Contribution of PV Generators with Energy Storage

to Grid Frequency and Voltage Regulation Via

Nonlinear Control Techniques
Hamed Taheri *, Ouassima Akhrif * and Aimé Francis Okou †
*
École de technologie Supérieure(ÉTS), Montreal, Quebec, Canada ([email protected], [email protected])
†
Royal Military College of Canada(RMC), Kingston, Ontario, Canada ([email protected])

Abstract—This paper proposes a nonlinear control strategy for a The most common practice is to use energy storage systems
hybrid PV-battery system insuring frequency and voltage (ESSs) as smoothing devices for a PV system’s output [3, 4].
support of the power system. The hybrid system includes a PV These works try to smooth the fluctuating PV power but do
panel and battery connected to three-phase DC-AC inverter via not consider power utility conditions such as load variation
DC-DC boost converter and bidirectional DC-DC boost
and the resulting frequency deviations and no frequency
converter. A synchronous generator represents the power grid.
The voltage regulators control DC-DC boost converter and DC- regulation is therefore attempted [5]. Some approaches
AC inverter while the frequency regulator controls the considering frequency control have been reported recently. A
bidirectional DC-DC boost converter. A conventional MPPT is state of the art survey of load-frequency control strategies can
used to adjust the reference for nonlinear PV voltage regulator. be found in [6]. Most of these results are for islanded power
The voltage regulator is designed based on multi-input multi- systems or micro-grids [5,7,8]. In these cases, the output
output exact feedback linearization technique. It consists of a variations of renewable plants are not given to the power
module that uses the terminal voltage deviation to generate q-axis utility and fuel cells as storage devices are used for frequency
voltage component. A module that maintains the DC-link voltage control. [14] introduces a maximum power point tracking
is also added to generate d–axis voltage component as well. The
method providing reserve power capability to PV systems so
proposed frequency regulator includes a module that changes the
reference signal of a battery current control module when the that they can participate in frequency regulation but again, it is
frequency deviation is significant. The battery current regulator in a micro-grid environment
is designed based on partial input-output feedback linearization In [9] a frequency control method based on simple
strategy. The proposed control system is evaluated in simulation. fuzzy logic is proposed for a PV–diesel hybrid system to
The results reveal that with the proposed control scheme, the PV- introduce the control by the PV while a PSO-based PI
battery generator reacts like a conventional synchronous Frequency Controller for Wind-Solar Diesel Hybrid Energy
generator when the grid frequency changes considerably. Generation/Energy Storage is proposed in [10]. Power
modulation of a photovoltaic (PV) generator for frequency
Keywords-grid connected PV; Nonlinear control; battery;
converter; frequency; voltage
regulation is proposed in [11]. The generator has a small
electric double-layer capacitor. The capacitor absorbs rapid
fluctuations of PV generation, and allows the generator to
change its output at a limited ramp rate. The generator output
I. INTRODUCTION
is modulated in proportion to the frequency deviation. [12]
Renewable energy sources, such as photovoltaic presents a minimal-order observer-based control method for
(PV), wind energy, or small-scale hydro, provide a realistic small isolated utility connected large PV system to reduce
alternative to generators for electricity generation in remote frequency deviations. A linear technique to implement
areas [1]. To date, it has not been necessary for small PV frequency and voltage regulation capability in solar
generators to provide frequency-regulation services to the photovoltaic power plant in the framework of the Smart Grid
isolated utility. In the future, with an increasing penetration of is discussed in [15]. All these works, are based on linear
PV generation, their impact upon the overall control of the control techniques and do not try to coordinate frequency
power system will become significant [2]. This will lead to regulation with an MPPT strategy. In [13] and [16], the
situation where the PV generators will be required to share authors implemented a nonlinear model-based voltage and
some of the duties, such as grid frequency and voltage control. frequency regulation scheme for PV generators [13] and wind
Several studies have been carried out to minimize energy-based system [16]. No energy-storage was considered
the harmful effects of connecting PV generators to the utility.

