IMPLEMENTATION OF A STAND-ALONE PHOTOVOLTAIC SYSTEM BASED

ON DECENTRALIZED DC-DC CONVERTERS

Diogo B. Cndido, Jonatan R. R. Zientarski, Rafael C. Beltrame,
Jos R. Pinheiro, Leandro Michels and Hlio L. Hey
Federal University of Santa Maria UFSM
Power Electronics and Control Research Group GEPOC
Santa Maria - RS - Brazil
diogoeng.ufsm@gmail.com, heliohey@gmail.com

Abstract This paper proposes a stand-alone

photovoltaic (PV) system that has as its input stage a set
of DC-DC converters, which aim to perform the input
energy processing with maximum power point tracking
in a decentralized form. As battery charger the system
possesses a bidirectional converter for controlling the
power flux from/to battery bank, which is out of the main
power path. The output stage is performed by an
inverter. Previous studies have demonstrated that the
decentralized input stage can provide an efficiency gain
of 16% in situations where one or more photovoltaic
panel is shaded, damaged or aged. The paper analyzes all
operation modes of the proposed PV system, considering
different levels of load energy demand and of solar
irradiation. The proposed system is validated by
simulation and experimental results.
1

Keywords Decentralized Photovoltaic Systems, MPPT.

I. INTRODUCTION
The constantly increasing of global energy demand
together with worldwide concern about environmental
impacts generated by energy production, mainly those that
make use of fossil fuels, has heightened the interest in
environmentally-friendly energy sources. One technology
that has been receiving a great deal of attention is
photovoltaic (PV) technology, mostly due to its advantages
such as direct energy conversion without intermediate
thermal processes or engines, easy building-integration in
residential or commercial buildings, low environmental
impact and modularity. Its environmentally-friendly qualities
have contributed to the expansion of the installed capacity of
PV systems around the world, which had reached more than
9.2 GW at the beginning of 2008 [1].
The increasing global interest in PV systems has led to
competition among manufacturers, who have been aiming to
obtain more efficient and cheaper PV panels, and to make PV
technology more attractive and competitive when compared to
other renewable technologies. Concomitantly, much research
has been focused on static converters in an effort to obtain
higher efficiency and better maximum power point tracking
and, thus, maximize PV energy conversion. PV systems can be
classified as Grid-Connected and Stand-Alone PV systems.
Manuscript received 31/08/2009. Revised on 11/04/2010. Accepted for
publication in 15/05/2010 by recommendation of the Editor Fernando L. M.
Antunes.

220

Grid-connected systems are designed to operate in parallel and

interconnected with the electric utility grid. This allows the
power produced by the PV system to either supply on-site
electrical loads or to back-feed the grid when the PV system
output is greater than the on-site load demand. Stand-alone PV
systems are designed to operate independently of the electric
utility grid. Most stand-alone PV systems present an additional
energy storage system to supply the load when there is no
sunlight. This is an attractive alternative to feed isolated loads
without access to distribution power lines or where its
extension is not economically viable [1].
Typical stand-alone PV systems use a battery bank as an
energy storage system. The battery bank represents the most
onerous part in a stand-alone PV system when considering its
maintenance cost [2]-[4]. In cyclic applications, where the
battery is daily charged and discharged, the battery is the
most expensive element of the system throughout its useful
life. Although it costs up to 15% of the initial installation
cost, it represents up to 46% of the cost when maintenance
costs are considered [5]. This high cost is due to the batterys
short lifetime compared to the other system components. It is
important to highlight that relative cost of the battery bank
will be even more significant if a suitable battery charge
process is not used. Therefore, it is imperative to choose a
charge/discharge method that maximizes battery lifetime. In
this paper some battery charging methods are discussed
and the most suitable for the proposed stand-alone PV
system is indicated.
In relation to PV stand-alone system configurations, the
centralized topology is the most used. Due to the
employment of a single converter to perform the maximum
power point tracking (MPPT), the generation capability of
the PV array is not fully utilized in situations in which the
panels are shaded, damaged and/or aged, contributing to
decreased efficiency and increased cost of the energy
generated by such systems. As an alternative to this
configuration, this paper presents a stand-alone PV system
which uses a decentralized energy configuration in the input
stage. Additionally, based on the operation modes of the
systems, the energy management strategy to be used is
discussed and presented. The proposed configuration
provides a better utilization of the energy generated by the
PV array by means of individualized MPPT, as well as a
better management of the energy stored in the battery bank,
aiming at an increased life cycle and reduction of
maintenance costs.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

