Automatica 42 (2006) 1583 – 1588

www.elsevier.com/locate/automatica

Technical communique
Structurally constrained H2 and H∞ control: A rank-constrained
LMI approach夡
Seog-Joo Kim a,∗ , Young-Hyun Moon b
a I&C Group, Korea Electrotechnology Research Institute (KERI), Sung Joo Dong, Changwon, Kyungnam 641-120, Korea
b Department of Electrical and Electronic Engineering, Yonsei University, Shinchon Dong, Seoul 120-749, Korea

Received 26 September 2005; received in revised form 8 March 2006; accepted 19 March 2006
Available online 6 June 2006

Abstract
This paper describes an iterative linear matrix inequality (LMI) approach to the design of a structurally constrained output feedback controller.
The structured synthesis is formulated as a novel rank-constrained LMI optimization problem, where the controller parameters are explicitly
described so as to impose structural constraints on the parameter matrices. An iterative penalty-function method based on heuristics is used to
solve the rank-constrained LMI problem. A numerical experiment illustrates the practicality of the proposed method.
䉷 2006 Elsevier Ltd. All rights reserved.

Keywords: H∞ control; Linear matrix inequality (LMI); Penalty-function method; Structured synthesis

1. Introduction
Structurally constrained control involves designing a lin-
ear time-invariant (LTI) controller subject to structural con-
straints on its parameter matrices. This is sometimes called
fixed-structure control or structured synthesis. Structured con-
trol problems have long been recognized as practically impor-
tant because structural constraints are inevitable in many fields
of control and systems engineering, such as decentralized con-
trol. Static output feedback (SOF) stabilization also belongs to
this class of synthesis problems since it can be regarded as an
unstructured fixed-order control problem.
Despite its usefulness in practical control applications, struc-
tured controller synthesis is a challenging task due to its in-
herent nonconvexity, and no complete solution has been found
yet. Nonetheless, SOF controller design is a well-studied field
in the linear matrix inequality (LMI) framework. The coupled
LMI formulation employing the celebrated elimination lemma
夡 This paper was not presented at any IFAC meeting. This paper was
recommended for publication in revised form by Associate Editor Keqin Gu
under the direction of Editor André Tits.
∗ Corresponding author. Tel.: +82 55 280 1472; fax: +82 55 280 1476.
E-mail addresses: sjkim@keri.re.kr (S.-J. Kim), moon@yonsei.ac.kr
(Y.-H. Moon).
(e.g., Skelton, Iwasaki, & Grigoriadis, 1997) converts SOF syn-
thesis to the well-known rank-constrained LMI problem that
can be solved relatively efficiently using various iterative meth-
ods (Iwasaki & Skelton, 1995; Ghaoui, Oustry, & AitRami,
1997; Grigoriadis & Skelton, 1996). However, the elimination
lemma cannot be used for general structured control problems,
since these require an LMI formulation in which the controller
parameters can be directly manipulated. Several suitable LMI
formulations along with the associated numerical methods have
been proposed in the literature (de Oliveira, Camino, & Skel-
ton, 2000; Ebihara & Hagiwara, 2003; Han & Skelton, 2003;
Tan & Grigoriadis, 2000), and the present paper presents an-
other heuristic algorithm.
In this paper, motivated by Ebihara and Hagiwara (2003)
and Han and Skelton (2003), we first propose a unified LMI
representation for the design of structurally constrained con-
trollers. The synthesis is characterized as a rank-constrained
LMI optimization problem that differs from the coupled LMI
formulation. The problem is subsequently solved iteratively
using a penalty-function method (PFM) (Kim, Moon, Kwon,
& Kim, 2005b). Our method can find a suboptimal solution
without using a low-efficiency bisection method or a specific
initialization procedure. The solution of our algorithm moves
towards the region satisfying the rank condition as the penalty
parameter increases.

