See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/236162695

Stabilization of Nonlinear Systems Subject to Actuator Saturation

Conference Paper  in  IEEE International Conference on Fuzzy Systems · July 2013

DOI: 10.1109/FUZZ-IEEE.2013.6622399

CITATIONS READS
9 207

4 authors:

Souad Bezzaoucha Rebaï Benoit Marx

Interdisciplinary Center for Security, Reliability and Trust (SnT), University of Luxem… University of Lorraine
50 PUBLICATIONS   209 CITATIONS    148 PUBLICATIONS   2,036 CITATIONS   

SEE PROFILE SEE PROFILE

Didier Maquin José Ragot

University of Lorraine University of Lorraine
385 PUBLICATIONS   3,891 CITATIONS    733 PUBLICATIONS   6,088 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Takagi-Sugeno model-based control View project

Control of nonlinear system View project

All content following this page was uploaded by José Ragot on 27 May 2014.

The user has requested enhancement of the downloaded file.

 Stabilization of Nonlinear Systems Subject to
Actuator Saturation
Souad Bezzaoucha∗† , Benoı̂t Marx∗† , Didier Maquin∗† , José Ragot∗†
∗ Université de Lorraine, CRAN, UMR 7039, 2 avenue de la Forêt de Haye, Vandoeuvre-lès-Nancy Cedex, 54516, France
† CNRS, CRAN, UMR 7039, France.
e-mail:firstname.name@univ-lorraine.fr.
Abstract—This paper addresses the stabilization of nonlinear
systems described by Takagi-Sugeno models affected by input
actuator saturation. A parallel distributed compensation design
is used for the state feedback controller. Stabilization conditions
in the sense of the Lyapunov method are derived and expressed as
a linear matrix inequality problem. The obtained gains depend
on the actuator saturation limits. A cart-pendulum example is
presented to illustrate the effectiveness of the proposed approach.
Keywords—Takagi-Sugeno systems, actuator saturation, parallel
distributed compensation control law
I. INTRODUCTION
Actuator saturation or control input saturation is probably
one of the most usual nonlinearity encountered in control
engineering due to the physical impossibility of applying
unlimited control signals and/or safety constraints.
In general, there are two main design strategies to deal with
actuator saturations. The first strategy is a two-step approach
in which a nominal linear controller is first constructed by
ignoring the actuator saturation. Then, after this controller has
been designed, usually using standard linear design tools, a
so called anti-windup compensator is designed to handle the
saturation constraints [1], [2], [3], [4]. A typical anti-windup
scheme consists in augmenting a nominal pre-designed linear
controller with a compensator based on the discrepancy be-
tween unsaturated and saturated control signals fed to the plant
[5]. The second strategy considers the saturation constraint
from the beginning of the design task. Several approaches were
developed, one of them is the invariant sets framework, which
has been significantly developed in control engineering over
the last decades, see [6], [7]. This framework ensures that
every state trajectory that starts inside the invariant set will
not exceed it, i.e. the system states will be bounded inside this
set.
An interesting approach for the determination of the invariant
sets framework was applied in [8] and [9]. This approach
consists in the so-called polytopic rewriting of the saturation
constraint. This polytopic representation is then used to deter-
minate (maximize) the largest invariant set in which the system
states remain bounded.
In the contrary of the previous method, that most likely ensures
to never reach the saturation level in order to preserve the
closed loop performances, in this paper, we explicitly consider
the saturation constraint, and even admit its occurrence. In fact,
the authors use the Takagi-Sugeno (T-S) representation of the
saturation (as well known as the polytopic representation) to
integrate the limitation constraints into the control synthesis,
such that the system stability is ensured and the control gains
are calculated depending on the saturation level.
The authors would like to precise that even if the expressions
of the T-S saturation has some similarity to the polytopic
one used in [8] and [9], the development, control strategy
and objectives are completely different since in the proposed
approach the invariant sets are never considered; moreover the
objective is the stabilization of nonlinear systems represented
with the Takagi-Sugeno (T-S) models and the synthesis of a
state feedback controller by parallel distributed compensation
(PDC) with control gains explicitly depending on the satu-
ration level. The controller is subject to actuator saturation
and the input saturation is directly taken into account in the
controller design. Stabilization conditions are derived from the
Lyapunov method and expressed as linear matrix inequalities
(LMI). A cart-pendulum example is considered to illustrate
the effectiveness of the proposed approach. The nonlinear
equations of the cart-pendulum are given and, using the Sector
Nonlinearity Transformation (SNT) [10], [11], a T-S model
of the system is deduced and used. The system is subject
to actuator saturations, it will be shown that these input
constraints may cause the instability of the nonlinear system,
but with the proposed approach the stability of the closed-loop
system is ensured.
The rest of this paper is organized as follows. Section 2
introduces the Takagi-Sugeno structure for modeling and some
preliminary results, mathematical notations and a brief descrip-
tion of the saturation. It is followed by the representation of the
nonlinear saturation by a T-S structure in section 3. In section
4 is designed a state feedback control law depending on the
saturation bounds. A numerical example and some simulation
results are given in section 5. Conclusions and future works
are exposed in section 6.
II. P RELIMINARIES
A. Takagi-Sugeno structure for modeling
Initially introduced in [12], the T-S models represent a
simple and accurate method to study the nonlinear behaviors.
The T-S modeling allows to represent the behavior of nonlinear
systems by the interpolation of a set of linear submodels [10],
[13], [14]. Each submodel contributes to the global behavior
of the nonlinear system through a weighting function µi (ξ(t)).
The T-S structure is given by
 where ujsat = sat(uj (t)), for j = 1, . . . , nu .
 n
X Considering the three parts of the saturated signal (4), each
component of the vector usat (t) is written as
 ẋ(t) = µi (ξ(t))(Ai x(t) + Bi u(t))



