Decentralized model predictive based load frequency control

in an interconnected power system

T.H. Mohamed
a,
, H. Bevrani
b
, A.A. Hassan
d
, T. Hiyama
c
a
High Institute of Energy, South Valley University, Egypt
b
Dept. of Electrical Engineering & Computer Science, University of Kurdistan, Iran
c
Dept. of Electrical Engineering & Computer Science, Kumamoto University, Kumamoto, Japan
d
Faculty of Engineering, Dept. of Electrical Engineering, Minia University, Minia, Egypt
a r t i c l e i n f o
Article history:
Received 11 January 2010
Received in revised form 1 September 2010
Accepted 19 September 2010
Available online xxxx
Keywords:
Load frequency control
Integral control
Model predictive control
Decentralized control
a b s t r a c t
This paper presents a new load frequency control (LFC) design using the model predictive control (MPC)
technique in a multi-area power system. The MPC technique has been designed such that the effect of the
uncertainty due to governor and turbine parameters variation and load disturbance is reduced. Each local
area controller is designed independently such that stability of the overall closed-loop system is guaran-
teed. A frequency response model of multi-area power system is introduced, and physical constraints of
the governors and turbines are considered. The model was employed in the MPC structures. Digital sim-
ulations for both two and three-area power systems are provided to validate the effectiveness of the pro-
posed scheme. The results show that, with the proposed MPC technique, the overall closed-loop system
performance demonstrated robustness in the face of uncertainties due to governors and turbines param-
eters variation and loads disturbances. A performance comparison between the proposed controller and a
classical integral control scheme is carried out conrming the superiority of the proposed MPC technique.
2010 Elsevier Ltd. All rights reserved.
1. Introduction
In LFC problem, area load change and abnormal conditions lead
to mismatches in frequency and scheduled power interchanges be-
tween areas. These mismatches have to be corrected by the LFC
system. LFC objectives, i.e. frequency regulation and tracking the
load demands, maintaining the tie-line power interchanges to
specied values in the presence of modeling uncertainties, system
nonlinearities and area load disturbances, determine the LFC syn-
thesis as a multi-objective optimization problem [1,2].
The xed parameters controller, like an integral controller or a
PI controller, is widely employed in the LFC application. Fixed
parameters controllers are designed at nominal operating points
and may no longer be suitable in all operating conditions. For this
reason, adaptive gain scheduling approaches have been proposed
for LFC synthesis [3,4].
This method could to overcome the disadvantages of the con-
ventional PID controller like. The need of adaptation of controller
parameters, but actually, it faces some difculties, like the instabil-
ity of transient response as a result of abruptness in the system
parameters, in additionally, Impossibility of obtaining accurate lin-
ear time invariant models at variable operating points [3].
Recently, the model predictive control (MPC) appears to be an
efcient strategy to control many applications in industry; it has
many advantages such as very fast response, robustness against
load disturbance and parameters uncertainty. Its straightforward
design procedure is considered as a major advantage of the MPC.
Given a model of the system, only an objective function incorporat-
ing the control objectives needs to be set up. Additional physical
constraints can be easily dealt with by adding them as inequality
constraints, whereas soft constraints can be accounted for in the
objective function using large penalties. Moreover, MPC adapts
well to different physical setups and allows for a unied approach
[5,6].
Recently, some papers have reported the application of MPC
technique on the load frequency control issue [79]. In [7], fast re-
sponse and robustness against parameter uncertainties and load
changes can be obtained using MPC controller, but, only for single
area load frequency control application. In [8] the usage of MPC in
multi-area power system is discussed, but, only by economic view-
point, it presented a new model predictive load frequency control
including economy logic for LFC cost reduction. In [9], Feasible
Cooperation-Based MPC (FC-MPC) method is used in distributed
LFC instead of Centralized MPC which is impractical for control
of large-scale, geographically expansive systems, such as power
systems, In spit of the good effort done in [9], the paper did not
0196-8904/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2010.09.016

