2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

An Advanced Excitation Control Scheme for a

Synchronous Generator in a Single Machine Infinite
Bus System Based on Detailed Dynamical Models
2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES) | 978-1-7281-5672-9/20/$31.00 ©2020 IEEE | DOI: 10.1109/PEDES49360.2020.9379338

Md Abu Hanif Pramanik Tushar Kanti Roy Subarto Kumar Ghosh Md Shamim Anower
Department of ECE Department of ETE Department of EEE Department of EEE
RUET RUET RUET RUET
Rajshahi, Bangladesh Rajshahi, Bangladesh Rajshahi, Bangladesh Rajshahi, Bangladesh
[email protected] [email protected] [email protected] [email protected]

Md Apel Mahmud
School of Engineering
Deakin University
Geelong, Australia
[email protected]

Abstract—In this paper, an advanced excitation control scheme excitation systems of synchronous generators. Excitation con-
based on the nonlinear partial feedback linearization technique trollers have the ability provide additional damping into power
is proposed to design an excitation controller for a synchronous systems during severe disturbances. The damping capability
generator in a single machine infinite bus (SMIB) system based
on the seventh-order dynamical model. The proposed partial of excitation controllers depends on dynamical properties of
feedback linearizing scheme is applied for simplifying and lin- synchronous generators and types of excitation controllers de-
earizing this detailed dynamical model of synchronous generators signed using these models. Therefore, it is essential to design
in an SMIB system using nonlinear coordinate transformations excitation controllers using advanced control techniques based
that cancel nonlinearities within the system. A linear state on detailed models of synchronous generators representing
feedback controller is then used to design for the partial feedback
linearized synchronous generators. The final excitation control their behaviors during the practical operation to ensure the
law include nonlinear terms and the linear state feedback stable operation.
control law where these nonlinear terms appear from nonlinear Traditionally, conventional power system stabilizers (PSSs)
coordinate transformations that cancel nonlinearities for making are utilized for damping oscillating behaviors of synchronous
the partial feedback linearized SMIB system independent of generators due to small changes in power systems [1]. How-
operating conditions. The stability of dynamics that do not
transform during the feedback linearization process is also ever, these PSSs do not work when power systems face
analyzed in this paper. Finally, simulation studies are carried severe disturbances. Some advanced linear control theories
out to evaluate the performance of the controller by applying the such as the zero dynamic linear control scheme [2], minimax
most severe three-phase short-circuit fault at the terminal of the linear quadratic Gaussian (LQG)-based scheme [3], etc. are
synchronous generator as well as at the middle of a transmission employed to design excitation controllers for further enhancing
line and results are compared with a traditional power system
stabilizer (PSS). the stability. However, all these controllers use linearized
Index Terms—Partial feedback linearizing controller, SMIB model of synchronous generators which restricting operating
system, excitation controller, synchronous generators, detailed points around the equilibrium point based on which the system
model. is linearized and excitation controllers are designed.
Different types of nonlinear control techniques such as adap-
I. I NTRODUCTION tive backstepping [4], robust adaptive backstepping [5], [6],
Power systems exhibit complexities and nonlinearities sliding mode [7], feedback linearizing model predictive [8],
which operate under stressed conditions due to steady growth exact feedback linearization [9], direct feedback lineariza-
in load demands. At the same times, external disturbances tion [10], and partial feedback linearization [11] schemes
such as short-circuit faults, changes in loads, etc. further are extensively used to design excitation controllers for syn-
stress power systems which lead to unstable operations. These chronous generators in either single machine or multimachine
unstable operating scenarios arise from oscillations with longer power systems. All these controllers ensure the operation
durations which need to be damped in order to maintain of power systems over a large operating scenarios along
the stable operations of power systems. Large synchronous some advantages and disadvantages. For example, adaptive
generators in traditional power systems are considered as controllers will be responding in a slow manner if the gains
the main sources of damping toques which are provided by are not appropriately selected and there are no systematic way