i=1 3
n (1) X
ujsat (t) = µji (uj (t)) (λji uj (t) + γij ), j = 1, . . . , nu (7)
 X
 y(t) = µi (ξ(t))(Ci x(t) + Di u(t))


i=1
i=1
with
where x(t) ∈ R is the system state variable, u(t) ∈ Rnu is
nx
λj1 = 0 λj2 = 1 λj3 = 0 (8)
the control input and y(t) ∈ Rm is the system output. ξ(t) ∈
Rq is the decision variable vector assumed to be measurable γ1j = ujmin γ2j = 0 γ3j = ujmax (9)
(as the system output) or known (as the system input). The
weighting functions µi (ξ(t)) of the n submodels satisfy the and the weighting functions defined by
convex sum property 
1−sign(uj (t)−ujmin )
 n 

 µ1 (uj (t)) = 2
 X
 sign(uj (t)−ujmin )−sign(uj (t)−ujmax ) (10)
µi (ξ(t)) = 1 µ2 (uj (t)) = 2
(2) 
i=1 1+sign(uj (t)−ujmax )


µ3 (uj (t)) =

 0 ≤ µ (ξ(t)) ≤ 1, i = 1, . . . , n
i 2

In the remaining of the paper, the following lemmas are used:
Lemma 1. Consider two matrices X and Y with appropriate
dimensions and G a symmetric positive definite matrix. The
following property is verified
X T Y + Y T X ≤ X T GX + Y T G−1 Y
Based on the convex sum property of the weighting functions
(2), the control input vector u(t) ∈ Rnu subject to actuator
saturation can be written in order to have the same activation
functions for all the input vector components as:
usat (t) =
(3) ! 
 
3
X nu X
Y 3

Lemma 2. (Congruence) Consider two matrices X and Y , if


 µ1i (λ1i u1 (t) + γi1 ) ×  µkj (uk (t)) 
 i=1 k=2 j=1
X is positive (resp. negative) definite and if Y is a full column

 
rank matrix, then the matrix Y XY T is positive (resp. negative)
 .. 
 . 
definite. 
 !  

