑖 = 1 , … , 𝑝 ) , 𝒱 2 βŠ‚ D ξ€· 𝐴 0 ξ€Έ , ( 2 . 3 2 ) then there exist a compensator ( 𝐾 , 𝐿 , 𝑀 , 𝑁 ) on 𝒲 ∢ = ( 𝒱 2 ∩ 𝒱 βŸ‚ 1 ) and a subspace 𝒱 𝑒 of 𝒳 𝑒 such that 𝒱 1 = 𝑆 1 , 𝒱 2 = 𝑆 2 , and 𝒱 𝑒 is generalized 𝑆 𝐴 𝑒 ( 𝑑 ) -invariant, where 𝑆 1 and 𝑆 2 are given in Lemma 2.7.

Proof. Suppose that there exists a pair ( 𝒱 1 , 𝒱 2 ) of subspaces satisfying the stated above conditions. Since 𝒱 1 and 𝒱 2 are closed subspaces and 𝒱 1 βŠ‚ 𝒱 2 , we have 𝒱 2 = 𝒱 1 βŠ• ( 𝒱 2 ∩ 𝒱 βŸ‚ 1 ) . Define 𝒲 ∢ = ( 𝒱 2 ∩ 𝒱 βŸ‚ 1 ) and 𝒳 𝑒 ∢ = 𝒳 βŠ• 𝒲 . Let 𝑅 ∢ 𝒱 2