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

to calculate these gains. Similarly, sliding mode excitation Vt Infinite Bus

controllers introduce chattering effects while the model pre-
dictive and all feedback linearizing controllers require exact G
parameters of the system. Furthermore, all these excitation Transformer Transmission line
controllers are designed using simplified third-order models Vs
repented by the transient voltage behind a reactance which Fig. 1. A typical SMIB system
do not capture some useful dynamics such as d-axis transient
voltage [11].
Recently, adaptive [12] and partial feedback linearizing [13] sion lines, and an infinite bus. The dynamical model discussed
excitation controllers are designed including the dynamic of in this section is mainly based on this typical SMIB system.
d-axis transient voltage. The adaptive excitation controller The SMIB system is modeled by considering mechanical
in [12] requires measurement of all state variables as it ensures and electrical dynamics of the synchronous generator along
the desired tracking of these states. However, it is extremely with the dynamic of the excitation system. Here, electrical
difficult to have measurements for all state variables within the dynamics are represented based on both subtransient and
system, especially if the order of the system becomes higher. transient behaviors while the dynamic of the excitation system
The partial feedback linearization scheme in [13] transforms is captured by considering an IEEE Type-II excitation system
a firth-order system into a third-order linear one where the as discussed in [16].
state variables for the transformed or feedback linearized The following set of equations represents the mechanical
system are generally the speed deviation and its derivative. behaviors of the synchronous generator [16]:
Hence, the partial feedback linearization scheme is simple
and easy to implement. However, all these controllers are not δ̇ = ω − ω0
ω0 
designed using the subtransient characteristics of synchronous ω̇ = Pm − Kd Eq Iq − Kd ψ1d Iq − Kq Ed Id (1)
generators. 2H
 D
The detailed dynamical model of synchronous generators +Kq ψ2q Id − (Xq − Xd )Id Iq − (ω − ω0 )
considering subtransient characteristics is used in [14] for 2H
designing excitation controllers of synchronous generators where δ is the rotor angle; ω is the speed; ω0 is the syn-
in multimachine power systems where the rotor angle is chronous speed; D is the damping co-efficient; H is the inertia
considered as the output function. It is worth noting that the constant; Pm is the mechanical power input; Xd and Xq
feedback linearizability of the dynamic models of synchronous are d- and q-axis subtransient reactances, respectively; Xd
generators relies on the output as the transformed states of and Xq are d- and q-axis transient reactances, respectively;
the feedback linearized system are obtained based on some ψ1d and ψ2q are d- and q-axis subtransient flux linkages in
operations on this output and these states are then used as the damper coil, respectively; Ed and Eq are d- and q-axis
feedback variables for the controller [15]. The selection of the transient voltages of the synchronous generator, respectively;
rotor angle as the output function as presented in [14] requires and Id and Iq are d- and q-axis currents of the synchronous
additional observers as this rotor angle cannot be measured generator, respectively. The terms Kd , Kq , Kd , and Kq can
directly [9]. be represented as:
This work focuses to design an excitation controller using (Xq − Xls )
(Xd − Xls )
the partial feedback linearization technique for a synchronous Kd = , K 
= ,
(Xd − Xls ) q
(Xq − Xls )
generator in an SMIB system where the speed deviation is (2)
used as the output function rather than the rotor angle as used (Xd − Xd ) (Xq − Xq )
Kd = , and K 
=
in existing schemes. This speed deviation can be measured (Xd − Xls ) q
(Xq − Xls )
directly and does not require any observers. Furthermore, there
with Xls as the armature leakage reactance.
will be more damping into the system if the speed deviation
The electrical dynamics representing the subtransient and
is used as the feedback. The partial feedback linearization
transient behaviors can be described as [16]:
scheme is applied to cancel nonlinearities so that the SMIB
system becomes a linear model and the linear quadratic 1  
ψ̇1d = −  −ψ1d + Eq − (Xd − Xls )Id
regulator (LQR) technique is then applied to determine the Tdo
intermediate control input for the feedback linearized SMIB 1  
system. Finally, simulation results are presented to justify ψ̇2q = −  −ψ2q − Ed − (Xq − Xls )Iq
Tqo
the effectiveness of the proposed controller which are also 1 
compared with a conventional PSS. Ėd = −  −Ed + (Xq − Xq )Iq − Kq Kq (ψ2q (3)
Tqo