 X 3 nu
Y X3 
µ`i (λ`i u` (t) + γi` ) ×  µkj uk (t)
 

B. Mathematical notations 
 i=1 k=1,k 6=` j=1


The following notations are used throughout the paper: a
 .. 
.
 
block diagonal matrix with the square matrices A1 , . . . , An
 
  

u −1 X
 3 ! nY 3
on its diagonal is denoted diag(A1 , . . . , An ). For any matrix, X n n

n k

M , S(M ) is defined by S(M ) = M + M T .The smallest and  µi (λi unu (t) + γi ) × 
u u u
µj uk (t)

largest eigenvalues of the matrix M are respectively denoted i=1 k=1 j=1
λmin (M ) and λmax (M ). The saturation function for a signal (11)
ν(t) is defined by (4), where νmax and νmin denote the For nu inputs, 3nu submodels are obtained. Thus, it is im-
saturation levels. portant to note that (11) is an analytical expression of the
actuators saturation directly expressed in term of the control
 ν(t) if νmin ≤ ν(t) ≤ νmax

variable, that can also be expressed as
sat(ν(t)) := νmax if ν(t) > νmax (4) nu
 3
νmin if ν(t) < νmin X
usat (t) = µi (t)(Λi u(t) + Γi ) (12)
i=1
III. P ROBLEM STATEMENT
The global weighting functions µi (t), the matrices Λi ∈
A. Takagi-Sugeno saturation control Rnu ×nu and vectors Γi ∈ Rnu ×1 are defined as follows
 nu
The main idea of this work is to model the nonlinear Y
µjσj (uj (t))

actuator saturation using the Takagi-Sugeno representation


 µ i (t) =
 i
(section II-A) and then propose a PDC control law ensuring  j=1
stability of the closed loop system. For that, it is proposed to Λi = diag(λjσj ) (13)
 i
re-write the saturation equation (4) for each component of the

 h iT
γσ1 1 , . . . , γσnnuu

control input vector under a particular form.  Γi

= i i
Let us consider a control input vector u(t) ∈ Rnu , defined by
T
where the indexes σij (i
= 1, . . . , 3 nu
and j = 1, . . . , nu ),
u(t) = ( u1 (t) . . . unu (t) ) (5) equal to 1, 2 or 3, indicate which partition of the j th input
(µj1 , µj2 or µj3 ) is involved in the i th submodel.
The control input under actuator saturation constraint is given The relations between i the number of the submodel) and the
by σij indices are given by the following equation

T
usat (t) = u1sat (t) . . . unsat
u
(t) (6)
 i = 3nu −1 σi1 +3nu −2 σi2 +. . .+30 σinu −(31 +32 +. . .+3nu −1 ) IV. S ATURATED STATE FEEDBACK CONTROL INPUT
The objective is to design a stabilizing time-varying state
The indices σij are such that ((σi1 − 1), . . . , (σinu − 1)) feedback controller ensuring the stability of the system, even
correspond to (i-1) in base 3. in the presence of control input saturation. The solution is
An illustrative example is given for two inputs (nu = 2), with obtained by representing the saturation as a T-S system and

T by solving an optimization problem under LMI constraints.
usat (t) = u1sat (t) u2sat (t) (14) In the nominal case, if no saturation is affecting the system
(1), the well-known following PDC (Parallel distributed com-
Since three partitions are defined for each input, the Takagi- pensation) can be applied
Sugeno model for usat (t) is then represented by 32 submodels n
X
9 u(t) = − µj (ξ(t))Kj x(t) (19)
X
usat (t) = µi (t)(Λi u(t) + Γi ) (15) j=1

i=1 Using the quadrative Lyapunov function

with the parameters µi , Λi and Γi given by the following table V (x(t)) = xT (t)P x(t), P = PT > 0 (20)
the stability of the closed-loop system is obtained by comput-
submodel i (σi1 , σi2 ) µi (t) Λi Γi
ing the gains Kj from [10]
T
µ11 µ21 diag(λ11 , λ21 ) γ11 γ12