0005-1098/$ - see front matter 䉷 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.automatica.2006.03.017
1584 S.-J. Kim, Y.-H. Moon / Automatica 42 (2006) 1583 – 1588

We use the following notation. Sn denotes the set of all From  > 0 and rank(W̃ ) = m, the following hold:
symmetric matrices in Rn×n , and In denotes the n × n identity
matrix (the dimension n is omitted where it can be inferred I ≺ W̃3 ⇔ −1 I − W̃3−1  0,
from the context). For a matrix A, its transpose, trace, and rank
W̃1 = W̃2 W̃3−1 W̃2T .
are denoted by AT , tr(A), and rank(A), respectively. If A is
a symmetric matrix, A  0 (A
0) means that A is positive Also, it follows that
definite (semidefinite). For a rectangular matrix A, A⊥ stands
for an orthogonal complement of A; i.e., A⊥ A=0, A⊥ A⊥T  0. (W̃3 − I )−1 = W̃3−1 + W̃3−1 (−1 I − W̃3−1 )−1 W̃3−1 .
Finally, Tzw means the transfer function from w to z.
Applying the Schur complement to (3) yields
2. Preliminaries 0  M − W̃1 + W̃2 (W̃3 − I )−1 W̃2T
This section describes a basic tool for fixed-structure control = M + W̃2 W̃3−1 (−1 I − W̃3−1 )−1 W̃3−1 W̃2T
problems. In this framework, all LMI variables are explicitly
M.
expressed in the LMIs at the expense of an increase in the
This ends the proof. 
decision variables and a nonconvex rank condition. Lemma 1
establishes the basis of our method.
In the next section we show that the structurally constrained
controller synthesis can be described in the form of LMI (1)
Lemma 1. For matrices Q ∈ Sn and B ∈ Rn×m with
subject to the rank condition (2).
rank(B) < n, the following statements are equivalent:
3. Rank-constrained LMI formulation of a structured
(i) B ⊥ QB ⊥T ≺ 0.
synthesis
(ii) There exists a scalar
∈ R such that Q −
BB T ≺ 0.
(iii) There exist a positive semidefinite matrix W ∈ Sn+m and
Consider the LTI system with the following state-space rep-
a scalar  > 0 such that

 resentation:
Q B
≺ W, (1) ẋ = Ax + B1 w + B2 u,
B T Im
rank(W ) = m. (2) z = C1 x + D11 w + D12 u,

y = C2 x + D21 w, (5)
Proof. Statements (i) and (ii) are equivalent to each other due
to Finsler’s lemma (e.g., Skelton et al., 1997). By applying a where x ∈ Rn is the state, w ∈ Rnw is the exogenous input, u ∈
congruence transformation to (1), condition (iii) can be equiv- Rnu is the control input, z ∈ Rnz is the output to be regulated,
alently described as the existence of W̃
0 and  > 0 such that and y ∈ Rny is the measured output. Our aim is to design an


 SOF controller, u = Ky that places all the eigenvalues of the
Q B M 0
T T =
T
≺ W̃ , (3) closed-loop system in a stability region described by
B T I 0 I

rank(W̃ ) = m, (4) D(p, q, r) = {s ∈ C : p + qs + q ∗ s ∗ + r|s|2 < 0}, (6)

where where C denotes the set of complex numbers and ∗ represents

 the complex conjugate. For example, the regions D(0, 1, 0) and
I −−1 B D(−1, 0, 1) correspond to the open left-half complex plane for
T= , W̃ = TW TT ,
0 I continuous-time systems and the open unit circle for discrete-
M = Q − −1 BB T . time systems, respectively.
In addition, the closed-loop system should satisfy the fol-
First, it is immediate that condition (ii) implies the existence of lowing performance specification for all t > 0:
 > 0 and W̃
0 satisfying (3) and (4). If condition (ii) holds for  t
T


some
, there exists a  > 0 such that M ≺ 0. Consequently, w Q S w
d < 0. (7)
we can select W̃ in (3) as 0 z ST R z