II. PHOTOVOLTAIC SYSTEMS

A. Centralized Photovoltaic Systems
A typical centralized PV system is shown in Figure 1. This
configuration has only one PV array, an input step-down
converter for maximum power point tracking (MPPT) and
battery charge control, a step-up converter for bus regulation
and an inverter to generate the AC output voltage. The
utilization of a series connection of several PV panels and a
single step-down converter to perform MPPT is an important
drawback for this topology, due to the low ratio of solar
energy conversion when one or more photovoltaic panel is
exposed to shading, or in cases of failure and defects [6]-[13].
In this scheme, the maximum extractable current of the array
is limited by the panel with lowest energy production. As a
result, if a single panel is shaded or suffers from aging, the
energy production is strongly reduced [11], [14].
As can be seen in Figure 1 (a), the employment of single
converters for energy conversion stages in centralized
stand-alone PV systems reduces the reliability of such
systems when compared to the systems which utilize a
decentralized configuration. In a centralized system, the
energy supplied to the load is completely shut down when a
failure in any power converter occurs or even during
replacement of panels or batteries.
B. Decentralized Photovoltaic Systems
The decentralized configurations for photovoltaic systems
were developed for utilization in the 1000 Roofs Program,
created by the German Federal Government, whose aim was to
promote the use and research of PV technologies in energy
generation [8]. These configurations present advantages in
relation to centralized configurations, such as better utilization
of the energy generated by the PV panels. This configuration
has two or more PV arrays, where each array has its own DCDC converter to perform the MPPT. This feature provides a
considerable reduction of the effects caused by mismatches or
partial shading of the panels, which can produce a gain of up
to 16% of generated energy compared to centralized systems
[15]. On the other hand, the main disadvantage of this
configuration is the control complexity.
The proposed decentralized system for stand-alone PV
applications shown in Figure 2 is based on the topology used

in grid-connected systems [7], [11], [12], [16]. The main

feature of this topology is the use of two or more
independent arrays of PV panels, each with its own DC-DC
converter. In order to increase the efficiency, the DC bus is
fed by the series connection of outputs of DC-DC boost
converters. When there is no difference of energy production
among PV arrays, the DC bus voltage is equally distributed
among the output of n-series boost converters. In this case,
the step-up voltage of each individual boost converter is
small, increasing the PV conversion efficiency.
The proposed topology uses a series connection of two
boost converters in the input stage. Both converters can be
controlled independently by the use of individual MPPT
algorithms, which control S1 and S2 switches, respectively.
These converters operate in a master/slave mode, which is
used to reduce the energy generation when load and battery
demand are lower than PV production capability. The slave
converter is the first one to disable the MPPT algorithm
when necessary, while the master converter is the last one to
disable the function to track the maximum power. In order to
perform the interface between the DC bus and the battery
bank, the proposed system presents a bidirectional DC-DC
converter, which is comprised of switches S3 and S4 (in leg
configuration) and a T filter [17]. It can operate in Buck mode
(using S3 and the antiparallel diode of S4), when it is necessary
to charge the battery bank, or in Boost mode (using S4 and the
antiparallel diode of S3), when the energy generated by the PV
array is not sufficient to supply the load demand.
The T filter is used to minimize the high frequency current
ripple through the batteries. This filter is important to
increase the batteries lifetime, since continuous ripple
current may reduce battery life due to battery heating. As a
result, this filter can reduce the costs involved with battery
maintenance. Moreover, in order to further increase the
batteries lifetime, the IU charging method is used (one level

L1

PV
PV

vPVs

L1

Sa

L2

D2

S1

S3

PV

Lout

PV

D1

Cpv

PV

C1

Bat

C2

Sb

S2

C1

vPVm

L2

D2

S2
CPV2

C2

S4

PV

PV

(a)

PV

Buck
DC
DC

Boost
DC

Inverter
AC IAC

DC

DC

Lf iLf

Lb iLb

Cout

z vAC

iBat
vC2

vBat

S4

Cf

Boost
IPV DC

IPV

IPV

vDC

S6

S8

(a)
Array

PV

Array

S7

PV

PV

Battery
Bank

S5

Lout i
AC

Cout
Rout

S3

vC1

CDC

PV

Dpv

CPV1

PV

PV

PV

D1

S1

DC

Inverter
AC IAC

Boost
DC

DC

DC

DC
DC

Load

Bidirectional
Converter

IBat
Load

(b)
Fig. 1. Centralized stand-alone PV system. (a) Schematic. (b)
Block diagram.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

Battery
Bank

(b)
Fig. 2. Decentralized stand-alone PV system. (a) Schematic. (b)
Block diagram.