Corresponding author.
E-mail addresses: [email protected] (T.H. Mohamed), [email protected]
(H. Bevrani), [email protected] (A.A. Hassan), [email protected]
(T. Hiyama).
Energy Conversion and Management xxx (2010) xxxxxx
Contents lists available at ScienceDirect
Energy Conversion and Management
j our nal homepage: www. el sevi er. com/ l ocat e/ enconman
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
deal with the problem of systems parameters mismatch, it only
discussed the effect of load change, in addition, the range of load
change used in the cases of study is very large and inappropriate
in load frequency control issue.
This paper sheds the light on the impacts of parametric uncer-
tainties beside the load change effect in an interconnected power
system with decentralized model predictive based load frequency
control. In this paper, each local area controller can be designed
independently. The MPC technique law produces its optimal out-
put derived from a quadratic cost function minimization based
on the dynamic model of the specied area. The technique calcu-
lates the optimal control signal while respecting the given con-
straints over the output frequency deviation and the load change.
The effects of the physical constraints such as generation rate con-
straint (GRC) and speed governor dead band [1] are considered.
The power system with the proposed MPC technique has been
tested through the effect of uncertainties due to governors and tur-
bines parameters variation and load disturbances using computer
simulation. A comparison has been made between the MPC and
the traditional integral controller, which is widely used in practical
industries, conrming the superiority of the proposed MPC tech-
nique. The simulation results proved that the proposed controller
guarantees the robust performance in the presence of uncertainties
due to governors and turbines parameters variation and loads
disturbances.
The rest of the paper is organized as follows: the description of
the dynamics of the interconnected power system is given in Sec-
tion 2. A general consideration about MPC and its cost function are
presented in Section 3. The proposed methodology is applied to
two and three-area power system as a cases study, in Section 4. Fi-
nally, the paper is concluded in Section 5.
2. System dynamics
A multi-area power system comprises areas that are intercon-
nected by tie-lines. The trend of frequency measured in each con-
trol area is an indicator of the trend of the mismatch power in the
interconnection and not in the control area alone. The LFC system
in each control area of an interconnected (multi-area) power sys-
tem should control the interchange power with the other control
areas as well as its local frequency. Therefore, the dynamic LFC sys-
tem model must take into account the tie-line power signal. For
this purpose, consider Fig. 1, which shows a power system with
N-control areas [1].
In this section, a frequency response model for any area-i of N
power system control areas with an aggregated generator unit in
each area is described [1].
The overall generator-load dynamic relationship between the
incremental mismatch power (DP
mi
DP
Li
) and the frequency
deviation (Df
i
) can be express
D
_
f
i

1
2H
i
_ _
DP
mi

1
2H
i
_ _
DP
Li

D
i
2H
i
_ _
Df
i

1
2H
i
_ _
DP
tie;i
1
the dynamic of the governor can be expressed as:
D
_
P
mi

1
T
ti
_ _
DP
gi

1
T
ti
_ _
DP
mi
2
the dynamic of the turbine can be expressed as:
D
_
P
gi

1
T
gi
_ _
DP
ci

1
R
i
T
gi
_ _
Df
i

1
T
gi
_ _
DP
gi
3
the total tie-line power change between area-i and the other areas
can be calculated as:
D
_
P
tie;i
2p

N
j1
ji
T
ij
Df
i

N
j1
ji
T
ij
Df
j
_

_
_

_
4
In a multi-area power system, in addition to regulating area fre-
quency, the supplementary control should maintain the net inter-
change power with neighbouring areas at scheduled values. This
is generally accomplished by adding a tie-line ow deviation to
the frequency deviation in the supplementary feedback loop. A suit-
able linear combination of frequency and tie-line power changes for
area i, is known as the area control error (ACE),
ACE
i
DP
tie;i
B
i
Df
i
5
Eqs. (1) to (4) represent the frequency response model for N power
system control areas with one generator unit in each area and can
be combined in the following state space model:
D
_
P
gi
D
_
P
mi
D
_
f
i
D
_
P
tie;i
_