0 0 For instance, the performance matrices for H∞ optimiza-
W̃ = ,  < .˜
0 I ˜ tion are given by Q = −2 I , S = 0, and R = I . We can
formulate this problem in terms of rank-constrained LMIs
Note that W̃
0 and rank(W̃ ) = m. To prove the converse, by Lemma 2.
suppose there exist  > 0 and W̃
0 such that (3) and (4) hold.
Partition W̃ as Lemma 2. For system (5), there exits a static output con-


W̃1 W̃2 trol law that ensures that the closed-loop system satisfies both
W̃ = .
W̃2T W̃3 the pole constraint (6) and the performance specification (7)
 S.-J. Kim, Y.-H. Moon / Automatica 42 (2006) 1583 – 1588 1585

if and only if the following rank-constrained LMI problem is Proof. See Iwasaki, Skelton, and Geromel (1994) and Chilali
feasible in P  0 (P ∈ Sn ), W
0 (W ∈ S3n+2nz +nw ), and a and Gahinet (1996). The proof is analogous to that for
scalar  > 0: Lemma 2. 
⎡ pP 0 qP 0 ATcl CclT ⎤ In the next section, we describe a computation method for
⎢ 0 Q 0 S BclT T ⎥
Dcl rank-constrained LMI problems.
⎢ ⎥
⎢ ∗ ⎥
⎢q P 0 rP 0 −In 0 ⎥
⎢ ⎥ ≺ W, (8)
⎢ 0 ST R −Inz ⎥
4. PFM for rank-constrained LMI problems
⎢ 0 0 ⎥
⎢ ⎥
⎣ Acl Bcl −In 0 In 0 ⎦ With a slight abuse of notation, we consider the following
Ccl Dcl 0 −Inz 0 Inz problem:

rank(W ) = n + nz , (9) min cT x

x

where s.t. W (x)
0, L(x)  0,

rank(W (x)) r, (14)

where x is the decision vector and W (x) ∈ Sn and L(x) ∈

Proof. According to Scherer, Gahinet, and Chilali (1997) and
Chilali and Gahinet (1996), satisfying the pole constraint (6)
and the performance condition (7) is equivalent to the existence
of a positive definite matrix P ∈ Sn such that
(10)
Applying Lemma 1 to (10) yields (8) and (9).  lemma provides the basis of the PFM.
Sm are matrices that are affine functions of x. The rank r is
assumed to be less than the full rank of W . It is worth noting
that the problems in the previous section can be represented
by (14) if we take the objective function as the H2 - or H∞ -
norm of the closed-loop system. For example, the H2 optimal
control problem in Lemma 3 can be stated as the minimization
of tr(B1T P B 1 ) subject to (12) and (13).
In the PFM, the rank-constrained problem (14) is first con-
verted to an LMI optimization problem without the rank con-
dition by incorporating a penalty function into the objective
function. Then, a sequence of convex LMI optimization prob-
lems are solved using an existing LMI solver.
To select a penalty function reflecting the violation of the rank
condition, we note that the rank condition in (14) is satisfied if
and only if the n − r eigenvalues of W are zero. The following