II. D ETAILED DYNAMICAL M ODEL OF A S YNCHRONOUS +(Xq − Xls )Iq + Ed )
G ENERATOR IN AN SMIB S YSTEM 1 
Ėq = −  −Eq − (Xd − Xd )Id − Kd Kd (ψ1d
Fig. 1 shows a typical SMIB system which comprises a Tdo

synchronous generator, a transformer, two parallel transmis- +(Xd − Xls )Id − Eq ) + Ef d

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

 
where Tdo and Tqo are d- and q-axis open-circuit subtransient relationship between states and parameters related to the
 
time constants, respectively; Tdo and Tqo are d- and q-axis excitation control signal; u = Vc is the excitation control
open-circuit transient time constants, respectively; Xd and Xq signal; y is the output function; and h(x) is a scalar function
are d- and q-axis reactances, respectively; and Ef d is the field representing the output. The vector, f (x) can be represented
excitation voltage. The terms Kd and Kq can be represented as:
as: f1 (x) = ω − ω0
(Xd − Xd ) (Xq − Xq ) ω0 
Kd = and Kq = (4) f2 (x) = Pm − Kd Eq Iq − Kd ψ1d Iq − Kq Ed Id

(Xd − Xls ) (Xq − Xls ) 2H
 D
The dynamic behavior of an IEEE Type-II exciter is rep- +Kq ψ2q Id − (Xq − Xd )Id Iq − (ω − ω0 )
2H
resented through the changes in the excitation coil voltage 1  
as [16]: f3 (x) = −  −ψ1d + Eq − (Xd − Xls )Id
Tdo
1 1  
Ėf d = − [−Ef d + KE (Vc + Vref − Vt )] (5) f4 (x) = −  −ψ2q − Ed − (Xq − Xls )Iq
TE Tqo
(9)
where TE is time constant of the exciter, KE is the gain of 1    
f5 (x) = −  −Ed + (Xq − Xq )Iq − Kq Kq (ψ2q
the exciter, Vc is excitation control signal, Vref is the reference Tqo
value of the terminal voltage, and Vt is the terminal voltage. 
+(Xq − Xls )Iq + Ed )
The dynamics of the synchronous generator represented by 1 
equations (1), (3), and (5) need to use some algebraic equations f6 (x) = −  −Eq − (Xd − Xd )Id − Kd Kd (ψ1d
Tdo
for Id , Iq , and Vt . By neglecting the resistances of the stator 
and transmission lines, Id and Iq can be calculated as follows: +(Xd − Xls )Id − Eq ) + Ef d
    1
1 f7 (x) = − [−Ef d + KE (Vc + Vref − Vt )]
Id =  Kd Eq + Kd ψ1d − Vsq ) TE
Xd + Xe
  (6) In this section, the controller is designed by considering the
1
Iq =  −Kq Ed + Kq ψ2q + Vsd ) speed deviation as the output function, i.e., y = h(x) = ω−ω0
Xq + Xe
and the feedback linearization scheme is used considering this
where Xe is the equivalent reactance of the line and Vsd = speed deviation as the output function as discussed through
Vs sin(δ − θvs ) and Vsq = Vs cos(δ − θvs ) are d- and q-axis different steps in the remaining of this section.
voltages of the infinite bus with Vs as the infinite bus voltage • Step 1: Calculation of the relative degree and feedback
and θvs as the angle of the voltage. The terminal voltage can linearizability
be written as: At the start of the design process, it is required to calculate