221

of voltage and one level of current). A Full-Bridge inverter is

connected to the DC bus to generate the AC voltage to the
loads. Figure 2 shows this converter, which is comprised of
S5S8 switches and a Lout Cout output filter. The impact of the
DC bus voltage ripple on the output voltage is minimized by
a closed loop control of the inverter.
The stand-alone PV systems present variations in solar
power generation, in the energy stored in the battery bank
and in the energy demanded by the load. Therefore, in order
to obtain a regulated DC bus, an overall control strategy for
power management was considered, which takes into
account different conditions for the input energy, the
storage energy and the load demand. Since load demand is
unpredictable and uncontrollable, the output power can be
considered a disturbance for DC bus regulation. Therefore,
the closed-loop control of the inverter is not addressed in
this paper.
III. METHODS OF CHARGING BATTERIES
Choosing the charging method for the battery bank is not
a trivial task. Important factors must be taken into account,
such as the type of battery, charging time, protection
against overload and excessive discharge. Among the most
used methods for charging batteries, it can mention: (i) one
level of voltage; (ii) one level of current; (iii) two levels of
voltage, two levels of current; (iv) two levels of current and
one of voltage; (v) one level of voltage and one of current;
and (vi) the method of charging by pulsed current [18].
The method used to charge the batteries when the
bidirectional converter operates as Buck mode is the IU
method. In this method, the charging current is regulated up
to 10% of the nominal Ah capacity of the battery bank until
the voltage reaches the level of equalization (2.4V/e). Thus,
when this condition is achieved, a constant voltage level of
fluctuation is applied. This method appears to be the most
suitable to recharge batteries because the charge current
imposed on the batteries is controlled, giving enough time
for water recombination and to prevent losses [18].
IV. OPERATION MODES
The control objective of the proposed PV system is the
regulation of the DC bus voltage independent of solar
irradiation, battery charge and load demand. This objective
is achieved by the DC bus voltage regulation in the
neighborhood of its nominal reference. In order to obtain a
constant voltage over CDC, the following power balance
must be ensured:
PBUS = PVM + PVS + PBB PL = 0 ,

(1)

where PBUS is the power flow on CDC, PVM and PVS are PV
power converted by master and slave boost converters,
respectively, PBB is the power extracted (+) or injected (-)
on the bidirectional converter, and PL is the load demand.
In order to analyze the operation modes, it is important to
highlight that the solar power generation, the energy stored
in the battery bank and the load demand are variable.

222

A. Operating Modes of Individual Converter

1) PV DC-DC Boost Converters There are three possible
operation modes for both PV master (PVCm) and slave
(PVCs) converters: MPPT Mode, DC Bus Regulation
(DCBR) Mode and Off Mode. MPPT Mode is used in
normal situations to extract the maximum power from each
array of panels. DCBR Mode is used to regulate the DC bus
when the energy supplied by the PV array is higher than the
energy demanded by the load and the battery bank. In this
case, the energy extracted from the PV array is that necessary
to ensure the power balance given in (1).
2) Bidirectional DC-DC Battery-Bank Converter The
Bidirectional DC-DC Battery Bank Converter (BBC) can
operate in four modes: Boost, Buck, Float and Off. Boost
Mode is used when the energy produced by the PV arrays is
not sufficient to supply the load. In this case, the battery bank
supplies the additional energy necessary to maintain the DC
bus regulated. On the other hand, Buck Mode is used to
charge the battery bank when the energy produced by the PV
arrays is higher than the energy demanded by the load. Float
Mode is a particular case of Buck mode, which is enabled
when batteries are completely charged, avoiding their
discharge in stand-by. The off mode occurs when the battery
bank is discharged and energy produced by PV arrays is
sufficient only to supply load.
3) Inverter Control strategy considers two operating
modes for inverter (INV): On Mode and Off Mode.
B. Operating Modes of Overall System
1) Mode 1: Power generated by the panels is equal to power
consumed by load Figure 3 (a) shows the energy flow when
the power generated by PV panels is exactly that required to
meet the power demand by the load. Both PVCm and PVCs
operate in MPPT Mode, INV is On, and BBC is Off.
2) Mode 2: Power generated by the panels is smaller than
the power consumed by load Figure 3 (b) shows the energy
flow when the power generated by the PV panels is lower than
the load. The battery bank supplies the additional power in order
to maintain the DC bus regulated. Both PVCm and PVCs are
operating in MPPT mode and BBC operates in boost mode.
3) Mode 3: Power generated by the panels is higher than
the power consumed by load and the batteries are not fully
charged Figure 3 (c)-(e) shows the energy flow when the
power generated by the PV panels is higher than the load.
DC bus is regulated by using the exceeding energy to charge
the batteries. Three sub-modes can be defined:
Mode 3a: Figure 3 (c) shows the case where batteries are
charged with a lower current than the maximum
charging current. Thus, both PVCm and PVCs
converters operate in MPPT Mode and the BBC operates
in Buck Mode.
Mode 3b: Figure 3 (d) shows the energy flow when
batteries are charged with the maximum charging
current. In this case, DC link power balance is achieved
by disabling MPPT algorithm of PVCs. The converter
operates now in DCBR Mode, i.e., it works out of

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

Array
PV

PVCs = MPPT
DC

PVS > 0

INV = ON
AC

DC

DC

DC

PV

Load

PL > 0

PVS > 0

Array
INV = ON
AC

DC

DC

DC

DC
BBC = OFF
PVM > 0 DC
DC
PVCm = MPPT
PBB = 0

PV

PVCs = MPPT
DC

Array

PV

PV

PVS > 0

PV

PVCs = OFF
DC

Array

DC

DC

DC

PV

Load

PVS > 0

DC
PV

PBB < 0 (IBat = IBatmax)

PV

PVS > 0

DC

PV

PVCs = MPPT
DC

Array

DC

PV

Load

PL > 0

PVS > 0

DC
DC

DC
BBC = FLOAT
PVM > 0 DC
DC
PVCm = DCBR
PBB = 0

PV

DC

DC

BBC = BUCK

PV

PBB < 0 (IBat = IBatmax)