_
_

1
T
gi
0
1
R
i
T
gi
0
1
T
ti

1
T
ti
0 0
0
1
2H
i

D
i
2H
i

1
2H
i
0 0 2p

N
j1
ji
T
ij
Df 0
_

_
_

_
DP
gi
DP
mi
Df
i
DP
tie;i
_

_
_

0 0
0 0

1
2H
i
0
0 2p
_

_
_

_
DP
Li
Dv
i
_ _

1
T
gi
0
0
_

_
_

_DP
ci
y ACE
i
0 0 B
i
1
DP
gi
DP
mi
Df
i
DP
tie;i
_

_
_

_
_

_
_

_
6
where DP
gi
is the governor output change of area i; DP
mi
the
mechanical power change of area i; Df
i
the frequency deviation of
area i; DP
Li
the load change of area i; DP
ci
the supplementary con-
trol action of area i; y
i
the system output of area i; H
i
the equivalent
inertia constant of area i; D
i
the equivalent damping coefcient of
area i; R
i
the speed droop characteristic of area i; T
gi
, T
ti
the governor
and turbine time constants of area i; ACE
i
the control error of area i;
B
i
the frequency bias factor of area i; T
ij
the tie-line synchronizing
coefcient with area j; DP
tie,i
the total tie-line power change be-
tween area i and the other areas; and DV
i
is the control area inter-
face, DV
i

N
j1
ji
T
ij
Df
j
.
3. Model predictive control
The MPC has proved to efciently control a wide range of appli-
cations in industry such as chemical process, petrol industry, elec-
tromechanical systems and many other applications. The MPC
scheme is based on an explicit use of a prediction model of the sys-
tem response to obtain the control actions by minimizing an objec-
tive function. Optimization objectives include minimization of the
difference between the predicted and reference response, and the
control effort subjected to prescribed constraints. The effectiveness
of the MPC is demonstrated to be equivalent to the optimal control.
It displays its main strength in its computational expediency, real-
time applications, intrinsic compensation for time delays, treat-
ment of constraints, and potential for future extensions of the
methodology. At each control interval, the rst input in the optimal
sequence is sent into the plant, and the entire calculation is re-
peated at subsequent control intervals. The purpose of taking
new measurements at each time step is to compensate for unmea-
sured disturbances and model inaccuracy, both of which cause the
2 T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
system output to be different from the one predicted by the model
[5,6].
Fig. 2 shows a simple structure of the MPC controller. An inter-
nal model is used to predict the future plant outputs based on the
past and current values of the inputs and outputs and on the pro-
posed optimal future control actions. the prediction has two main
components: the free response which being expected behavior of
the output assuming zero future control actions, and the forced re-
sponse which being the additional component of the output re-
sponse due to the candidate set of future controls. For a linear
system, the total prediction can be calculated by summing both
of free and forced responses; reference trajectory signal is the tar-
get values the output should attain. The optimizer is used to calcu-
late the best set of future control action by minimizing a cost
function (J), the optimization is subject to constraints on both
manipulated and controlled variables [10].
The general object is to tighten the future output error to zero,
with minimum input effort. The cost function to be minimized is
generally a weighted sum of square predicted errors and square fu-
ture control values, e.g. in the Generalized Predictive Control
(GPC):
JN
1
; N
2
; N
u