Vt = Vtd2 + Vtq2 (7) the relative degree in order to determine the order of the
transformed system, i.e., the feedback linearizability. For this
where Vtd = Vsd − Xe Iq and Vtq = Vsq + Xe Id are d- and purpose, it is essential to start with the following calculations:
q-axis terminal voltages.
L1−1
f h(x) = h(x) = ω − ω0 = Δω (10)
Equations (1), (3), and (5) represents the complete dynam-
ical model of the synchronous generator used in an SMIB and
system while algebraic equations (6) and (7) are used solving Lg L1−1 h(x) = Lg h(x) = 0 (11)
f
these dynamics. The proposed controller is designed using the
model as represented by equations (1), (3), and (5) whose where Lf h(x) and Lg h(x) represent the Lie derivative of
detailed discussions are provided in the following section. along the vector fields f (x) and g(x), respectively whose
definitions can be found in [11]. The relative degree can be
III. P ROPOSED E XCITATION C ONTROLLER D ESIGN represented by r under the following condition [17], [18]:
The dynamic properties of the SMIB system described
Lg Lr−1
f h(x) = 0 (12)
through equations (1), (3), and (5) can be expressed as:
Hence, it is essential to calculate the following:
ẋ = f (x) + g(x)u
(8) L2−1 h(x) = L1f h(x) = f2
y = h(x) f
(13)
Lg L3−1 h(x) = Lg L1f h(x) = 0
where x = [δ ω ψ1d ψ2q Ed Eq Ef d ]T f

represents the state of the system; f (x) = Since Lg L2−1

f h(x) = 0, it is essential to continue to the next
[f1 (x) f2 (x) f3 (x) f4 (x) f5 (x) f6 (x) f7 (x)]T represent a step of similar calculations for which it can be written as:
vector with the relationship among parameters and states of the ∂f2
system; g(x) = [g1 (x) g2 (x) g3 (x) g4 (x) g5 (x) g6 (x) g7 (x)]T L3−1
f h(x) = L2f h(x) = f (x)
∂x
with g1 (x) = g2 (x) = g3 (x) = g4 (x) = g5 (x) = g6 (x) = 0 ∂ 2 f2
(14)
and g7 (x) = − K E
TE , represents a vector depicting the
Lg L2−1 h(x) = L L
g f
1
h(x) = g(x) =
 0
f
∂x2

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

Hence, r = 3 which means that the seventh-order system will • Step 4: Excitation control law design
be transformed into a third-order system and the feedback At this stage, the original excitation control input, i.e., u =
linearized system will be a partially linearized one. Vc can be obtained from v = a(x) + b(x)u which can be
• Step 2: Partial feedback linearization and feedback lin- written as follows:
earized system v − a(x)
Since the SMIB system will be reduced to a system lower u= (21)
b(x)
than that of the original system, the feedback linearized states
Here, the linear control input is obtained by using a linear
can be written as:
quadratic regulator (LQR) controller for a similar condition as
z1 = L0f h(x) = h(x) discussed in [13] for which it can be calculated as:
z2 = L1f h(x) = f2 (15) v = −kz (22)
∂f2
z3 = L2f h(x) = f (x)
∂x where k represents linear gain with k = [1 2.29 2.14]. Here,
and partially linearized system can be written as: u is the partial feedback linearizing excitation control input
with nonlinear terms a(x) and b(x) for canceling nonlinearities
ż1 = z2 within the system and v is the linear control input stabilizing
ż2 = z3 (16) the system while providing the desired damping into the
ż3 = v = a(x) + b(x)u system. Simulation results are provided in the next section
based on this newly designed controller.
where v is a linear control input with a(x) = L3f h(x) and
b(x) = Lg L2f h(x). Here, a(x) can be calculated as: IV. S IMULATION R ESULTS
For evaluating the performance characteristics of this con-
∂ 2 f2
a(x) = L3f h(x) = Lf [L2f h(x)] = f (x) (17) troller, an SMIB system with the frequency of 60 Hz and
∂x2 230 kV transmission line with a single generator connected
The linear control input v for the feedback linearized system to an infinite bus through a network of transformers and
as represented by equation (16) and the original excitation transmission lines as shown in Fig. 1 is considered in this
control signal u can be obtained from v = a(x) + b(x)u if work. Here, the voltage at the primary side of the transformer
the remaining dynamics, that are not transformed through the is considered as 13.8 kV while the same in the secondary side
feedback linearization process, are stable. as 230 kV. The nominal parameters of the SMIB power system
• Step 3: Stability analysis of remaining dynamics used for the simulation in this paper are:
The stability of remaining four dynamics can be analyzed ω0 = 314.59 rad/s, Xd = 1.305 pu, Xd = 0.296 pu,
by selecting the nonlinear coordinate transformations (φ̃) in Xd = 0.252 pu, Xq = 0.474 pu, Xq = 0.3 pu, Xq =
 