INV = ON
AC

DC

PL > 0

DC

PVM > 0 DC
DC
PVCm = MPPT

DC
DC

PVM > 0 DC
DC
PVCm = DCBR

Array
PV

Load

PL = 0

PVCs = DCBR
DC

PVS > 0

DC

DC

DC

PV

PL = 0

Battery
Bank

(j) Mode 5c

INV = OFF
AC

Load

PL = 0

DC
BBC = BUCK
PVM > 0 DC
DC
PVCm = MPPT
PBB < 0 (IBat = IBatmax)

(i) Mode 5b

PV

PBB < 0 (IBat = IBatmax)

PBB = 0

Battery
Bank

Load

BBC = BUCK

BBC = FLOAT

(f) Mode 4a

INV = OFF
AC
DC

Load

PL > 0

Battery
Bank

Array

INV = OFF
AC

DC
DC

PV

PVS > 0

(h) Mode 5a

PVCs = OFF
DC

PVS > 0

Load

PVCs = DCBR
DC

Battery
Bank

(g) Mode 4b

PV

PV

DC
BBC = BUCK
PVM > 0 DC
DC
PVCm = MPPT
PBB < 0 (IBat < IBatmax)

Battery
Bank

Array

(c) Mode 3a

(e) Mode 3c

INV = ON
AC

DC

DC
DC

Load

PL > 0

DC
BBC = BUCK
PVM > 0 DC
DC
PVCm = MPPT
PBB < 0 (IBat < IBatmax)

Battery
Bank

(d) Mode 3b
PVCs = OFF
DC

PV

Array

PVM > 0 DC
DC
PVCm = DCBR

Battery
Bank

Array

DC

DC

Battery
Bank

INV = ON
AC

DC

PL > 0

BBC = BUCK

PVM > 0 DC
DC
PVCm = MPPT

INV = ON
AC

DC

(b) Mode 2

INV = ON
AC

DC

PVS > 0

Battery
Bank

(a) Mode 1
PVCs = DCBR
DC

PL > 0

DC
BBC = BOOST
PVM > 0 DC
DC
PVCm = MPPT
PBB > 0

Battery
Bank

Array

PV

Load

PVCs = MPPT
DC

PVCs = OFF
DC

PVS > 0

DC

INV = OFF
AC
DC

DC

PV

Load

PL = 0

DC
BBC = FLOAT
PVM > 0 DC
DC
PVCm = DCBR
PBB = 0
Battery
Bank

(l) Mode 6

Fig. 3. Operation modes of decentralized stand-alone PV system.

maximum power point (MPP) to ensure the DC bus

power balance.
Mode 3c: Figure 3 (e) shows the energy flow when
batteries are also charged with maximum charging
current. However, in this case, the power required to
charge the batteries is lower than only PVCm can
supply. So DC link power balance is achieved by
disabling PVCs and changing the PVCm to operate in
the DCBR Mode.
4) Mode 4: Power generated by the panels is higher than
the power consumed by load and the batteries are fully
charged In this mode, the exceeding power cannot be used
to charge the batteries, since they are fully charged. In this
case, BBC operates in Float Mode. Thus, the DC bus is
regulated by disabling the MPPT algorithm of PVCs and/or
PVCm. Figure 3 (f) and 3 (g) show two sub-modes, where

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

the regulation of vDC is performed analogously to Modes 3b

and 3c, respectively.
5) Mode 5: Inverter is off and the batteries are not fully
charged This mode occurs when the power generated by
the panels is lower than the power consumed by load and the
batteries are fully discharged. In this case, the inverter must
be turned off to avoid the collapse of the DC bus voltage.
However, the power generated by the panels is used to
charge the batteries. Figure 3 (h)-(j) show three possible submodes, where DC bus regulation is analogous to modes 3a to
3c, respectively.
6) Mode 6: Inverter is off and the batteries are fully
charged In this mode, shown in Figure 3 (l), it was
considered that the inverter was previously turned off to
avoid the DC bus voltage collapse, that the batteries are fully
charged and BBC operates in float mode. Thus, the DC link