978-1-4799-0224-8/13/$31.00 ©2013 IEEE 42

  which is used for charging
the identical dynamics except for 
however in these 2 papers and no MPPT strategy was Similarly, bidirectional DC-DC boost converter has
implemented.

 1 1
This paper extends the results in [13,16] by mode as:
 = − +

considering a hybrid PV-battery generator connected to the (2)
 

 1 1  !
grid. More importantly and contrary to the results in the
=   −  !  + .
 
  
literature, it supports the grid frequency and voltage while
insuring maximum power extraction from the PV generator.
An other advantage of the proposed control system is that it Consider now the DC-AC inverter. The inverter
takes into account the nonlinear dynamics of the PV-battery terminal voltage, the inductor current and the voltage at the
generator. It can therefore significantly improve the transient transformer primary are related by the following relationship in

 −"# 1
behaviour of the power system. the d-q reference frame.
= + %. + 
 
$ &
$ 
The paper is organized as follows. The PV power
system model is firstly obtained in Section 2. Then the
 1
= −%. + 
proposed approach to design the frequency and voltage
 & 
$ &
regulators is introduced in Section 3. The proposed control
system is evaluated in Section 4 and the simulation results are (3)
presented. Conclusion remarks are given in Section 5. id and iq stand for inductance currents in the d-q
reference frame. vd and vq are the converter terminal voltage
components. The d-q reference frame components for the
II. MODELLING OF SYSTEM voltage at the transformer primary are Vs and 0. The active
The hybrid grid-connected PV-battery generator power balance between the AC side and the DC side provides

 3"#. 
system is shown in Figure 1. This system includes a PV panel the following equation,
 = −

 2.
in series with DC-DC boost converter, battery stack in series (4)

The variable  is the complement of duty ratio of the boost

with bidirectional DC-DC boost converter, three-phase DC-
AC inverter, filtering inductors and a step-up transformer. The
secondary of the transformer is connected to the rest of the converter switch.
grid via a transmission line. A synchronous generator is used It is also important to represent the algebraic relationship
to represent the power grid system. The model-based control between the generator terminal voltage and the grid for the
design approach that will be used requires the dynamic model design of the voltage regulator. The relationship has the
of the entire system. This section presents the model that will following expression in the d-q reference frame. It is assumed

"# = ") − ) . %. &

be used to find the controller equations. that the grid is in quasi-steady state,

where ") represents the d component for the grid voltage.

(5)

Finally, the nonlinear model of three-phase DC-AC inverter is

 3"#. 
obtained as:
 =  −

 2.
 −"# 1
= + %. & + 
  $ $
 1
= −%. + 
Figure. 1 Model of hybrid PV-battery-diesel power system

Switching converters in this system include DC-DC  & 

$ &
boost converter, bidirectional DC-DC boost converter and "# = ") − ) . %. & (6)
three-phase DC-AC inverter. For DC-DC boost converter,
Defining:
Ω = +. ("- − "-
 /0$ 3
)and 2 = 2 − 2 4 5
 /0$ 2
based on KCL and KVL laws and nonlinear relationships
between inductor (or capacitor) voltage and current, nonlinear
where 2 = 2 ( )2 then, the following multi-input multi-


 1 1
models are extracted easily as:
  = − +
     
output nonlinear model is obtained:
 1 1  Ω = ") − ). %. & − "-
/0$

(1)


  =   −   +  .  
3
2 = 7  . : −
89 ; <-.8=
 ,  ,  ,  , ,  and  are representing PV voltage  ! 2 !
 −"# 1
=[ + %. & ] + 
 
$ $
and current, inductor current, DC bus voltage, inductance and

= −%. +  &
 
input capacitance of boost converter and dutycycle of switch
 & 
respectively. (7)
@

43
 and & are inputs of the system and  32 and Ω are the J =  
J = − J2 + 
 

[ J2 =  , ]    ! !
J = 7 J −  ! : + _` . a
 ^
\ = J2
(9)
") − ) . %. JL − "-
outputs. The system nonlinear dynamics canbe presented as:
/0$  2
P 00 V
  
JK or JK = b(J) + c(J)a
U P 0 0 V 
(8)
J2K O 789  . : − ; <-.QR
I M=O ! U+OO  0 UU X Y \ = d(J)
U O@  U &
2 !
J;K O + %. J
S<-
O U N @ T
h̀ h̀
S Qi j 8klm
JLK with f(x) = g h n, g(x) = pqr and h(x) = x2 and u is the
L .