such a way that these satisfy the following condition [11], 0.243 pu, Xls = 0.18 pu, Tdo = 1.01 pu, Tdo = 0.053 pu,
 
[17], [18]: Tqo = 0.4 s, Tqo = 0.1 s, TE = 0.0957s, KE = 1, H = 3.2 s,
Xe = 0.0014 pu, Vs = 1 pu, θsd = 0o , and D = 0.
Lg φ̃ = 0 (18) The performance characteristics of the designed partial feed-
The above condition will be satisfies if these coordinations are back linearizing excitation controller (PFBLEC) are analyzed
selected as follows: by applying a three-phase short-circuit fault at the point where
the synchronous generator is connected to the SMIB system
z̃1 = φ̃1 = δ and the middle of one transmission line out of two parallel
z̃1 = φ̃2 = ψ1d lines. Here, the fault is applied at t=5 s and clears at t=5.2 s
(19) i.e., the fault duration is 0.2 s for both scenarios. In order to
z̃1 = φ̃3 = ψ2q
show the superiority of the PFBLEC, the performance is also
z̃1 = φ̃4 = Ed compared with a conventional power system stabilizer (PSS).
for which corresponding dynamics can be represented as: With these fault different responses (terminal voltage, speed
deviation, rotor angle, and control signal) are observed.
z̃˙1 = Lf φ̃1 = f1 = z1
The terminal voltage of the synchronous generator will be
z̃˙2 = Lf φ̃2 = f4 zero due to application of a three-phase short-circuit fault on
(20)
z̃˙3 = Lf φ̃3 = f5 its terminal. At this stage the generator will be disconnected
from the system through the circuit breaker and it will be
z̃˙4 = Lf φ̃4 = f6
reconnected when the fault is cleared, i.e., after 0.2 s. Hence,
The stability of the first equation in (20) can be analyzed in a the terminal voltage becomes zero during the faulted condition,
similar way as discussed in [11], [13] and remaining equations i.e., from t=5.0 s to t=5.2 s as shown in Fig. 2(a). However,
can be expressed in terms of z̃1 , z̃2 , and z̃2 whose stability this voltage will exhibit oscillating behaviors which can also be
can be then analyzed using the same analogy. Finally, it can seen from Fig. 2(a). However, the severity of these oscillations
be found that all these dynamics are stable. are less in the terminal voltage while comparing with the speed

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

Terminal voltage (pu)

Speed deviation (pu)

1 0.02

PFBLEC 0
0.5 CPSS
PFBLEC
-0.02 CPSS
0
0 5 10 15 0 5 10 15
Time (s) Time (s)
(a) Terminal voltage (b) Speed deviation

150 20
PFBLEC PFBLEC
Rotor angle (degree)

Control input (pu)

CPSS 10 CPSS
100
20
19 0
50 18
8 9 10 11
-10
0
-20
0 5 10 15 0 5 10 15
Time (s) Time (s)
(c) Rotor angle (d) Control signal

Fig. 2. Responses for an SMIB system with a three-phase short-circuit fault at the terminal of the synchronous generator

1.2
Terminal voltage (pu)

Speed deviation (pu)