Lemma 3 states the structured SOF synthesis for H2 optimal Lemma 4. The rank of a positive semidefinite matrix W ∈ Sn
control. is less than or equal to r if and only if there exists a matrix
V ∈ Rn×(n−r) with V T V = In−r such that
Lemma 3. For system (5) with D11 = D21 = 0, we can find
an SOF control law such that the closed-loop system satisfies tr(V T W V ) = 0. (15)
the H2 performance, Tzw 2 = Ccl (sI − Acl )−1 Bcl 2 < , and
also the pole constraint (6) if and only if there exist P  0 Proof. Suppose that the eigenvalues of W are 1  · · · n .
(P ∈ Sn ), W
0 (W ∈ S3n+2nz ), and a scalar  > 0 such that Then, the following inequality holds (Horn & Johnson, 1985,
p. 191):
tr(B1T P B 1 )2 , (11)
⎡ ⎤ 1 + · · · + n−r tr(V T W V ) (16)
pP qP 0 ATcl CclT
⎢ q ∗P rP −In 0 ⎥ for any matrix V ∈ Rn×(n−r) such that V T V = In−r . The
⎢ 0 ⎥
⎢ ⎥ equality in (16) holds when the columns of V consist of the
⎢ 0 −Inz ⎥
⎢ 0 Inz 0 ⎥ ≺ W, (12) eigenvectors corresponding to the n − r smallest eigenvalues
⎢ ⎥
⎣ Acl −In 0 In 0 ⎦ of W . Therefore, the sufficiency is trivial from (16). To prove
Ccl 0 −Inz 0 Inz the necessity, suppose that a matrix W
0 with rank(W ) r is
given. Then, for V consisting of the eigenvectors corresponding
rank(W ) = n + nz , (13) to the n − r smallest eigenvalues of W , we have tr(V T W V ) =
1 + · · · + n−r = 0. This ends the proof. 
where
Also, notice that if rank(W ) > r, then tr(V T W V ) > 0.
Acl = A + B2 KC 2 , Ccl = C1 + D12 KC 2 . Therefore, tr(V T W V ) can be used as a penalty function for
1586 S.-J. Kim, Y.-H. Moon / Automatica 42 (2006) 1583 – 1588
the rank condition of problem (14). We now define the penalty
function as
p(x; V ) = tr(V T W V ), (17)
where V ∈ Rn×(n−r) , V T V = In−r . From (16), we see that the
proposed penalty function (17) is an upper bound on the sum
of the n − r smallest eigenvalues of W . Furthermore, note that
the penalty function is linear in x for a given V . As will be
seen below, we iteratively solve the following penalized convex
optimization to obtain a solution to (14) instead of minimizing
(17) and cT x
xk = arg min{(x; k ,k ,Vk−1 ) : x ∈ C}, k=1, 2, . . . , (18)
x
where
(x; , , V ) = cT x + tr(W ) + p(x; V ) (19)
and
C = {x : W (x)
0, L(x)  0}.
In (18), k > 0 is the penalty parameter and k > 0 is the opti-
mization weight. Also, Vk−1 is constructed from the eigenvec-
tors corresponding to the n−r smallest eigenvalues of W (xk−1 )
since the eigenvectors of W can be taken to be orthonormal to
each other.
To understand the necessity of the additional term tr(W ) in
(19), consider the following simple feasibility problem where
c = 0: find x1 , x2 satisfying the convex condition