T 0
@
N −%. JL
h
Q S ^
9 h 9 _` s

3
input or duty cycle of the system. To apply input-output
J = (J , J2 , J; , JL )Z = (Ω, 2 ,  , & )
where
Z
\ = d(J) = J2
feedback linearization technique, the output considered is:

Differentiating y with respect to time, we obtain:

\K = 7 J −  ! : + .a
  ^_`
  
(10)
III. DESIGN OF CONTROLLER
The relative degree of this system is therefore r = 1 < 2.
In order for the hybrid PV-battery generator to react like a Therefore partial input-output feedback linearization method
conventional generator, a unified nonlinear controller is
transformation, we choose a new state variable u(x) according
should be employed. In order to complete the coordinate
designed. In case of any frequency deviations, the hybrid
system should support grid frequency by providing appropriate
to the following condition:
. c(J) = 0
vw(x)
amount of active power to the power grid. Also, under any
vQ
contingencies in the grid, the developed hybrid system should (11)
Choosing u(x) = J satisfies Eq. 11. The nonlinear coordinate
transformation Z = ϕ(x ) is therefore:
be able to maintain the terminal voltage close to its nominal

{ = | (J) = u(x) = J
value. A three-level controller is designed to this end. To

[ 1 1
control the bidirectional DC-DC boost converter, a frequency
{2 = |2 (J) = −  + 
regulator is proposed. Voltage regulators fulfilling PV, DC bus
and terminal voltages are designed to control DC-DC boost
converter and three-phase DC-AC inverter as well. and the new state feedback law is given by,

u=
S~i w(x)j€
A. Design of frequency regulator
~ ~ w(x)
This regulator is responsible for controlling the output (12)
power of grid connected PV inverter. The power should be
With,
L2ƒ u(x) = [ ~ U†„ +  J ] and L‡ Lƒ u(x) =
S S  Sˆ‰Š
equal to a given reference value in steady state. Since PV
„ ~.„
Applying the coordinate transformation Z = ϕ(x) and the new
always generates MPP power, the battery is used to provide
some of its power in reserve for regulation purposes. A sub-

0
module is used to readjust the reference value of the injecting
0 1
control law to the model (Eq. 9) gives
K
[Z = 70 0: ∙ Z + 1 . v
current of battery in order to support the grid when the

y = z
frequency deviation becomes significant. Figure 2 shows the (13)
structure of the frequency regulator. It includes a current

with v as the new auxiliary control variable.

control module and a frequency control module. Next the Eq. 13 represents a completely controllable linear system
design of these two modules is discussed.
∆™
E∗
A.2 Linear control design
PI + PI Battery The nonlinear state equations of bidirectional DC-DC boost

A, A
šA› A∗ , DA
DDD∗
current PWM converter, battery side have been converted into simple linear
- controller
state equations by partial feedback linearization strategy. Note
that the affine nonlinear system (2) and the linear system (13)
EFGH are related via a diffeomorphism. Therefore achieving battery
current regulation is equivalent to having the output of the
Figure. 2 Frequency regulator linear system (13) asymptotically track a reference output
function. To facilitate the study of asymptotic tracking, the

Firstly, we define the error e(t) which is the difference of the

A.1 Nonlinear current control module design system equations are converted into error state variable form.