1 (1, 1)
T
P1 ATi +Ai P1 −RjT BiT −Bi Rj < 0i = 1, . . . , n; j = 1, . . . , n
µ11 µ22 diag(λ11 , λ22 ) γ11 γ22

2 (1, 2) (21)
3 (1, 3) µ11 µ23 diag(λ11 , λ23 )

γ11 γ32
T with Kj = P1−1 Rj (where P1 = P −1 ).

4 (2, 1) µ12 µ21 diag(λ12 , λ21 )


γ21 γ12
T In the presence of input saturation, the gains computed
T from (21) do not ensure the closed-loop stability. The objective
µ12 µ22 diag(λ12 , λ22 ) γ21 γ22

5 (2, 2) is then to design a nonlinear state feedback controller (19) in
6 (2, 3) µ12 µ23 diag(λ12 , λ23 )

γ21 γ32
T order to guarantee the stability of the saturated system (18)
T such that the control gains Ki depend on the saturation limits.
µ13 µ21 diag(λ13 , λ21 ) γ31 γ12

7 (3, 1) By replacing the control law (19) in the T-S system (18), the
8 (3, 2) µ13 µ22 diag(λ13 , λ22 )

γ31 γ22
T obtained closed-loop system is the following
T nu
µ13 µ23 diag(λ13 , λ23 ) γ31 γ33

9 (3, 3) n X
X n X
3

TABLE I. W EIGHTING FUNCTIONS , MATRICES Λi AND Γi , FOR

ẋ(t) = µi (ξ(t))µj (ξ(t))µsat
k (t)
nu = 2 i=1 j=1 k=1
((Ai − Bi Λk Kj )x(t) + Bi Γk ) (22)
The determination of the controller gains is now detailed.
B. Problem statement
Theorem 1. There exists a time-varying state feedback con-
Let us now consider a T-S nonlinear system represented by troller (19) for a saturated input system (18) ensuring that the
the following state equation system state converges toward an origin-centred ball of radius
n bounded by β if there exists P1 = P1T > 0, R, Σk = ΣTk > 0
ẋ(t) =
X
µi (ξ(t))(Ai x(t) + Bi u(t)) (16) solutions of the following optimization problem
i=1
min β (23)
The control input u(t) is subject to actuator saturation, then P1 , R, Σk
the system (16) becomes s.t.  
Qijk I
n <0 (24)
I −βI
X
ẋ(t) = µi (ξ(t))(Ai x(t) + Bi usat (t)) (17)
i=1 with
S(Ai P1 − Bi Λk Rj )
 