x 0
W 1
0, x2 1 (20)
0 x2
and the rank condition rank(W ) = 1. The solution to this
problem is x1 = 0, x2 1. Assume that (18) is started with
x1 > 1, x2 =1 and that only penalty function (17) is minimized.
Since the smaller eigenvalue is 1, and the associated eigenvec-
tor is [0 1]T , the resulting penalty function becomes x2 . Then,
minimizing only (17) yields the problem:
min x2 s.t. (20), (21)
and its solution can be any value x1 0 and x2 = 1. Therefore,
we observe that x1 is not guaranteed to be zero by (21). This
implies that if the solution to (21) by existing optimization
software is x1 > 1, x2 =1, we cannot obtain the desired solution
despite successive iterations. However, by minimizing (19) as
min x1 + x2 + x2 s.t. (20),
we can obtain x1 = 0, x2 = 1 in a single iteration. Note that
tr(W ) = ni=1 i . This example demonstrates why we should
minimize the sum of the n − r eigenvalues relative to the total
sum of the eigenvalues.
Let us now examine certain properties of the algorithm.
Lemma 5. In (18), suppose that  and  are fixed, and that
only Vk is updated. Then, the following inequality holds:
(xk+1 ; , , Vk+1 )(xk ; , , Vk ). (22)
Proof. We know from (16) that p(xk ; Vk ) p(xk ; Vk−1 ) holds.
From the optimality condition, the following inequalities hold:
(xk+1 ; , , Vk+1 ) (xk+1 ; , , Vk )
(xk ; , , Vk ).
This ends the proof. 
Lemma 5 implies that for fixed  and , the sequence
{(xk ; , , Vk )} is nonincreasing. When the value of the
penalty function p(xk ; Vk ) is not sufficiently small with fixed
 and , we can make it smaller by increasing  based on the
following lemma.
Lemma 6. In (18), suppose that  and V are fixed. If
k+1 > k , the following inequalities hold:
(xk+1 ; , k+1 , V ) (xk ; , k , V ), (23)
p(xk+1 ; V ) p(xk ; V ), (24)
c xk+1 + tr(W (xk+1 )) c xk + tr(W (xk )).
T T
(25)
Proof. Let us denote
c̃T x = cT x + tr(W ).
The inequality (23) can be proved by the following inequalities
using the optimality condition and k+1 > k :
(xk ; , k , V ) = c̃T xk + k p(xk ; V )
 c̃T xk+1 + k p(xk+1 ; V )
 c̃T xk+1 + k+1 p(xk+1 ; V )
= (xk+1 ; , k+1 , V ).
In addition, from the inequalities
c̃T xk + k p(xk ; V )  c̃T xk+1 + k p(xk+1 ; V ) (26)
and
c̃T xk+1 + k+1 p(xk+1 ; V )  c̃T xk + k+1 p(xk ; V ),
the following holds:
(k+1 − k )p(xk+1 ; V ) (k+1 − k )p(xk ; V ),
which implies (24). Finally, from (24) and (26), we can see that
(25) holds. 
In a similar manner, the following lemma holds for fixed 
and V .
Lemma 7. In (18), suppose that  and V are fixed. If
k+1 > k , the following inequalities hold:
(xk+1 ; k+1 , , V ) (xk ; k , , V ),
cT xk+1 cT xk ,
tr(W (xk+1 )) + p(xk+1 ; V ) tr(W (xk )) + p(xk ; V ).
Proof. The proof is omitted because it is similar to that for
Lemma 6. 
 S.-J. Kim, Y.-H. Moon / Automatica 42 (2006) 1583 – 1588 1587

Remark 1. Lemmas 6 and 7 shed light on the roles of the com- Table 1
putation parameters  and . By increasing , we can reduce Comparison of the H2 norm for Example 1
the value of the penalty function, and by increasing , we can Method de Oliveira et al. Ebihara and Hagi- PFM
reduce the value of the original objective function. (2002) wara (2003)

Tzw 2 0.33 0.27331 0.27296

We can now proceed to the implementation of the PFM for
rank-constrained LMI optimization problems.
Penalty function
Algorithm 1. The PFM for rank-constrained LMI optimization
problems: 100

p (x; V)
(1) Initialization: Find an initial x0 by solving x0 ={x : x ∈ C}. 10-4
Compute V0 from W (x0 ) using eigenvalue decomposition,
10-8
and choose 1 > 1, 1 > 1,  ∈ (0, 1), >1,  > 1,
> 1, 1 >1, 2 >1, and 3 >1. Set k = 1. 10-12
(2) Convex optimization: Compute xk by solving the convex
LMI optimization problem (18). H2 norm
(3) Computation of V : Compute Vk from W (xk ) using eigen- 0.3
value decomposition.
0.25
(4) Feasibility test: If p(xk ; Vk ) < 1 or p(xk ; Vk )/tr(W (xk ))
< 3 , an almost feasible solution is obtained and the com- 2
0.2
putation proceeds to step 6.
(5) Penalty-parameter update: If p(xk ; Vk ) cannot be
0.15
reduced to the prescribed extent with k , that is,
p(xk ; Vk ) > p(xk−1 ; Vk−1 ), then increase the penalty pa- Penalty parameter and optimization weight
rameter by k+1 =k ; otherwise, k+1 =k . Set k ← k +1 106
and go to step 2. 
(6) Optimality test: If |cT xk −cT xk−1 |/|cT xk−1 | 2 , then cT xk
104
, 

cannot be further reduced; stop. 