223

power balance is achieved by disabling PVCs and changing

the PVCm from MPPT to DCBR mode. It is worth
mentioning that it was assumed that the inverter could not be
turned on automatically due to safety requirements.
V. MPPT ALGORITHM
One of the obstacles to the dissemination of PV
technology is its high installation cost. Thus, it is essential to
drain the maximum available power from PV panels to
increase the global efficiency of the system and to reduce
generated energy costs.
An important characteristic of PV panels is the existence
of only one point of operation from which the maximum
power can be extracted. The MPP (Maximum Power Point)
for a given level of solar radiation and temperature
correspond to a single value of voltage (VMPP) and current
(IMPP). Therefore, the objective of all MPPT techniques is to
find this point independently of environmental conditions.
The following features should be observed in the choice of
the MPPT technique: time taken to reach the MPP, steadystate oscillation, and sensitivity to parametric variations.
In the proposed system, a modified P&O (Perturb &
Observe) algorithm was used for MPPT. This algorithm
changes the MPP by means of duty ratio modification in
each Boost converter. Table I summarizes the operation of
this algorithm [19].
Although the conventional and modified P&O algorithms
present identical control logic to define the signal of the next
perturbation, the conventional P&O algorithm uses a
constant perturbation whereas the modified P&O algorithm
uses a variable perturbation whose module is proportional to
variations of power. This slight modification in the
conventional algorithm results in a short tracking time and
minimization of oscillations around the MPP.
VI. SIMULATION RESULTS
This section presents the simulation results for the
proposed PV system. The system was subjected to variations
on load and/or solar radiation in order to analyze its behavior
under the different operation modes analyzed. The
specifications of the PV system are summarized in Table II.
As detailed in Section IV, this system has several
operation modes and most of them have similar
characteristics. In order to not include repetitive information,
this section presents simulation results for each operating
mode of the input boost converters (MPPT and DCBR
functions) and for each operating mode of the bidirectional
converter (buck and boost modes).
Figure 4 (a) shows the simulation results for a single input
boost converter tracking the maximum power point of the PV
arrays. A sudden radiation change from 700W/m2 to
1000W/m2 was imposed on the PV array to show that the
proposed modified P&O algorithm presents a fast dynamic
response and reduces steady state oscillations.
Figure 4 (b) presents the power extracted from PV arrays
when the input slave Boost converter changes from MPPT to
DCBR mode. DCBR mode is used when there is an excess of

224

TABLE I
Summary of P&O algorithm
Pertubation
Positive
Positive
Negative
Negative

Change in Power
Positive
Negative
Positive
Negative

Next Pertubation
Positive
Negative
Negative
Positive

TABLE II
Specifications of the PV System
Converter/Stage

PV Boost
Converters

Bidirectional
Converter

Voltage-Source
Inverter

Battery Bank

Parameter
Total Power (Pout)
Individual Power (PPV)
Switching Frequency
PV Array Capacitor (CPV1,2)
Inductor (L1,2)
Output Capacitor (C1,2)
Boost Power
Boost Switching Frequency
Buck Power
Buck Switching Frequency
Buck/Boost Inductor (Lb)
Filter Capacitor (Cf)
Filter Inductor (Lf)
Current Ripple though Lb
Power
DC Bus Voltage Level (vDC)
Output AC Voltage (vAC)
Switching Frequency
Filter Capacitor (Cout)
Filter Inductor (Lout)
Voltage Level (vbat)
Max. Battery Charging Current
(IBatmax)
Charge Current Ripple

Value
480 W
240 W
30 kHz
68 F
800 H
100 F
480 W
40 kHz
216 W
100 kHz
245 H
1000 F
3,4 H
10%
480 W
200 V
127 Vrms
30 kHz
4 F
600 H
48 V
4,5 A
0,2%

energy generated by PV power. DCBR operation disables the

MPPT algorithm of the slave input boost converter in order
to extract only the energy demanded by load and batteries. In
this simulation, a load step is performed at 0.33s to force
commutation from Mode 1 to Mode 4a (Section IV). As can
be seen, DCBR mode is imposed on the slave boost
converter while the master continues extracting the
maximum power from its PV array. Figure 4 (c) presents the
DC bus voltage being regulated during the DCBR operation.
Figure 4 (d) and 4 (e) present simulation results when
power consumed by load is greater than power generated by
PV arrays and the batteries are not fully discharged. In this
case, the condition given in (1) is reached by activation of the
bidirectional converter in boost mode to transfer energy from
battery bank to DC bus. During this condition, the input
boost converters are extracting all power available in the
panels. Figure 4 (d) shows the DC bus voltage when a
commutation from Mode 1 to Mode 2 occurs. The DC bus
voltage is regulated by the bidirectional converter after
commutation between modes. Figure 4 (e) presents the
current extracted from the battery bank. As can be seen, the
bidirectional converter injects current to regulate the DC bus
voltage. Moreover, it is worth mentioning that a
commutation of the bidirectional converter from
discontinuous conduction mode (DCM) to continuous
conduction mode (CCM) occurs at 0.63s.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

500

250
MPP @ 1000W/m2

MPP @ 700W/m2

160
140

300

PPV (W)

700W/m2

PPV_Master

250
200

100

50

0
0

0,05 0,1

0,15 0,2

100

60

100

40

50

20

0,25 0,3 0,35 0,4 0,45 0,5

Time (s)

120

80

PPV_Slave

0,05 0,1

0,15 0,2

(a)

0,25 0,3 0,35 0,4 0,45 0,5

Time (s)

205

205

200

200

195

195

190

190

185

VDC (V)

210

5
4

180

175

175

170

165

Boost ON

165
0,5

0,6

0,7
Time (s)

0,8

0,9

0
0,4

0,5

0,6

(d)

0,7
Time (s)

0,8

0,9

160

0,6

0,8

(e)

ILb
ILf

2
58

1,2

1,4
1,6
Time (s)

1,8

2,2

ILb
ILf

1,5

57

IBat (A)

VBat (V)

57,5

1,5

2,5

58,5

0,25 0,3 0,35 0,4 0,45 0,5

Time (s)