0.02 PFBLEC
1 CPSS

0.8 0
PFBLEC
10-3
2
0.6 CPSS
-0.02 0
0.4 -2
8 9 10 11 12
0.2 -0.04
0 5 10 15 0 5 Time (s) 10 15
Time (s)
(a) Terminal voltage (b) Speed deviation
20
60 PFBLEC
Rotor angle (degree)

PFBLEC
10 CPSS
Control signal

40 CPSS

20 0
20
0 -10
19.5
-20 19
10 11 12 13 -20
0 5 10 15 0 5 10 15
Time (s) Time(s)
(c) Rotor angle (d) Control signal

Fig. 3. Responses for an SMIB system with a three-phase short-circuit fault at the middle of one transmission line out of two parallel lines

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.
 2020 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES)

deviation and rotor angle behaviors as depicted in Figs. 2(b) synchronous generators and evaluating its performance over a
and (c), respectively. The main reason behind such severity range of real operating scenarios.
is that the stability issue in the synchronous generator is
R EFERENCES
mostly dominated by electromechanical properties. From these
figures, it can be seen that oscillations persist for a longer [1] Hongxia Wu, K. S. Tsakalis, and G. T. Heydt, “Evaluation of time delay
effects to wide-area power system stabilizer design,” IEEE Transactions
period with higher amplitudes as compared to the terminal on Power Systems, vol. 19, no. 4, pp. 1935–1941, 2004.
voltage response. However, the designed PFBLEC outperforms [2] M. Mahmud, “An alternative lqr-based excitation controller design for
the conventional PSS by damping oscillations in a faster way power systems to enhance small-signal stability,” International Journal
of Electrical Power & Energy Systems, vol. 63, pp. 1 – 7, 2014.
from where it can easily be evidenced that the designed [3] M. Hossain, H. Pota, V. Ugrinovskii, and R. Ramos, “Voltage mode sta-
controller has better damping capability. Furthermore, the bilisation in power systems with dynamic loads,” International Journal
designed controller requires to utilize less control efforts as of Electrical Power & Energy Systems, vol. 32, no. 9, pp. 911 – 920,
2010.
compared to the conventional PSS which can be found from [4] T. K. Roy, M. A. Mahmud, W. Shen, and A. M. T. Oo, “Nonlinear exci-
the corresponding control signal in Fig. 2(d). This figure tation control of synchronous generators based on adaptive backstepping
demonstrates that the conventional PSS hits the maximum method,” in 2015 IEEE 10th Conference on Industrial Electronics and
Applications (ICIEA), 2015, pp. 11–16.
physical limit (±10 pu) of the excitation system whereas the [5] T. K. Roy, M. A. Mahmud, W. Shen, and A. M. Oo, “Nonlinear
PFBLEC uses less than half of this limit. adaptive excitation controller design for multimachine power systems
Another fault is applied at the middle of one line as shown with unknown stability sensitive parameters,” IEEE Transactions on
Control Systems Technology, vol. 25, no. 6, pp. 2060–2072, 2017.
in Fig. 1. The severity of this fault is less than the short- [6] T. Roy, M. Mahmud, W. Shen, A. Oo, and M. Haque, “Robust nonlinear
circuit fault at the terminal of the synchronous generator. In adaptive backstepping excitation controller design for rejecting external
this case, the terminal voltage does not becomes zero though disturbances in multimachine power systems,” International Journal of
Electrical Power & Energy Systems, vol. 84, pp. 76 – 86, 2017.
it significantly reduces as the generator is still connected [7] X. Liu and Y. Han, “Decentralized multi-machine power system exci-
to the system. The terminal voltage, speed deviation, rotor tation control using continuous higher-order sliding mode technique,”
angle responses in Figs. 3(a) to (c) exhibit similar oscillatory International Journal of Electrical Power & Energy Systems, vol. 82,
pp. 76 – 86, 2016.
behaviors as discussed for the fault at the terminal of the [8] T. F. Orchi, T. K. Roy, M. A. Mahmud, and A. M. T. Oo, “Feedback lin-
generator. All these responses are quickly settled to their pre- earizing model predictive excitation controller design for multimachine