I
From (12), equation (17) can be written as Qijk = (25)
I −Σk
nu
n X
X 3 and
ẋ(t) = µi (ξ(t))µsat
k (t)(Ai x(t) + Bi (Λk u(t) + Γk )) ΓTk BiT Σk Bi Γk < β (26)
i=1 k=1
(18) nu
for i = 1, . . . , n, j = 1, . . . , n, k = 1, . . . , 3 .
with n the number of sub-models, nu the dimension of The controller gains are given by
the input vectors, µi (ξ) and ξ respectively are the system
weighting functions and the measurable decision variables. Kj = P1−1 Rj (27)
µsat
k , Λk and Γk are respectively the saturation weighting
functions and the parameters defined by equations (13). Proof:
 Let us define the quadratic Lyapunov function (20). is bounded by β (33) and the LMIs (26). From (24), with a
According to (22) and (20), the time derivative of V (x(t)) is Schur complement, it obviously follows that
given by
n X
X n X
3 nu
(1/β) I < −Qijk , i = 1, . . . , 3nu (39)
V̇ (x(t)) = µi (ξ(t))µj (ξ(t))µsat
k (t)
i=1 j=1 k=1 (28)
  implying that all the eigenvalues of (−Qijk ) are larger that
S(xT (t)P Bi Γk + xT (t)P (Ai − Bi Λk Kj )x(t)) 1/β. As a consequence 1/β < ε holds, and finally the radius
is bounded by β.
Using Lemma 1, it follows that
Remark 1. It is important to highlight that the proposed
S(xT (t)P Bi Γk ) ≤ ΓTk BiT Σk Bi Γk + xT (t)P Σ−1
k P x(t) (29) approach ensures the stability of nonlinear systems that may be
The time derivative of the Lyapunov function (28) is bounded destabilized by saturated control with a control proportional
as follows to the saturation limits. If the submodels are initially unstable,
the proposed approach is not suitable since the LMI condition
nu
n X
X n X
3 cannot be fulfilled. Thus, to overcome the submodels stability
V̇ (x(t)) ≤ µi (ξ(t))µj (ξ(t))µsat
k (t) constraint, a solution would be to consider a dynamic output
i=1 j=1 k=1 feedback control instead of the state feedback one [16].
 
xT (t)(S(P (Ai −Bi Λk Kj )+P Σ−1 T T
k P )x(t)+Γk Bi Σk Bi Γk
(30)
Let us define V. N UMERICAL EXAMPLE
Qijk = S(P (Ai − Bi Λk Kj )) + P Σ−1
k P (31)
Let us consider a nonlinear cart-pendulum system
ε= min λmin (−Qijk ) (32) illustrated in figure 1.
i=1:n,j=1:n,k=1:3nu

δ= max ΓTk BiT Σk Bi Γk (33)

i=1:n,k=1:3nu

Since Σk > 0 and from equation (30), according to Lyapunov

stability theory [15], V̇ (t) < −ε k x k2 +δ. It follows that
V̇ (t) < 0 for
 Qijk < 0

and (34)
 2 δ
kxk > ε
which means that x(t) is uniformly bounded
q and converges to
δ
a small origin-centered ball of radius ε .
Applying Lemma 2, Qijk < 0 is equivalent to
S(P −1 ATi − P −1 KjT ΛTk BiT ) + Σ−1
k <0 (35)
Using the variable changes Fig. 1. Cart-pendulum system

P −1

P1 =
(36) The pendulum rotates in a vertical plan around an axis
Rj = Kj P1
located on a cart. The cart can move along a horizontal rail,
condition (35) is linearized as lying in the rotation plane. The state of the system is a vector
T
x = ( x1 x2 x3 x4 ) where x1 = z(t) is the cart
S(P1 ATi − RjT ΛTk BiT ) + Σ−1k <0 position, x2 = ż(t) the cart velocity, x3 = θ(t) is the angle
i = 1, . . . , n; j = 1, . . . , n; k = 1, . . . , 3nu (37) between the downward direction and the pendulum (x2 = 0
Applying Schur’s complement to (37), it becomes for the upright position of the pendulum), and x4 = θ̇(t) is the
pendulum angular velocity. A control force F (t) parallel to the
S(P1 ATi − RjT ΛTk BiT )
 
I rail is applied to the cart. The pendulum mass and cart mass
Qijk = <0 (38) are denoted m = 1Kg and M = 5Kg respectively. l = 0.1m
I −Σk
is the pendulum length and I the moment of inertia of the
As the weighting functions satisfy (2) and Σk > 0, if (38) pendulum with respect to its axis on the cart. The cart friction
is satisfied for i = 1, . . . , n, j = 1, . . . , n, k = 1, . . . , 3nu , is composed of the viscous friction proportional to the cart
and k x k2 > δε , then V̇ (x(t)) < 0, which implies q that x(t) velocity f ż(t) (f = 100N/m/s), a friction proportional to
converges to an origin centered ball of radius δ
. the cart position ks z(t) (ks = 0.001N/rad/s) and a friction
qε torque in the angular motion of the pendulum proportional to
δ
The objective is now to minimize the radius ε. Firstly δ the angular velocity k θ̇(t) (k = 0.003N/rad/s). The state
equations are as follows: with
zi1 = max zi (t), zi2 = min zi (t), i = 1, . . . , 4 (47)