actual output y(t) with the reference output y‘’ƒ (t). y‘’ƒ (t) is
This module is responsible for maintaining the charging-
discharching current of the battery equal to the reference value.
This value is modified by using a frequency sub-module which the reference value for the battery current. This reference may
is explained afterward. By controlling the battery current, the be modified by the frequency control module when the
output power is also controlled. Voltage and current dynamics frequency deviation becomes large.
derived in Eq. 2 are used to design the input-output feedback “() = \() − \”0$ () (14)
linearization-based controller. They can be written in the form:
{K = {2
In state space form, and given that,
∆
\• =  → ˜
{2K = 
(15)

44
Defining “ = { − \”0$ () and “2 = {2 − \K ”0$ () , it is easy to
find a relationship between “ and “2 as:
equations of modules, exact input-output feedback

“K = {K − \K ”0$ () = “2

linearization schemes are employed.

Ÿ 
/0$
“K2 = {K2 − \• ”0$ () =  − \• ”0$ ()
(16)
! Nonlinear DC-link


Voltage Controller
!

=−  “ − 2 “2 + \• ”0$ () 

The linear auxiliary control law can be defined as:
œGF
∗
(17) œž
"Ø
/0$
dq/abc PWM
Eqs 16 and 17 can be transformed into general state equation

“K 0 1 “ &
form as follow. Nonlinear Terminal

 =   "Ø
“K2 − − 2 “2
(18) Voltage Controller

The question now is to choose appropriate coefficients  Figure. 3 the voltage regulator
and 2 to ensure the system global stability. Let the desired
poles be #,2 = −¡%¢ £ ¤%¢ ¥1 − ¡ 2 , supposing known ¡ , the
V ) − L) . ω. xL − V°±’ƒ 0 0
Recall that the nonlinear dynamics of system are:
value %¢ is obtained as: xK P ® V
O 7²s U®³ . D: − ; µ°.xR U PO 0 0 VU v®
(20)
0.9257 x2K
%¢ = “J.ª;L ¡ I M= O „ 2 „ U+O  0 U Xv Y
©- x;K O U O~  U ¶
+ ω. x
Sµ°
xLK O ~ L U N0 ~ T
N −ω. xL T
where controller gains are found simply by solving the 
following equations:
3
x = (x , x2 , x; , xL )· = (Ω, U ®³2 , i® , i¶ )
where
·
∆ (#) = # + 2¡%¢ # + %¢
(19)
2 2

# −1 u = (u , u2 ) = (v® , v¶ )
· ·
and
∆(#) = « « = (# + 2 ). # + = #2 + y = (y , y2 )· = (U 3
# + 2 ®³2 , Ω)
 ·

2 # + 
i~
Let
= %¢ 2 θ(t) = U .D
⇒ ∆(#) = ∆ (#) ⇒ ˜  C ®³
2 = 2¡%¢
 which represent the duty cycles for the
The signals D and  y = x2 =z
Then the first output is,

\K = J2K =»() −

bidirectional boost converter switches are passed through a ; <-.QR
2 ! 2
={
K −
PWM module. The outputs of the PWM module are connected
; <K-.QR
\• =[»() + . a= ,
to the corresponding switches. ; <¼i ; <-.½.Q¾ ; <-
2 ! 2 !@ 2 ! 2 @.`
– ]- (21)
A.3 Frequency control module design The relative degree for this output is therefore ¿ = 2 and the nonlinear

K − ; µK°.xR + ; µÁ − ; µ°.Â.x¾ ]Ã
When grid frequency changes significantly due to control input is:
u = . Àv − [[θ(t)
perturbations in load power, some power should be provided S2 ~.Š i

; µ° 2 „ 2 „~ 2 „
either by synchronous generator or PV-battery generator in
\2 = J ={;
order to maintain the frequency at its nominal value. This For the second output,

