International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

32

Design and Analysis of a Model Predictive

Unified Power Flow Controller (MPUPFC) for
Power System Stability Assessment
Md.Shoiib Shahriar , Md.Ashik Ahmed and Md.Shahid Ullah
Abstract This paper addresses model predictive controller
(MPC) as an effective solution for improving the oscillations in a
single machine infinite bus (SMIB) power system connected with
a FACTS device, unified power flow controller (UPFC). UPFC is
mainly used in the transmission systems which can control the
power flow by controlling the voltage magnitude, phase angle
and impedance. And as a controller, MPC, not only provides the
optimal control inputs, but also predicts the system model
outputs which enable it to reach the desired goal. So, model
predictive unified power flow controller (MPUPFC), a
combination of UPFC and MPC along with proper system model
parameters can provide a satisfactory performance in damping
out the system oscillations making the system stable. Simulation
is done inMatlab. Response is shown for 2 different states
controlling 4 different control signals of UPFC.

Index Term Power system stability, FACTS, system model,

UPFC, MPC.
I

INTRODUCTION

Power system oscillations are an inevitable phenomenon of

the system. Though it starts with normal small changes in
system loads, it can turn into a disaster following a large
disturbance. Faults and weak protective relaying operation can
cause the oscillation to collapse the system [1]. In order to
damp these power system oscillations, different devices and
control methods are used [2]. Power system stabilizer (PSS)
has been used for years for this purpose as an economical and
effective solution. But, excessive variation in voltage profile
as well as the lack of control over three phase large
disturbances at generator terminals causes the demand of

Md.Shoiib Shahriar
Departmnet of Electrrical and Electronics Engineering , Islamic universuty of
Technology (IUT)Board Bazar 1704 ,Gazipur
Shoibee05@gmail.com
Md.Ashik Ahmed
Departmnet of Electrrical and Electronics Engineering , Islamic universuty of
Technology (IUT)Board Bazar 1704 ,Gazipur
ashikhmd@yahoo.com
Md.Shahid Ullah
Departmnet of Electrrical and Electronics Engineering , Islamic universuty of
Technology (IUT)Board Bazar 1704 ,Gazipur
msullah@iut-dhaka.edu

finding out a better solution to this problem [3].

FACTS technology is the newest way of improving power
system operation controllability and power transfer limits
which has been added in the stream with the progress in the
field of power electronics devices. FACTS devices can cause a
substantial increase in power transfer limits during steady state
operation [3-4].Among the FACTS devices, UPFC is the one
having very attractive and effective features. It is capable of
providing simultaneous control of voltage magnitude and
active and reactive power flows, in adaptive fashion. It has the
ability to control the power flow in transmission line, improve
the transient stability, mitigate system oscillation and provide
voltage support [1, 5].
A number of researches had been done to find out the control
schemes for performing the oscillation-damping task of
UPFC. Huang et al. [6] attempted to design a conventional
fixed-parameter lead-lag controller for a UPFC installed in the
tie line of a two area system to damp the inter area mode of
oscillation. Wang and Swift [7] developed a unified FillipsHeffron model for a power system equipped with a SVC, a
TCSC and a TCPS. Chen et al. [8] used an output feedback
controller designed by simulated annealing (SA) for TCSC to
improve power system low frequency oscillations. Mok et al.
[9] considered the design of an adaptive fuzzy logic controller
for the samepurpose. Dash et al. [10] suggested the use of a
radial basis function NN for a UPFC to enhance system
damping performance.
Among the controllers which are used to control the FACTS
devices, model predictive controller (MPC) has got a wide
range of attractive and versatile features. Now a day, this
controller is widely used in the field of industry as it has the
ability to implement constraints in control process system. A
good overview of industrial linear MPC techniques can be
found in [11].
In this paper, model predictive Controller (MPC) has been
chosen to control UPFC. Attractive features of these two tools
jointly provide a very satisfactory solution to the system
stability. System model of SMIB system along with UPFC is
controlled here with MPC for four different control signals of
UPFC. Impact of the system for different combinations of
control signals is also investigated.

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS

International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

I.

SYSTEM DYNAMIC MODEL

The power system [12] that is studied in this paper consists of

a synchronous generator connected to transmission lines and
an infinite bus via two transformers.The UPFC consists of an
excitation transformer (ET), a boosting transformer (BT)and
twothree-phase GTO based voltage source converters (VSCs);
which are connected to each other with a common dc link
capacitor [1].
Fig. 1 shows a SMIB system equipped with an UPFC. The
four input control signals to the UPFC are mE, mB, E, and B.
Where, mEis the excitation amplitude modulation ratio,
mBis the boosting amplitude modulation ratio,
Eis the excitation phase angle, and
Bis the boosting phase angle.