 (m + M )z̈(t) + ks z(t) + f ż(t) − mlθ̈(t)cos(θ(t))

 (
 +mlθ̇2 (t)sin(θ(t)) = F (t) Fi1 (t) = zzii1
(t)−zi2


−zi2
(48)

zi1 −zi (t)
 Fi2 (t) = zi1 −zi2
−mlz̈(t)cos(θ(t)) + (ml2 + I)θ̈(t)


The two partitions Fi1 (t) and Fi2 (t) of each premise variable



+k θ̇(t) + mglsin(θ(t)) = 0

zi (t), i = 1, 2, 3, 4 will contribute to the reconstruction of each
(40) weighting function of the T-S model. The sub-model number
In order to ensure the nonlinear system stability, the system depends on the non-linearities, such that r = 24 = 16.
equations (40) is written under a T-S from by applying the The T-S model is then given by the following equations:
sector nonlinearity approach which allows to exactly represent
16
the nonlinear system without any loss of information. Since X
the transition from the nonlinear system to the T-S one is not ẋ(t) = µi (t)(Ai x(t) + Bi u(t)) (49)
the main subject of this paper and for place limitation, only i=1
the main steps are given. with  16
First, the quasi-LPV form (quasi-linear with parameter vary-  Y
ing) is derived


 µi (t) = Fj,σj (zj (t))
i
j=1 (50)
ẋ(t) = A(x(t))x(t) + B(x(t))u(t) (41) 

 Ai = A(z1,σi1 , z2,σi2 , z3,σi3 , z4,σi4 )

Bi = B(z1,σi1 , z2,σi2 , z3,σi3 , z4,σi4 )
with :

T σij may take the value 1 or 2 and represents the number of the
x(t) = z(t) ż(t) θ(t) θ̇(t) , u(t) = F (t) (42) j th partition.
In order to simplify the LMI implementation, it is assumed that
0 1 0 0
 
the control gains Kj are equal for all the submodels. Then, two
 −ks /(m + M ) a1 (t) a2 (t) a3 (t)  control laws are proposed:
A(x(t)) =  
 0 0 0 1  u(t) = −Kn x(t) and u(t) = −KT S x(t) (51)
a7 (t) a4 (t) a5 (t) a6 (t)
T for the nominal controller (without input saturation) and the
B(t) = ( 0 b1 (t) 0 b2 (t) )
(43) proposed T-S controller respectivelly.
and the parameters ai (t) and bi (t) are defined as follows: The nominal control gain calculated from (21) is equal to:
 mlf z1 (t)z22 (t)
Kn = ( 3.4068 −95.7038 −29.8001 3.8235 ) (52)
f
 a1 (t) = − m+M −
 m+M
 Considering the saturation levels equal umin = 0 and umax =
 a2 (t) = − mlzm+M 3 (t)z4 (t)



 − 3, The control gain KT S is then calculated by solving the LMI
m2 l2 z22 (t)z1 (t)z3 (t)z4 (t)


 − mlgz1 (t)z2 (t)z3 (t) given by theorem 1 and is equal to:
m+M



a (t) = kz1 (t) KT S = ( 0.0003 5.9835 0.6139 −0.0024 ) (53)