fault values during the post-fault condition when the PFBLEC power systems,” IEEE Access, vol. 6, pp. 2310–2319, 2018.
[9] M. Mahmud, H. Pota, and M. Hossain, “Full-order nonlinear observer-
is used. The superiority of the PFBLEC can further be proven based excitation controller design for interconnected power systems via
from Fig. 3(d) which shows the control signals for both exact linearization approach,” International Journal of Electrical Power
controllers and the PFBLEC does not use the full limit of & Energy Systems, vol. 41, no. 1, pp. 54 – 62, 2012.
[10] G. Kenne, R. Goma, H. Nkwawo, F. Lamnabhi-Lagarrigue, A. Arzande,
the exciter while the conventional PSS utilizes the maximum and J. C. Vannier, “An improved direct feedback linearization technique
limit. for transient stability enhancement and voltage regulation of power gen-
erators,” International Journal of Electrical Power & Energy Systems,
Simulation results with both faults clearly demonstrate vol. 32, no. 7, pp. 809 – 816, 2010.
the enhanced damping capability, faster response time, and [11] M. A. Mahmud, H. R. Pota, M. Aldeen, and M. J. Hossain, “Partial feed-
less stresses on the excitation system with the PFBLEC as back linearizing excitation controller for multimachine power systems
to improve transient stability,” IEEE Transactions on Power Systems,
compared to the conventional PSS. vol. 29, no. 2, pp. 561–571, 2014.
[12] T. K. Roy, M. A. Mahmud, and A. M. T. Oo, “Robust adaptive backstep-
V. C ONCLUSIONS ping excitation controller design for higher-order models of synchronous
generators in multimachine power systems,” IEEE Transactions on
Power Systems, vol. 34, no. 1, pp. 40–51, 2019.
A partial feedback linearizing excitation controller is de- [13] M. A. Mahmud, M. J. Hossain, H. R. Pota, and A. M. T. Oo, “Robust
signed for an SMIB system by considering the speed devia- partial feedback linearizing excitation controller design for multimachine
tion as the output function. The designed controller reduces power systems,” IEEE Transactions on Power Systems, vol. 32, no. 1,
pp. 3–16, 2017.
the original order of the synchronous generator to a third- [14] A. K. Singh and B. C. Pal, “Decentralized control of oscillatory
order system using nonlinear coordinate transformations. The dynamics in power systems using an extended lqr,” IEEE Transactions
dynamic stability of remaining dynamics are briefly discussed on Power Systems, vol. 31, no. 3, pp. 1715–1728, 2016.
[15] M. A. Mahmud, M. J. Hossain, and H. R. Pota, “Selection of output
and finally, a linear quadratic regulator-based control technique function in nonlinear feedback linearizing excitation control for power
is used to derive the linear control input for this third-order systems,” in 2011 Australian Control Conference, 2011, pp. 458–463.
feedback linearized system which uses the speed deviation [16] P. W. Sauer and M. A. Pai, Power System Dynamics and Stability. New
Jersey: Prentice Hall, 1997.
and its derivative terms as feedback variables. The enhanced [17] Q. Lu, Y. Sun, and S. Mei, Nonlinear Control Systems and Power System
damping capability of the designed partial feedback linearizing Dynamics. Norwell: Kluewe Academic Publishers, 2001.
controller can clearly be evidenced from the simulation re- [18] A. Isidori, Nonlinear Control Systems. London: Springer-Verlag, 2001.
sults where the oscillations are damped effectively comparing
with the PSS. Furthermore, the designed excitation controller
lengthens the lifetime of the exciter by reducing stresses as it
does not require to use the full physical limit of the excitation
system. Future works will include detailed stability analysis
along with dealing large-scale power systems with multiple

978-1-7281-5672-9/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY. Downloaded on June 28,2024 at 07:44:51 UTC from IEEE Xplore. Restrictions apply.