(7) Optimization weight update: If |cT xk − cT xk−1 |/|cT xk−1 |
< , then increase the optimization weight by k+1 = k ; 102
otherwise, k+1 = k . Set k ← k + 1 and go to step 2.
20 40 60 80 100 120 140 160 180 200
Iteration
Remark 2. The PFM first tries to make p(xk ; Vk ) as small as
possible by adjusting k only. By Lemma 5, k is increased Fig. 1. Behavior of the PFM for Example 1.
only when the value of the penalty function is not sufficiently
reduced. If we find an almost feasible solution, we start increas-
Nevertheless, the PFM did find solutions for several problems
ing k to decrease cT xk . However, by Lemma 7, increasing k
(Kim, Moon, Kwon, & Kim, 2005a; Kim et al., 2005b).
might increase p(xk ; Vk ). If the value of the penalty function
becomes large while increasing k , we can reduce it again by
5. Numerical example
re-increasing k . This process is repeated until we cannot fur-
ther reduce cT xk while maintaining p(xk ; Vk ) small.
To illustrate the proposed PFM, we selected an example, and
the algorithm was implemented in MATLAB using SeDuMi
Remark 3. In the PFM, the most important computation
(Sturm, 1999) and YALMIP (Löfberg, 2004). More examples
parameters that affect the convergence of the algorithm are 1 ,
can be found in Kim et al. (2005a).
, and . Although their values require further investigation, us-
ing the adaptive approach in Fares, Apkarian, and Noll (2000)
Example 1. This example is the discrete-time H2 optimal con-
to adjust the penalty parameter allows the example in the next
trol problem discussed in Ebihara and Hagiwara (2003) and de
section to be solved without much difficulty. Moreover, unlike
Oliveira, Geromel, and Bernussou (2002), to which the reader
the algorithms in Ebihara and Hagiwara (2003) and Han and
is referred for the system data. A decentralized H2 suboptimal
Skelton (2003), ours does not require any specific initialization
controller was designed using Lemma 3 and the PFM. Our goal
procedures. We start the algorithm with a point x0 over the
is to minimize tr(B1T P B 1 ) subject to (12) and (13). The com-
convex set {x : x ∈ C}.
putation parameters used in the PFM were
Remark 4. Like other local algorithms (Ghaoui et al., 1997; 1 = 10−8 , 2 = 10−6 , 3 = 10−9 ,  = 1.1, = 1.03,
Grigoriadis & Skelton, 1996; Iwasaki & Skelton, 1995), the
proposed method does not guarantee global convergence. 1 = 1000, 1 = 100,  = 0.95, = 10−4 ,
1588 S.-J. Kim, Y.-H. Moon / Automatica 42 (2006) 1583 – 1588
and the obtained controller was