N
2
jN
1
bj^ yk jjk wk j
2

Nu
j1
kjuk j 1
2
7
where N
1
, N
2
are the lower and upper prediction horizons over the
output, N
u
is the control horizon, b(j), a(j) are weighting factors. The
control horizon permits to decrease the number of calculated future
control according to the relation: Du(k + j) = 0 for j PN
u
.
The w(k + j) represents the reference trajectory over the future
horizon N.
Constraints over the control signal, the outputs and the control
signal changing can be added to the cost function as follows:
u
min
uk u
max
u
min
Duk Du
max
y
min
yk y
max
8
Solution of Eq. (7) gives the optimal sequence of control signal over
the horizon N while respecting the given constraints of Eq. (8).
The MPC technique has many advantages, in particularly it can
pilot a big variety of process, being simple to apply in the case of
multivariable system, can compensate the effect of pure delay by
the prediction, inducing the anticipate effect in closed loop, being
a simple technique of control to be applied and also offer optimal
solution while respecting the given constraints. On the other hand,
this type of restructure required the knowledge of model for the
system, and in the present of constraints it becomes a relatively
more complex regulator than a simple conventional controller
Fig. 1. Dynamic model of a control area in an interconnected environment.
Model
Model
Past
Past
outputs
controls
Optimizer
Free responce
Reference
trajectory
Forced
response
Total
response
Future
controls
Cost function
J
Constraints
_
+
+ +
Future
errors
Fig. 2. A simple structure of the MPC controller.
T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx 3
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
such as a PID for example, and it takes more time for on-line
calculations.
4. Results and discussions
Computer simulations have been carried out in order to validate
the effectiveness of the proposed scheme. The Matlab/Simulink
software package has been used for this purpose.
The parameters of the decentralized MPC controllers are set as
follows:
prediction horizon = 13,
control horizon = 2,
weights on manipulated variables = 0,
weights on manipulated variable rates = 0.1,
weights on the output signals = 1,
sampling interval = 0.0002 s.
Constraints are imposed over the control action, and frequency
deviation are considered as follows:
max control action = 0.25 pu,
min control action = 0.25 pu,
max frequency deviation = 1 pu,
min frequency deviation = 1 pu.
For all studied areas in this paper, the simulation studies are
carried out for the proposed controllers with generation rate con-
+V
t1
-V
t1
1
ST
t1
1
2H
1
S + D
1
1
1+ST
g1
1
R
1
Turbine1
Governor1
Rotating
mass
and load
f
1
P
L1
P
g1
m1
P
2.T
12
/ S
B
1
ACE1
P
c1
tie
P
+V
t2
-V
t2
1
ST
t2
1
2H
2
S + D
2
1
1+ST
g2
Turbine2
Governor2
Rotating
mass
and load
f
2
P
L2
P
g2
m2
P
MPC
2
ACE2
P
c2
1
R
2
B
2
0
MPC
1
-1
0
Area 2
Area 1
Fig. 3. The block diagram of two area power system including the proposed MPC controllers.
Table 1
Parameters and data of a practical two area power system.
Area K (s) D (pu/HZ) 2H (pu s) R (Hz/pu) T
g
(s) T
t
(s) T
12
Area-1 0.3/s 0.015 0.1667 3.00 0.08 0.40 0.20
Area-2 0.2/s 0.016 0.2017 2.73 0.06 0.44
4 T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
straint (GRC) of 10% per minute and the maximum value of dead
band for governor is specied as 0.05 pu for each area [1].
4.1. Scenario A
In this scenario, two-control area power system, shown in Fig. 3
is considered as a test system to illustrate effectiveness of the pro-
posed control strategy. Each area consists of the overall rotating
mass and load and an aggregated generator unit including one
nonlinear turbine with GRC, and one governor with dead-band
constraint [1], as each local area controller can be designed inde-
pendently. On the other hand, the frequency deviation is used as
a feedback for the closed loop control system. The measured and
reference area control error ACE
i
, ACE
ref
= 0 Hz are fed to the model
-0.06
-0.04
-0.02
0

f
1

(
p
u
)

f
2

(
p
u
)
-0.04
-0.02
0
0 5 10 15 20 25 30
-0.01
0
0.01
0.02
(c)
(b)
(a)
Time, sec

P
t
i
e

(
p
u
)
Fig. 4. Power system responses to case1 of scenario A with MPC (solid) and conventional (dotted) controller: (a) frequency deviation in area-1, (b) frequency deviation in
area-2, and (c) tie-line power change.
(c)
(b)
(a)

f
1

(
p
u
)

f
2

(
p
u
)