\2K = JK =") − ) . %. JL − "- = {L

essential power for frequency compensation is provided by
/0$
power reserved in battery in DC side. The battery current
y2• = Ä−VK°±’ƒ + L) . ω2 . x; Å − Æ u2 = v2
reference should be changed by another frequency module to ~ .Â
~
(22)
The relative degree for this output is , r2 = 2 and the second control
increase some power for contribution to frequency regulation
by PV generator. The frequency control module is a two loop
controller; DC bus power controller and frequency controller.
Î2 − [−"-K
input is
a2 = + ) . % 2 . J; ]Ï
Frequency deviation along with PI controller determines S@ /0$
Æ .½
(23)
reference for DC bus power which is summation of PV and
The total relative degree is r = r + r2 = 4 = n .
battery powers. Likewise DC bus power finally assigns the
reference for battery current control module. PV power is
determined by MPPT and is therefore assumed to be always Therefore this system satisfies the exact feedback
constant. The output of frequency module is used to adjust the linearization condition. This means that the system can be
battery current reference value for current control module as linearized entirely by means the appropriate change of

coordinates z = (z , z2 , z; , zL )· and control inputs a and a2 the

described in previous subsection. coordinates and nonlinear state feedback. In the new

00
B. Design of smooth multi-input multi-output voltage
0 1 0 0
system can be rewritten as

ÓK = Ô0 0 0 0Õ ∙ Ó + Ö1 0× . 
regulator
Ò
(24)

0 0 0 1 00
In this section, a multi-input multi-output voltage regulator is
 0 0 0 0 01
designed. It has two duties; control of system terminal voltage
Ò

\ = ({ , {; )Z , Ó = ({ , {2 , {; , {L )Z ,  = ( , 2 )
and DC bus voltage. The structure of this controller is
presented in Figure 3 including terminal voltage module and

linear system. v is the new auxiliary two-dimensional control

DC bus voltage module. The detailed designed controllers for The system defined by Eq. 24 is a completely controllable
both modules are presented in the following. To obtain the

45
variable. The same aforementioned method is used to design figure 5. As seen during applied scenario, terminal voltage is
this linear controller. controlled at its nominal value, 1800V even after fault.
300
C. PV controller design

DC-link voltage
The PV controller includes two parts; PV voltage 200
controller and MPPT. It is represented by the single-input
single-output affine nonlinear system (1). It can easily be seen 100
that the relative degree of this system is r =2. Therefore it
satisfies the exact feedback linearization conditions, i.e., the 0
system can be linearized entirely by means of a change of 40 60 80 100 120 140

In this case, the coordinate transformation Z = ϕ(x ) is:

coordinates and nonlinear control. Time(Sec.)

z = φ (x) = y(x) = x
Figure. 4 Simulation result of DC bus voltage under load
perturbation
z2 = φ2 (x) = Lƒy(x) = − „Ú€ x2 + „Ú€ iڀ = zK
(13)
 
(25)
z2K = v
2100

Terminal voltage(V.)
and the new state feedback law is as follows, 1800

Si@ Û(Q)j
a= Ü @Û(Q)
(26) 1500

L2ƒ ω(x) = ıڀK − x + Udc and

Calculating:
  
1200
59 59.5 60 60.5 61 61.5 62 62.5 63 63.5 64

„Ú€ ~.„Ú€ ~.„Ú€

Time(Sec.)

L‡ Lƒω(x) =
Sˆ®³ Figure. 5 Simulation result of terminal voltage of transformer
~.„Ú€
(27)
the input is easily obtained as: Nominal grid frequency is 60Hz; but during very big
a= À − [ ßK − J + ]Ã
S.!    load perturbation the grid frequency is dropped while the
^ ! .! .!
(28)
conventional synchronous generator is not able to provide
the new coordinates Z = ϕ(x) and the new control law.
the linearized system can be rewritten as follows in terms of required power for returning frequency to its nominal value.