33

angle and speed, respectively; E q,Efd, and v the generator

internal, field and terminal voltages, respectively; T dothe
open circuit field time constant; xdandx dthe d-axis reactance,
d-axis transient reactance respectively; KAand TAthe exciter
gain and time constant, respectively; Vrefthe reference voltage;
and uPSSthe PSS control signal.
By applying Parks transformation and neglecting the
resistance and transients of the ET and BT transformers, the
UPFC can be modeled by the following equations (5-7):

vEtd
v [
Etq

iEd

]
i +
0
Eq

mE cos E vdc

(5)
mE s in E vdc
2

mB cos B vdc

iBd
2
(6)
i +
m
s
in

v
Bq

B
B
dc

2
3m
3m
vdc E (cosEiEd sinEiEq ) B (cosBiBd sinBiBq )
4Cdc
4Cdc

vBtd
v [
Btq

Fig. 1. SMIB power system equipped with UPFC

In stability and control studies of power system oscillations,

the linearized model can be used. In this paper dynamic model
of the system for small signal stability improvement is used.
Nominal parameters used for system modeling are given in
Appendix-I.

(7)
Where, vEt , iE , vBt , andiBare the excitation voltage, excitation
current, boosting voltage, and boosting current, respectively;
Cdc and vdcare the DC link capacitance and voltage.The
relations of excitation and boosting transformers parameters
and line currents can be written as:
m Sin v x
x
m Sin v
x
E dc Bd dE (v Cos B
B dc )
i BB E ' E
Ed x
q
b
2x
x
2
d
d
d

The nonlinear model of the SMIB system of Fig 1 can be

expressed by the following differential equations (1-4) [12]:

base ( 1) (1)
1

[ Pm Pe D( 1)] (2)
2H
1
E q' ' [ E fd Eq' ( xd xd' )id ] (3)
Tdo
1
E fd K A ((Vref v U pss ) E fd )
(4)
TA
Where Pm and Peare the input and output power, respectively.

Pe vd id vqiq , vt

2
d

vq2 ,

vd xq iq xq iiq ilq , vq Eq ' xd ' id ,

id iEd iBd , iq iEq iBq
Where M and D the inertia constant and damping coefficient,
respectively; basethe synchronousspeed; and the rotor

0 ]

Eq

m Cos v x
x
m Sin v
E
E dc Bq qE
B dc )
(v Sin B
2 xq
xq b
2

m Sin v x
x
m Sin v
x
i E E ' E E dc dE dt (v Cos B B dc )
Bd x
q
b
2x
x
2
d
d
d
m Cos v x
x
m Cos v
E
E dc qE
qt
B dc )
i

(v Sin B
Bq
2 xq
xq b
2

Where, XE and XB are the ET and BT reactance, respectively.

By combining and linearizing the equations (1-7) state space
equations of system will be obtained which are present in state
space represented equation (8). In this process there are 28
constants denoted by k are being used.
= AX + BU . (12)
Where the state vector Xand control vector Uare
X = [E qEfdVdc]T
U = [UpssmEEmBB]T

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS

International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

Where
0

K1
M

K
A 4
T 'd 0

K A K5
TA

K7
And
0

B 0

KA
TA

D
M

K2
M
K3

T 'd 0

0
0

0
K pe
M
Kqe
T 'd 0

K A Kve
TA

Kce

III.

1
T 'd 0

K8

K A K6
TA

1
TA
0

0
K pde

M
Kqde

K pd

M
K qd

T 'd 0

K K
A vd
TA

K9

0
K pb

M
Kqb

T 'd 0

T 'd 0

K AKvde
TA

Kcde

34

K A Kvb
TA
Kcb

0
K pdb

M
Kqdb

T 'd 0

K K
A vdb
TA

Kcdb

Fig. 2. Principle of model predictive control

DAMPING CONTROL

The unique feature of MPC which has made it different from

other controllers is the ability to predict the future response of
the plant.At each control interval an MPC algorithm attempts
to optimize future plant behavior by computing a sequence of
future manipulated variable adjustments [13]. Figure 2 shows
the basic principle of MPC.
Prediction horizon (TP) is the time range which, future system
outputs are predicted in it. Control horizon (Tc) is the time
steps number that input control sequence calculations for the
prediction horizon are done [13].In this plant model, value of
prediction horizon is taken 10 and control horizon is 2.
MPC approach can be expressed considering the
following finite horizon cost function [15]
H1
rh