(f)

59

2,5

0,15 0,2

185

180

160
0,4

0,05 0,1

(c)

10

170

(b)

210

IBat (A)

VDC (V)

VCC (V)

150

DCBR ON

180

350

150

IBat (A)

200

400

200

Power (W)

220

PPV_Total

450

1000W/m2

56,5

1
56

0,5

0,5
0

55,5

0,5

1
Time (s)

1,5

55
0,3

0,32 0,34 0,36 0,38 0,4 0,42 0,44 0,46 0,48 0,5
Time (s)

(g)

(h)

0
0,3

0,32 0,34 0,36 0,38 0,4 0,42 0,44 0,46 0,48 0,5
Time (s)

(i)

Fig. 4. Simulation results. (a) Power extracted from PV array under a sudden radiation change.(b)-(c) Power extracted from PV arrays and
DC bus voltage Modes 1 and 4a. (d)-(e) DC bus voltage and current extracted from the battery bank Modes 1 and 2. (f)-(g) DC bus
voltage and current drained by the bidirectional converter Modes 1 and 3. (h)-(i) Battery bank voltage level and current injected into the
battery bank Modes 3a and 4a.

In this operating mode, the battery bank voltage decreases

during the discharge of the batteries. The battery bank must
be turned off when its voltage level reaches the minimum
harmless value. In this case, the power available to feed the
load is reduced and DC bus voltage can drop. In this
situation, the energy quality delivered to the load is not
penalized if only critical loads remain powered (Mode 1).
Another alternative is to turn off the inverter and use the PV
energy to recharge the battery bank (Modes 5a, 5b and 5c).
Figure 4 (f) and 4 (g) present simulation results when the
power generated by the PV array is higher than the power
consumed by the load and the batteries are not fully charged.
In this case, there is a transition from Mode 1 to Mode 3a at
0.33s (Figure 4 (g)) to avoid the excessive increase of the DC

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

bus voltage. Figure 4 (f) shows the DC bus voltage and

Figure 4 (g) shows the current through the batteries. As can
be observed, the regulation of the DC bus voltage is achieved
by charging the battery bank with a constant current. This
current is the difference between the energy generated by the
PV array and the energy consumed by the load.
Figure 4 (h) and 4 (i) show the simulation results when the
voltage of the battery bank reaches its maximum value (fully
charged condition). The charging controller changes the
battery bank control strategy from the current loop to the
voltage loop at 0.38s (transition from Mode 3a to Mode 4a).
After this, the charging of the batteries is stopped and the
converter is controlled to maintain the equalization voltage
level on the batteries (IU charging method).

225

250

1000W/m

vDC

Power (W)

200

vDC
150

100
0

700W/m

iBat

10
12
Time (s)

14

16

18

20

Ch1: 50 V/div

(a)

Ch1: 50 V/div

Time: 2 s/div

Ch2: 0,5 A/div

(b)

Time: 5 s/div

(c)
VAC
IAC

vDC
iBat

iBat
Ch1: 50 V/div

Ch2: 0,5 A/div

Time: 5 s/div

Ch1: 0,5 A/div

Time: 1 s/div

Ch1: 50 V/div

Ch2: 2 A/div

Time: 5 ms/div

(d)
(e)
(f)
Fig. 5. Experimental Results. (a) MPPT under a sudden radiation change. (b) DC bus voltage Modes 1 and 4a. (c) DC bus voltage and
battery bank current drained by bidirectional converter Modes 1 and 2. (d) DC bus voltage and battery bank current supplied by
bidirectional converter Modes 1 and 3a. (e) Battery bank current Modes 3a and 4a. (f) Output voltage and current.

VII. EXPERIMENTAL RESULTS

This section presents experimental results that
demonstrate the behavior of the overall PV system and
confirm the analyses presented in the previous sections. The
bidirectional converter was implemented with IRG4PC50W
discrete IGBTs and 15ETH06 discrete diodes. Both input
Boost converters were implemented with IRFP360LC
MOSFETs and RHRP970 discrete diodes. The control
strategy was implemented in a Xilinx Spartan3E FPGA.
Additionally, the PV arrays were emulated by the Agilent
E4350B Solar Array Simulator.
Figure 5 (a) presents the performance of the MPPT
algorithm when subjected to a sudden variation in solar
radiation from 700W/m2 to 1000W/m2. It can be observed
that the proposed algorithm quickly tracks the MPP with
slight oscillations in steady state, around the MPP. Figure 5
(b) shows the DC bus voltage being regulated when the input
slave boost converter changes from MPPT to DCBR mode
(Mode 1 to Mode 4a). In DCBR Mode, the slave input Boost
converter operates out of the maximum power point to
ensure the power balance given in (1).
Figure 5 (c) presents experimental waveforms of DC bus
voltage (vDC) and battery bank current (iBat) when the power
generated by PV arrays is lower than the power consumed by
the load (transition from Mode 1 to Mode 2). The
supplementary power to guarantee the power balance is
supplied by the battery bank. One can observe that the
current drained from the battery bank is that required to
regulate the DC bus voltage at 200V. The current peak
during the start-up is a characteristic of the bidirectional
converter operating in boost mode. It is worth mentioning