−0.4104 −0.3536 0 0
K= . conference on decision and control, Hawaii, USA (pp. 5632–5637).
0 0 −0.3492 −0.1648
Table 1 lists the design result of the PFM and those of previous
studies; comparison of the PFM data with those obtained pre-
viously demonstrates the good numerical performance of the
PFM. The behavior of the PFM for Example 1 is shown in
Fig. 1, which indicates that the value of the penalty function
decreases monotonically until the PFM converges to an almost
feasible solution in about 90 iterations. After the penalty func-
tion tends to almost zero, adjusting the optimization weight
results in a suboptimal solution.
6. Conclusion
We have described a practical procedure for finding struc-
turally constrained H2 and H∞ suboptimal output feedback
controllers for linear systems. The synthesis is formulated as a
new type of rank-constrained LMI optimization problem, which
was solved by an iterative PFM. Our method is applicable to
both continuous- and discrete-time systems. Although the pro-
posed algorithm does not exhibit global convergence, it per-
formed well in the numerical experiment presented here.
Acknowledgments
The authors would like to thank the reviewers for their helpful
and constructive comments.
References
Chilali, M., & Gahinet, P. (1996). H∞ design with pole placement constraints:
An LMI approach. IEEE Transactions on Automatic Control, 41(3),
358–367.
de Oliveira, M. C., Camino, J. F., & Skelton, R. E. (2000). A convexifying
algorithm for the design of structured linear controllers. Proceedings
of the 39th IEEE conference on decision and control, Sydney, Australia
(pp. 2781–2786).
de Oliveira, M. C., Geromel, J. C., & Bernussou, J. (2002). Extended H2 and
H∞ norm characterization and controller parameterizations for discrete-
time systems. International Journal of Control, 75(9), 666–679.
Ebihara, Y., & Hagiwara, T. (2003). Structured controller synthesis using
LMI and alternating projection method. Proceedings of the 42nd IEEE
Fares, B., Apkarian, P., & Noll, D. (2000). An augmented Lagrangian
method for a class of LMI-constrained problems in robust control
theory. Proceedings of the American control conference, Chicago, USA
(pp. 3702–3706).
Ghaoui, L. E., Oustry, F., & AitRami, M. (1997). A cone complementarity
linearization algorithm for static output feedback and related problems.
IEEE Transactions on Automatic Control, 42(8), 1171–1176.
Grigoriadis, K. M., & Skelton, R. E. (1996). Low-order control design for
LMI problems using alternating projection methods. Automatica, 32(8),
1117–1125.
Han, J., & Skelton, R. E., (2003). An LMI optimization approach for structured
linear controllers. Proceedings of the 42nd IEEE conference on decision
and control, Hawaii, USA (pp. 5143–5148).
Horn, R. A., & Johnson, C. R. (1985). Matrix analysis. Cambridge: Cambridge
University Press.
Iwasaki, T., & Skelton, R. E. (1995). The XY-centering algorithm for the dual
LMI problem: A new approach to fixed order control design. International
Journal of Control, 62(6), 1257–1272.
Iwasaki, T., Skelton, R. E., & Geromel, J. C. (1994). Linear quadratic
suboptimal control with static output feedback. Systems & Control Letters,
23(6), 421–430.
Kim, S. J., Moon, Y. H., Kwon, S., & Kim, K. H. (2005a). Design of
a structurally constrained suboptimal controller using an LMI method.
Proceedings of the 44th IEEE conference on decision and control and the
European control conference 2005, Seville, Spain (pp. 4318–4323).
Kim, S. J., Moon, Y. H., Kwon, S., & Kim, K. H. (2005b). Rank-constrained
LMI approach to mixed H2 /H∞ static output feedback controllers.
Proceedings of the 16th IFAC world congress, Prague, Czech.
Löfberg, J. (2004). YALMIP: A toolbox for modeling and optimization in
Matlab. Proceedings of the IEEE computer-aided control system design
conference, Taipei, Taiwan.
Scherer, C., Gahinet, P., & Chilali, M. (1997). Multiobjective output-feedback
control via LMI optimization. IEEE Transactions on Automatic Control,
42(7), 896–911.
Skelton, R. E., Iwasaki, T., & Grigoriadis, K. M. (1997). A unified approach
to linear control design. London: Taylor & Francis.
Sturm, J. F. (1999). Using SeDuMi 1.02, a Matlab toolbox for optimization
over symmetric cones. Optimization Methods and Software, 11–12(1–4),
625–653.
Tan, K., & Grigoriadis, K. M. (2000). Robust decentralized control using
an alternating projection approach. Proceedings of the American control
conference, Chicago, USA (pp. 801–805).