J (xt ,[u0(t),...,uH1(t)]) h(xt iT (u),ui (t)) g(xtHT (u))

i1

Based on measurements obtained at time t, the controller

predicts the future dynamic behavior of the system over a
prediction horizon Tpand determines the input over a control
horizon Tc such that a predetermined open-loop performance
can be achieved. Ifthere were no disturbances and no modelplant mismatch, and ifthe optimization problem could be
solved for infinite horizons, then one could apply the input
function found at time t=0 to the system for all times t0.
However, this is not possible in general. Due to disturbances
and model-plant mismatch, the true system behavior is
different from the predicted behavior. In order to incorporate
some feedback mechanism, the open-loop manipulated input
function obtained will be implemented only until the next
measurement becomes available. Using the new measurement
at time t+ , the whole procedure will be repeated to find a
new input function with the control and prediction horizons
moving forward [14].

system equilibrium. Thiscondition on g(.) ensures

that the control design attempts to reach the system
equilibrium. Fig 3 shows a complete block
representation of UPFC connected SMIB system
controlled with MPC.

wheret is the current time; H is the length of the

optimization horizon; T is the sample period. If i >
0,then xt iT (u ) denotes the controlled trajectory at
time t iT from

xt

under

piecewise

controls

u [u0 (t ),..., ui 1 (t )] U ; h is the running cost;

and g is the terminal cost. We assume that h is nonnegative function and g satisfies

Linearized System Model

x Ax Bu

x
yref

g ( x) x xeq

for all x, where xeq is somedesired equilibrium and

>0 is some positive constant.That is, g is an
upward function whose lowest point is at the

Fig 3: Block diagram of MPC with UPFC connected plant.

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS

International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

IV.

35

SIMULATION RESULTS

A disturbance of pulse type signal is given in the systems Pe.

Disturbance duration (period) is 0.1 sec, disturbance amplitude
(size) is 1 unit and disturbance occurring time is 1 sec.
Response of MPC on proposed model is observed for 2 states

and . Individual effects of 4 different control signalsmE,

E, mB and B on system states are observed first in figure (47)

Fig. 4. Responses for control signal mE for states (i) , (ii)

Fig. 5. Responses for control signal mB for states (i) , (ii)

Fig. 6.Responses for control signal E for states (i) , (ii)

Fig. 7. Responses for control signal B for states (i) , (ii)

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS

International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

It has been seen that E(Fig. 6) and B(Fig. 7) give the worse
response. They can bring stability only in state , but takes a
long time which is not desired. Control signal mE (Fig. 4) also
doesnt have a good response characteristics. It can stable state
at a time of around 5 seconds but cant stable state .

36

Among the four signals, mB gives the best output making both
the states stable at a reasonable time of below 6 seconds (Fig.
5). As, MPC is a MIMO supported controller, so responses
will be observed now for control signals two at a time. Five
different combinations of two control signals are observed in
this
paper
in
fig
(8-12).

Fig. 8. Responses for control signal mB and mE together for states (i) , (ii)

Fig. 9. Responses for control signal B and E together for states (i) , (ii)

Fig. 10. Responses for control signal mE and E together for states (i) , (ii)

Fig. 11. Responses for control signal mB and B together for states (i) , (ii)

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS

International Journal of Electrical & Computer Sciences IJECS-IJENS Vol:12 No:04

37

Fig. 12. Responses for control signal mB and E together for states (i) , (ii)

Observing the combined effect of two different control

signals, it has been found that the responses are significantly
improved for every cases than applying them individually.
Among the 5 combinations, combination of mEand E provides
the worst response (Fig. 10). It cant make any of the states
stable. B-Ecombination brings stability in every states but
takes a long period of time (10 secs.) to bring stability in
(Fig. 9). Combination of mB-B(Fig. 11), mE-mB(Fig. 8)and
mB-E(Fig. 12) show the best responses here making both the
states stable within a short time of 3.5 seconds.

3.