226

that an inverse direction was used for reference current in

relation to Figure 5 (c).
Figure 5 (d) presents the waveforms for DC bus voltage
(vDC) and battery bank current (iBat) when there is a transition
from Mode 1 to Mode 3a. In this case, the exceeding power
generated by PV arrays is stored in the battery bank. As can
be seen, the bidirectional converter operates in buck mode to
charge the battery bank with the current necessary to regulate
the DC bus at the specified voltage (200V in this case).
Figure 5 (e) shows the battery bank current waveform when
a transition occurs in the bidirectional converter from buck
mode to float mode (Mode 3a to Mode 4a). Initially, the
bidirectional converter is used to regulate the DC bus voltage
while the batteries are charged. When the battery bank voltage
reaches its maximum value, the bidirectional converter control
commutes to the voltage loop. In this mode, an equalization
voltage level was maintained on the battery bank by injecting
the necessary current to achieve natural discharge
compensation. The slave input boost converter changes from
MPPT to DCBR mode to regulate the DC bus voltage.
Experimental results for the sinusoidal output waveforms are
presented in Figure 5 (f). These waveforms were obtained for
the nominal output voltage (127Vrms) and full load condition.
VIII. CONCLUSION
A decentralized stand-alone PV system is presented and
analyzed in this paper. This system presents important
advantages when compared with centralized systems, such as
the improvement in efficiency in cases of partial shading, the
possibility of placing panels in different orientations and
increased reliability, among others.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

However, the proposed topology requires a more complex

overall power management strategy to control power flows
among PV arrays, battery bank, and inverter in order to
maintain the DC bus regulation. The control strategy used
was validated by simulation and experimental results. A
control scheme was proposed for each individual power
electronic interfacing converter. In order to extract the
maximum power available from the PV arrays, an individual
MPPT algorithm based on the modified P&O algorithm was
used for each PV array. The IU charging method was chosen
to control the battery bank charging current, which
maximizes the lifetime of the batteries.
Simulation and experimental results were provided to
validate the system operation in different operating modes.
As the focus of this paper is to present and analyze the
behavior of the decentralized PV system under different load
and radiation conditions, the efficiency analysis of the
converters was not addressed.
ACKNOWLEDGEMENT
The authors would like to thank Dr. Mrio Lcio da Silva
Martins, M.Sc. Felipe Grigoletto, M.Sc. Paulo Ficagna and
Eng. Matheus Desconzi for helpful discussions in order to
improve the paper, and the Conselho Nacional de
Desenvolvimento Cientfico e Tecnolgico CNPQ (proc.
311096/2006-9 and 307798/2009-7) for financial support.
REFERENCES
[1] European Photovoltaic Industry Association, Solar
generation V, 2008.
[2] D.C. Martins, Conversores estticos de potncia
utilizados no processamento da energia solar
fotovoltaica, in IV Conferncia Internacional de
Aplicaes Industriais, vol. 2, pp. 67-77, 2000.
[3] J.H.R. Enslin, M. S. Wolf, D. B. Snyman, and
W. Swiegers, Integrated photovoltaic maximum
Power point tracking converter, IEEE Transactions on
Industrial Electronics, vol. 44, no. 6, pp. 769-773,
1997.
[4] G. F. Rodrigues, Estudo sobre baterias para aplicao
em sistemas fotovoltaicos isolados, Relatrio Interno
GEPOC-UFSM, 2005.
[5] International Energy Agency, Management of storage
batteries used in stand-alone photovoltaic power
systems, Report_IEA_PVPS_T3-10:2002, 2002.
[6] B. Lindgren, A power converter for photovoltaic
applications, Technical Report no. 335L, Chalmers
University of Technology, Gteborg, Sweden, 2000.
[7] M. Meinhardt and G. Cramer, Multi-string converter:
the next step in evolution of string-converter
technology, in 9th European Conference on Power
Electronics and Applications, 2001.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010