P
t
i
e

(
p
u
)
-0.06
-0.04
-0.02
0
0.02
-0.04
-0.02
0
0.02
0 5 10 15 20 25 30 35 40 45 50
-0.01
0
0.01
0.02
Time, sec
Fig. 5. Power system responses to case2 of scenario A with MPC (solid) and conventional (dotted) controller: (a) frequency deviation in area-1, (b) frequency deviation in
area-2, and (c) tie-line power change.
T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx 5
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
predictive controller in order to obtain the supplementary control
action DP
ci
which add to the negative frequency feedback signal.
The resulting signal is fed the governor giving the governor
valve position which supplies the turbine to give the mechanical
power change DP
mi
which is affected by the load change DP
Li
and
the tie-line power change DP
tie,i
giving the input of the rotating
mass and load block to provide the actual frequency deviation
Df. In addition, the tie-line ow deviation is added to the frequency
deviation in the supplementary feedback loop to give the area con-
trol error ACE
i
.
A practical two areas power system having the following nom-
inal parameters [1] listed in Table 1.
The simulation studies are carried out for the proposed control-
lers with generation rate constraint (GRC) of 10% per minute and
the maximum value of dead band for governor is specied as
0.05 pu for each area [1].
4.1.1. A-1 Case 1
The system performance with the proposed MPC controllers at
nominal parameters is tested and compared with the system per-
Fig. 6. Three-control area power system.
Table 2
Parameters and data of a practical three-control area power system.
Area K (s) D (pu/HZ) 2H (pu s) R (Hz/pu) T
g
(s) T
t
(s) T
ij
Area-1 0.3/s 0.015 0.1667 3.00 0.08 0.40 T
12
= 0.20
T
13
= 0.25
Area-2 0.2/s 0.016 0.2017 2.73 0.06 0.44 T
21
= 0.20
T
23
= 0.15
Area-3 0.4/s 0.015 0.1247 2.82 0.07 0.3 T
31
= 0.25
T
32
= 0.12
0 5 10 15 20 25 30 35 40 45 50
-0.04
-0.02
0
0.02