 3



a4 (t) = −f z1 (t)z2 (t) T
For an initial position x0 = ( 0 0 π/12 0 ) , figure 2
 a5 (t) = −mlz1 (t)z2 (t)z3 (t)z4 (t) − (m + M )gz1 (t)z3 (t)

 shows the system states for the following three cases: xn for

a6 (t) = − k(m+M ) the nominal case without saturation, xnsat is the system states



 ml z1 (t)


 a (t) = −k z (t)z (t) with saturated nominal control, it is clear from the depicted
 7 s 1 2

 1 ml
figure that the saturation has a destabilizing effect on the
b (t) = m+M + m+M z22 (t)z1 (t)

 1 system. And finally xsatT S denotes the system state obtained



b2 (t) = z1 (t)z2(t) with the proposed T-S approach.
(44) From the depicted figures, it is clear that with the proposed
with the following non-linearity : T-S approach the system stability is ensured as well as the
 1
convergence to an origin centered-ball. However, since we

 z1 (t) = −mlcos2 (x3 (t))+(l+I/ml)(m+M ) considered the same gain for all the 16 sub-models, the
 z2 (t) = cos(x3 (t))

obtained solution is slightly conservative, which may justify
(45) the oscillatory response for the third and fourth state.
 z3 (t) = sin(x 3 (t))
x3

 In order to relax these constraints, the nonlinear system may
z4 (t) = x24 (t)

be simplified by considering the small angles approximations:
The reader will note that the nonlinearities choice cos(θ) ≈ 1 and sin(θ) ≈ θ. Hence, instead of the four previous
(z1 (t), z2 (t), z1 (t) and z4 (t)) is not unique and that the non-linearities, only one is considered, namely: z4 (t) = x24 (t)
premise variables are bounded and given by the angle and and a T-S system with two sub-models is then defined with
speed limitation. Now, the convex polytopic transformation is the T-S controller is then given by
applied 2
X
u(t) = − µi (ξ(t))KiT S x(t) (54)
zi (t) = Fi1 (t)zi1 + Fi2 (t)zi2 , i = 1, . . . , 4 (46) i=1
 saturation, these saturations are expressed in terms of the con-
0.2
x
1n
0.4
x
2n
trol variables and of the saturation limits. As a consequence,
0.15
x
x
1nsat
0.3
x
x
2nsat these limits are taken into account when solving the LMI
1satTS 2satTS
optimization problem and when computing the gains of the
0.1 0.2
controller. Moreover, the proposed Takagi-Sugeno approach
0.05 0.1 allows to extend the use of linear tools, namely the LMI
0 0
formalism, to a nonlinear control problem.
As an illustrative example, the state feedback controller design
−0.05
0 2 4 6 8 10
−0.1
0 2 4 6 8 10 was applied to a cart-pendulum nonlinear T-S system. The ob-
tained results show the effectiveness of the proposed approach
for the studied example.
0.3 2
x3n x4n
0.2 x3nsat x4nsat

0.1
x3satTS
1
x4satTS R EFERENCES
0 0 [1] A. Syaichu-Rohman and R. Middleton, “Anti-windup schemes for
−0.1 discrete time systems: An LMI-based design,” in 5th Asian Control
−1
Conference, Australia, 20-23 July 2004.
−0.2