[8] M. Calais, J. Myrzik, T. Spooner, and V. G. Agelidis,

Inverters for single-phase grid connected photovoltaic
systems an overview, in 33rd IEEE Power
Electronics
Specialists
Conference,
vol.
4,
pp. 1995-2000, 2002.
[9] J. Myrzik and M. Calais, String and module integrated
inverters for single-phase grid connected photovoltaic
systems a review, in IEEE Bologna Power Tech
Conference Proceedings, vol. 2, 2003.
[10] F. Blaabjerg, Z. Chen, and S. B. Kjaer, Power
electronics as efficient interface in dispersed power
generation systems, IEEE Transactions on Power
Electronics, vol. 19, no. 5, pp. 1184-1194, 2004.
[11] G. R. Walker and P. C. Sernia, Cascaded DC-DC
converter connection of photovoltaic modules, IEEE
Transactions on Power Electronics, vol. 19, no. 4,
pp. 1130-1139, 2004.
[12] G. R. Walker and J. C. Pierce, Photovoltaic DC-DC
module integrated converter for novel cascaded and
bypass grid connection topologies, in 37th IEEE
Power Electronics Specialist Conference, pp. 30943100, 2006.
[13] S. B. Kjaer, J. K. Pedersen, and F. Blaabjerg, A
review of single-phase grid-connected inverters for
photovoltaic modules, IEEE Transactions on Industry
Applications, vol. 41, no. 5, pp. 1292-1305, 2005.
[14] G. R. Walker and P. C. Sernia, Cascaded DC-DC
converter connection of photovoltaic modules, IEEE
Transactions on Power Electronics, vol. 19, no. 4, pp.
1130-1139, 2004.
[15] J. Imhoff, J. R. Pinheiro, J. L. Russi, D. Brum, R. Gules
and H. L. Hey, DC-DC Converters in a multi-string
configuration for stand-alone photovoltaic systems, in
39th IEEE Power Electronics Specialists Conference,
pp. 2806-2812, 2008.
[16] P. C. Sernia and G. R. Walker, Multi-converter
topology evaluation for connection of low voltage DC
sources, in Australasian Universities Power
Engineering Conference, vol. 1, pp. 495-500, 2001.
[17] L. Schuch, C. Rech, H. L. Hey, H. A. Grndling, H.
Pinheiro, and J. R. Pinheiro, Analysis and design of a
new high-efficiency bidirectional integrated ZVT PWM
converter for DC-bus and battery-bank interface, IEEE
Transaction on Industry Applications, vol. 42, no. 5,
pp. 1321-1332, 2006.
[18] L. Schuch, Sistema CA/CC com um conversor PWM
bidirecional para interface entre o barramento CC e o
banco de baterias (Master Thesis), Programa de Ps
Graduao em Engenharia Eltrica, Universidade
Federal de Santa Maria, 2001.
[19] T. Esram and P.L Chapman, Comparison of
photovoltaic array maximum power point tracking
techniques, IEEE Transactions on Energy Conversion,
vol. 22, no. 2, pp. 439-449, 2007.

227

BIOGRAPHIES
Diogo B. Cndido was born in Porto Alegre, RS, Brazil,
in 1985. He received the B.S. and M.S degrees in electrical
engineering from the Federal University of Santa Maria,
Santa Maria, Brazil, in 2007 and 2010, respectively.
From 2004 to 2010, he was with the Power Electronics and
Control Research Group (GEPOC), Federal University of
Santa Maria. Currently, he is Product Development Engineer
with WEG S.A., Jaragu do Sul, Brazil. His research
interests include power electronics, frequency AC motor
drivers and renewable energy applications.
Jonatan Rafael Rakoski Zientarski was born in Iju,
Brazil. He received the B.S. degree in electrical engineering
from the Regional University of Northwest Rio Grande do
Sul State, Brazil, in 2006 and the M.S. degree in electrical
engineering at the Federal University from Santa Maria,
Brazil, in 2009. Since 2010, he has been with Instituto
Federal Catarinense, where he is currently a Professor. His
research interests include power converters modeling, design
and optimization, power factor correction and EMI
suppression techniques.
Rafael Concatto Beltrame was born in Santa Maria,
Brazil, in 1984. He received the B.S. and M.S. degrees in
electrical engineering from the Federal University of Santa
Maria, Brazil, in 2008 and 2009, respectively, where he is
currently working toward the Ph.D. degree. Since 2005, he has
been with the Power Electronics and Control Research Group
(GEPOC), Federal University of Santa Maria. His research
interests include electric motor drives, high-performance
power converters, and soft-switching techniques.

228

Jos Renes Pinheiro was born in Santa Maria, Brazil,

1958. He received the B.S. degree from the Federal
University of Santa Maria, Santa Maria, Brazil, in 1981, and
the M.S. and Ph.D. degrees from the Federal University of
Santa Catarina, Florianpolis, Brazil, in 1984 and 1994,
respectively, all in electrical engineering. His current
research interests include high-frequency and high-power
conversion, power supplies, multilevel converters, powerfactor-correction techniques, and modeling and control of
converters.
Leandro Michels was born in No-Me-Toque, Brazil, in
1979. He received the B.S and Ph.D. degrees from the
Federal University of Santa Maria, Santa Maria, Brazil, in
2002 and 2006, respectively, both in electrical engineering.
Since 2009 he has been with the Power Electronics and
Control Research Group (GEPOC) at Federal University of
Santa Maria, where he is currently an Adjunct Professor. His
current research interests include modeling and control of
power converters, applied digital control and renewable
energy applications.
Hlio Lees Hey was born in Santa Maria, Brazil, in
1961. He received the B.S. degree from the Catholic
University of Pelotas, Pelotas, Brazil, in 1985, and the M.S.
and Ph.D. degrees from the Federal University of Santa
Catarina, Florianpolis, Brazil, in 1987 and 1991,
respectively. Since 1994, he has been with the Power
Electronics and Control Research Group (GEPOC), Federal
University of Santa Maria, where he is currently a Full
Professor.

Eletrnica de Potncia, Campinas, v. 15, n. 3, p.220-228, jun./ago. 2010