f
1
(
H
z
)
0 5 10 15 20 25 30 35 40 45 50
-0.04
-0.02
0
0.02

f
2
(
H
z
)
0 5 10 15 20 25 30 35 40 45 50
-0.01
0
0.01
0.02

P
t
i
e
1
(
p
u
)
0 5 10 15 20 25 30 35 40 45 50
-0.01
0
0.01
Time, sec

P
t
i
e
2
(
p
u
)
(a)
(b)
(c)
(d)
Fig. 7. Power system responses to scenario B with MPC (solid) and conventional (dotted) controller: (a) frequency deviation in area-1, (b) frequency deviation in area-3, (c)
tie-line power change in area-1, and (d) tie-line power change in area-3.
6 T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
formance with a conventional integrator and at only load change in
area-2. Fig. 4 shows the simulation results in this case. The results
from the top to the bottom are: the frequency deviations of area-1
Df
1
, the frequency deviations of area-2 Df
2
, and the tie-line power
change between area-1 and area-2 DP
tie,1
using both the proposed
MPC and classical integrator controllers following a step load
change in area-2 (DP
L2
assumed to be 0.02 pu at t = 3 s). It is note-
worthy that with the proposed MPC controller the system is more
stable and fast comparing with the system with traditional
integrator.
4.1.2. A-2 Case 2
The robustness of the proposed MPC controller against wide
rang of parameter uncertainty is validated. In this case, the gover-
nor and turbine time constants of the two areas are increased to
T
g1
= 0.105 s (31% change), T
t1
= 0.785 s (95% change), T
g2
=
0.105 s (66% change) and T
t2
= 0.6 s (38% change), respectively.
Fig. 5 depicts the response of the MPC controllers in the presence of
above uncertainty, at same load change described in the rst case.
It has been indicated that a desirable performance response has
been achieved using the MPC controller while with conventional
integrator, the performance and stability is seriously degraded.
4.2. B Scenario B
To illustrate the behavior of LFC system with the proposed
decentralized MPC controller in a multi-area power system, con-
sider three identical interconnected control areas as shown in
Fig. 6. The simulation parameters [1] are given in Table 2. The sys-
tem is tested at a simultaneous 0.02-pu load step disturbance in
control area-2 and against wide rang of parameter uncertainty is
validated. In this case, the governor and turbine time constants
of each area is increased to T
g1
= 0.105 s (31% change), T
t1
=
0.785 s (95% change), T
g2
= 0.105 s (66% change) and T
t2
= 0.6 s
(38% change), T
g3
= 0.15 s (100% change) and T
t3
= 0.7 s
(100% change), respectively. Fig. 7 depicts the response of both
the proposed MPC and classical integrator controllers in the pres-
ence of above uncertainty. This gure shows that even at this se-
vere condition of uncertainties, system with proposed MPC
controllers keeps stable, while goes to oscillation with integrator
controllers.
5. Conclusion
This paper investigates robust load frequency control for inter-
connected power system based on the model predictive control
technique. The proposed method was applied to two and three-
control area power systems with parametric uncertainty and vari-
ous loads conditions. Digital simulations have been carried out in
order to validate the effectiveness of the proposed scheme. The
proposed controller has been tested for several mismatched
parameters and load disturbance.
A performance comparison between the proposed MPC and
conventional integrator controllers is carried out. The simulation
results demonstrate that the closed-loop system is robust against
the parameter perturbation of the system and has desirable perfor-
mance in comparison of classical integral control design in all of
the performed test scenarios.
References
[1] Bevrani H. Robust power system control. New York: Springer; 2009.
[2] Bevrani H, Hiyama T. On-line load-frequency regulation with time delays:
design and real-time implementation. IEEE Trans Energy Convers
2009;24(1):292300.
[3] Chang CS, Fu W. Area load frequency control using fuzzy gain scheduling of PI
controllers. Electr Power Syst Res 1997;42(1):14552.
[4] Talaq, Al-Basri F. Adaptive fuzzy gain scheduling for load frequency control.
IEEE Trans Power Syst 1999;14(1):14550.
[5] Thomas J, Dumur D, Buisson J, Gueguen H. Model predictive control for hybrid
systems under a state partition based MLD approach (SPMLD). International
conference on informatics in control, automation and robotics ICINCO04, vol.
3, Setbal; 2004. p. 7885.
[6] Richalet J, Rault A, Testud JL, Japon J. Model predictive heuristic control,
application to industrial processes. Automatica 1978;14(5):41328.
[7] Mohamed TH, Bevrani H, Hassan AA, Hiyama T. Model predictive based load
frequency control design. 16th international conference of electrical
engineering, Busan, Korea, July 2010.
[8] Rerkpreedapong D, Atic N, Feliachi A. Economy oriented model predictive load
frequency control. Power engineering, 2003 large engineering systems
conference on volume; 2003. p. 126.
[9] Venkat Aswin N, Hiskens Ian A, Rawlings James B, Wright Stephen J.
Distributed MPC strategies with application to power system automatic
generation control. IEEE Trans Control Syst Technol 2008;16(6). November.
[10] De Silva Clarence W. Mechatronic systems: devices, design, control, operation
and monitoring, book, Taylor and Francis Group, LLC; 2008.
T.H. Mohamed et al. / Energy Conversion and Management xxx (2010) xxxxxx 7
Please cite this article in press as: Mohamed TH et al. Decentralized model predictive based load frequency control in an interconnected power system.
Energy Convers Manage (2010), doi:10.1016/j.enconman.2010.09.016
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y
A
u
t
h
o
r

P
e
r
s
o
n
a
l

C
o
p
y