−0.3
−2 [2] L. Zaccarian and A. Teel, “A common framework for anti-windup,
bumpless transfer and reliable design,” Automatica, vol. 38, no. 10,
−0.4 −3
0 2 4 6 8 10 0 2 4 6 8 10 pp. 1735–1744, 2002.
[3] E. Mulder, M. Kothare, and M. Morari, “Multivariable antiwindup con-
troller synthesis using linear matrix inequalities,” Automatica, vol. 37,
no. 9, pp. 1407–1416, 2001.
Fig. 2. System states [4] G. Grimm, J. Hatfield, I. Postlethwaite, A. Teel, M. Turner, and L. Za-
ccarian, “Antiwindup for stable linear systems with input saturation:
an LMI-based synthesis,” IEEE Transactions on Automatic Control,
with vol. 48, no. 9, pp. 1509–1525, 2003.
[5] A. Syaichu-Rohman, “Optimisation based feedback control of input
K1T S = ( 0.0127 −15.0442 15.8757 0.7854 ) constrained linear systems,” Ph.D. dissertation, The University of New-
(55) castle Callaghan, Australia, 2005.
K2T S = ( 0.0079 −19.0341 8.7667 0.5326 )
[6] A. Alessandri, M. Baglietto, and G. Battistelli, “On estimation error
As seen on Figure 3, displaying the obtained state responses, bounds for receding-horizon filters using quadratic boundedness,” IEEE
Transactions on Automatic Control, vol. 49, no. 8, pp. 1350–1355, 2004.
[7] F. Blanchini and S. Miani, Set-theoretic methods in control. Systems
0.2 0.4
and Control: Foundations and Applications. Boston-Basel-Berlin:
x1n x2n
Birkhauser, 2008.
x1nsat x2nsat
0.15 0.3
x1satTS x2satTS [8] Y. Cao and Z. Lin, “Robust stability analysis and fuzzy-scheduling
control for nonlinear systems subject to actuator saturation,” IEEE Fuzzy
0.1 0.2
Sets and Systems, vol. 11, no. 1, pp. 57–67, 2003.
0.05 0.1 [9] S. Tarbouriech, G. Garcia, J. Gomes da Silva Jr., and I. Queinnec,
Stability and Stabilization of Linear Systems with Saturating Actuators.
0 0
London: Springer-Verlag, 2011.
−0.05
0 2 4 6 8 10
−0.1
0 2 4 6 8 10
[10] K. Tanaka and H. Wang, Fuzzy Control Systems Design and Analysis:
A Linear Matrix Inequality Approach. New York: John Wiley & Sons,
Inc., 2001.
0.3 2 [11] A. Nagy, G. Mourot, B. Marx, J. Ragot, and G. Schutz, “Systematic
x3n x4n
0.2 x3nsat x4nsat
multimodeling methodology applied to an activated sludge reactor
1
x3satTS x4satTS model,” Industrial & Engineering Chemestry Research, vol. 49, no. 6,
0.1
0 pp. 2790–2799, 2010.
0
−1 [12] T. Takagi and M. Sugeno, “Fuzzy identification of systems and its
−0.1 applications to modeling and control,” IEEE Transactions on Systems,
−2
−0.2 Man, and Cybernetics, vol. 15, no. 1, pp. 116–132, 1985.
−0.3 −3 [13] K. Tanaka, T. Hori, and H. Wang, “A multiple Lyapunov function
−0.4 −4 approach to stabilization of fuzzy control systems,” IEEE Transactions
0 2 4 6 8 10 0 2 4 6 8 10
on Fuzzy Systems, vol. 11, no. 4, pp. 582–589, 2003.
[14] K. Tanaka, H. Ohtake, and H. Wang, “A descriptor system approach
to fuzzy control system design via fuzzy Lyapunov functions,” IEEE
Transactions on Fuzzy Systems, vol. 15, no. 3, pp. 333–341, 2007.
Fig. 3. System states
[15] K. Zhang, B. Jiang, and P. Shi, “A new approach to observer-based
fault-tolerant controller design for Takagi-Sugeno fuzzy systems with
it is clear that the system states are stable with less oscillations state delay,” Circuits, systems, and signal processing, vol. 28, no. 5, pp.
and converge to an origin centered ball. 679–697, 2009.
[16] S. Bezzaoucha, B. Marx, D. Maquin, and J. Ragot, “Contribution to the
constrained output feedback control,” in American Control Conference,
VI. C ONCLUSIONS ACC, Washington, DC, USA, June 2013.
Considering the saturation nonlinearity, a nonlinear system
can be represented in an augmented T-S form. It is important
to note that, using the proposed representation of the actuator

View publication stats