0 1 0
ZK = 7
PV system contributes into grid frequency as soon as grid is
[ : ∙ Z +  .v
0 0 1
(29) perturbed by load and sends required power to AC side. The
y = z simulation result shows that grid frequency is regulated to its
nominal value as shown in figure 6.
v is the new auxiliary control variable. The PV reference
The above equation is a completely controllable linear system.
70
voltage value is determined with MPPT. In this paper, a
Grid frequency(Hz.)

conventional MPPT technique based on perturbation and 60

observation is used. 50

40

30
IV. SIMULATION RESULTS
20
40 50 60 70 80 90 100 110 120 130 140 150
To evaluate the performance of the proposed control Time(Sec.)

strategy, a power system consisting of three-phase grid-tied Figure. 6 Simulation result of grid frequency regulation
PV generator supplying an 1800V, 60Hz power grid with
active and reactive power is deployed. The PV generator The simulation results indicate also that maximum
includes PV panel with 173W maximum power PV power, 173W is extracted from PV panel and verifies both
(Vmpp=69.2V, Impp=2.5A) and a lead-acid battery (100V, MPPT and controller responses very well.
10Ah). The diesel turbine is applied to the system at time
equal 2 second. The initial AC loads are set at 450 W. To 200
190
examine the frequency and voltage regulating aptitude of the 180
PV power(W)

170
proposed control scheme, the requisite active and reactive 160
150
power is altered by abrupt AC load increment of 350W and 140

100var at time t equal 60 second at AC bus. Before and after

130
120

load disturbance, the PV generator is expected to generate its 110

100
40 50 60 70 80 90 100 110 120 130 140 150
maximum power because of MPPT. Recall that the main Time(Sec.)

objective of project is to develop a PV system which reacts Figure. 7 Simulation results of PV power
just as a conventional generator, i.e., the PV generator should
contribute to regulations of both grid frequency and voltage When perturbation happens, diesel generator tries to
during any grid contingency. provide some powers for maintaining grid frequency at its
As seen in Figure 4, DC bus voltage is regulated to nominal value as indicated in figure 8. It increases its
its reference value. Terminal voltage is plotted as well in

46
 generating power, while its power is not big enough. As REFERENCES
expected, DC bus power increases its generated power at
perturbation time to provide the power for maintaining grid [1] Bakirtzis, A.G. and Gavanidou, E.S.,”Optimum operation of a small
frequency at 60Hz. The DC bus power stands for power sent autonomous system with unconventional energy sources”, Electric
from PV panel which is always MPP power plus battery Power System Research, 1992, 23, 933102
power. Battery power depicted in figure 8 shows the power [2] Y.T. Tan, D.S. Kirschen, and N. Jenkins, “A model of PV generation
suitable for stability analysis,” IEEE Trans. Energy Convers., vol. 19,
increment at perturbation time to provide required power for no. 4, pp. 748-755, Dec 2004
AC side. The PV power which is always MPP power and [3] T. Kinjo, T. Senjyu, N. Urasaki, and H. Fujita, “Output levelling of
battery power together contribute into grid frequency renewable energy by electric double layer capacitor applied for energy
regulation as hybrid PV-battery generator. The negative storage system,” IEEE Trans. Energy Convers., vol. 21, no.1, pp. 221-
battery power means that battery is discharging. It confirms 227, March 2006
that the PV system is able to contribute to grid frequency as [4] Md. H. Rahman and S. Yamashiro, “Novel distributed power generating
system of PV-ECaSS using solar energy estimation,” IEEE Trans.
well as voltage regulations. Energy Convers., vol. 22, no.2, pp. 358-367, June 2007.
[5] Datta, M.; Senjyu, T.; Yona, A.; Funabashi, T.; "Minimal-order
500 observer-based coordinated control method for isolated power utility
250
connected multiple photovoltaic systems to reduce frequency
Battery power(W.)