V.
CONCLUSIONS
In this paper, by linearizing and combining the equations of
SMIB power system and UPFC, a complete state space model
was presented to study the oscillations. Effect for model
predictive controller for stability analysis is observed then for
different states of the system model. Different combinations
and individual impacts of four different control signals of
UPFC are analyzed. It has been found that the best solution of
stability analysis for the proposed model is the use of MPC
with multiple control signals of UPFC at a time.

7.

APPENDIX
The parameters used in system model:
Generator:
M = 8 MJ/MVA, Td0 = 5.044 s,D = 0 ,Xq =
0.6 pu, Xd = 1.0 pu, Xd = 0.3 pu
Transformers: XT = 0.1 pu, XE = 0.1 pu, XB = 0.1 pu
Transmission line : XL = 0.1 pu
Operating condition: Pe = 0.8 pu,Q=.1670pu , Vb = 1 pu, Vt =
1 pu
DC link parameter: VDC = 2 pu, CDC = 1.2 pu

4.

5.

6.

8.

9.
10.
11.

12.

13.
14.
15.

16.

REFERENCE
1.

2.

Shayeghi H. - Jalilizadeh S. - Shayanfar H. - Safari A. : Simultaneous

coordinated designing of UPFC and PSS output feedback controllers
using PSO, Journal of ELECTRICAL ENGINEERING, VOL. 60, NO.
4, 2009, 177184
Anderson P. M. - Fouad A.A., Power System Control and Stability :
Ames, IA: Iowa State Univ.Press, 1977.

17.

Keri A. J. F. Lombard X. Edris A. A. : Unified power flow

controller: modeling and analysis, IEEE Transaction on Power Delivery
14 No. 2 (1999), 648-654.[4] Hingorani N. G. Gyugyi L. :
Understanding FACTS: concepts andtechnology of flexible AC
transmission systems, Wiley-IEEE Press, 1999.
Hingorani N. G. Gyugyi L. : Understanding FACTS: concepts and
technology of flexible AC transmission systems, Wiley-IEEE Press,
1999.
Tambey N. - Kothari M.L. : Unified power flow controller based
damping controllers for damping low frequency oscillations in a power
system, IEE Proc. on Generation, Transmission and Distribution, Vol.
150, No. 2 (2003), 129-140.
Huang Z. Nil Y, et al. : Control strategy and case study, IEEE trans
Power System 2000;15(2):817-24.
Wang H.F. Swift F.J. : An unified model for the analysis of FACTS
devices in damping power system oscillations part I: Single machine
infinite bus power systems, IEEE Trans Power Deliver 1997; 12:941-6.
Chen X.R Pahalawaththa. N.C Annakkage U. D - Cumble C. S :
Design of decentralized output feedback TCSC damping controllers by
using simulated annealing, IEE Proc. on Generation, Transmission and
Distribution 145 No. 5 (1998), 553-558.
T. K. Mok, Y. Ni, and F. F. Wu, Design of fuzzy damping controller of
UPFC through genetic
algorithm, IEEE Power Eng. Society Summer Meeting, vol. 3, pp.
18891894, July 1620, 2000.
Dash P. K. - Mishra S. Panda G : A radial basis function neural
network controller for UPFC, IEEE Trans. Power Systems, vol. 15, no.
4, pp. 12931299, November 2000.
Qin S.J. and Badgwell T.A. An overview of industrial model predictive
control technology, In F. Allgower and A. Zheng, editors, Fifth
International Conference on Chemical Process Control CPC V, pages
232256. American Institute of Chemical Engineers, 1996.
Al-Awami A. T. - Abdel-Magid Y. L. - Abido M. A. : Simultaneous
Stabilization of Power System
Using UPFC-Based Controllers, Electric Power Components and
Systems, 34:941959, 2006.
Ahmadzade B. Shahgholian G. - MogharrabTehrani F. Mahdavian
M. : Model Predictive control to improve power system oscillations of
SMIB with Fuzzy logic controller.
Allgower F, Badgwell T.A, Qin J.S, Rawlings J.B and Wright S. J.
Nonlinear predictive control and moving horizon estimation An
introductory overview, In P. M. Frank, editor, Advances in Control,
Highlightsof ECC99, pages 391449. Springer, 1999.
Ford J.J. - Ledwich G. - Dong Z.Y : Efficient and robust model
predictive control for first swing transient stability of power systems
using flexible AC transmission systems devices, IET Generation,
Transmission & Distribution Received on 26th September 2007.

128704-6565-IJECS-IJENS August 2012 IJENS

IJENS