0
-250
deviations," Renewable Power Generation, IET , vol.4, no.2, pp.153-
-500 164, March 2010
-750
[6] H. Shayeghi, H.A. Shayanfar , A. Jalili, “Load frequency control
-1000
-1250
strategies: A state-of-the-art survey for the researcher”, Energy
Conversion and Management 50 (2009) 344–353
40 50 60 70 80 90 100 110 120 130 140 150
Time(Sec.) [7] D. Kottick, M. Blau and D. Edelstein, “Battery Energy Storage for
700
Frequency Regulation in an Island Power System”, IEEE Transactions
on Energy Conversion, Vol. 8, No. 3, September 1993
Synchronous generator

600
500 [8] X. Li, Y.-J. Song, and S.-B. Han, “Frequency control in micro-grid
power(W.)

400 power system combined with electrolyzer system and fuzzy PI

300 controller,” J. Power Sources, vol. 180, pp. 468–475, 2008.
200
100
[9] Datta, M.; Senjyu, T.; Yona, A.; Funabashi, T.; Chul-Hwan Kim; , "A
0
Frequency-Control Approach by Photovoltaic Generator in a PV–Diesel
40 50 60 70 80 90 100 110 120 130 140 150
Hybrid Power System," Energy Conversion, IEEE Transactions on ,
Time(Sec.)
vol.26, no.2, pp.559-571, June 2011
Figure. 8 Battery and synchronous generator powers [10] Das, D.C.; Roy, A.K.; Sinha, N.; , "PSO based frequency controller for
wind-solar-diesel hybrid energy generation/energy storage
system," Energy, Automation, and Signal (ICEAS), 2011 International
V. CONCLUSION Conference on , vol., no., pp.1-6, 28-30 Dec. 2011
[11] Kakimoto, N.; Takayama, S.; Satoh, H.; Nakamura, K., "Power
This paper proposes an advanced control scheme for PV Modulation of Photovoltaic Generator for Frequency Control of Power
generator in PV-battery-diesel hybrid power system System," Energy Conversion, IEEE Transactions on , vol.24, no.4,
supporting both grid frequency and voltage. The proposed pp.943-949, Dec. 2009
controller includes MPPT regulator which is used to control [12] Senjyu, T.; Datta, M.; Yona, A.; Chul-Hwan Kim; , "A Control Method
boost converter, voltage regulator to control three-phase for Small Utility Connected Large PV System to Reduce Frequency
Deviation Using a Minimal-Order Observer," Energy Conversion, IEEE
bidirectional inverter and frequency regulator to control a Transactions on , vol.24, no.2, pp.520-528, June 2009
bidirectional boost converter. An input-output feedback [13] Okou, A. F.; Akhrif, O.; Beguenane, R.; Tarbouchi, M.; , "Nonlinear
linearization method is employed to design controller control strategy insuring contribution of PV generator to voltage and
equations. The proposed control system is tested by frequency regulation," Power Electronics, Machines and Drives (PEMD
2012), 6th IET International Conference on , vol., no., pp.1-5, 27-29
simulation. A large load variation that provokes a significant March 2012
frequency deviation and a large voltage change is introduced [14] Pappu, V.A.K.; Chowdhury, B.H.; Bhatt, R.; , "Implementing frequency
to evaluate the new regulators. The simulation results reveal regulation capability in a solar photovoltaic power plant," North
that the hybrid PV-battery generator behaves exactly like American Power Symposium (NAPS), 2010 , vol., no., pp.1-6, 26-28
conventional synchronous generator. It increases the active Sept. 2010
power generator when the grid frequency decreases. It is also [15] Bhatt, R.; Chowdhury, B.; , "Grid frequency and voltage support using
PV systems with energy storage," North American Power Symposium
capable of maintaining its terminal voltage close to its (NAPS), 2011 , vol., no., pp.1-6, 4-6 Aug. 2011
reference value when a severe contingency occurs. [16] Sow, T.; Akhrif, O.; Okou, A.F.; Ba, A.O.; Gagnon, R.; , "Control
strategy insuring contribution of DFIG-Based wind turbines to primary
and secondary frequency regulation," IECON 2011 - 37th Annual
Conference on IEEE Industrial Electronics Society , pp.3123-3128, 7-10
Nov.2011

47

Powered by TCPDF (www.tcpdf.org)