Passivity-Based Control for Charging Batteries in

Photovoltaic Systems
Valentim E. Neto
1
, Filipe Perez
1
,Andr G. Trres
1
, Allan F. Cupertino
1,2
and Heverton A. Pereira
1,2

1
Gerncia de Especialistas em Sistemas Eltricos de Potncia
Universidade Federal de Viosa
Av. P. H. Rolfs s/n, 36570-000
Viosa, MG, Brazil
filipebmx@hotmail.com, valentimernandes@hotmail.com,
angoto@ufv.br
2
Graduate Program in Electrical Engineering
Federal University of Minas Gerais
Av. Antnio Carlos 6627, 31270-901
Belo Horizonte, MG, Brazil
allan.cupertino@yahoo.com.br, heverton.pereira@ufv.br
Abstract This paper proposes the use of passivity-based
control in a battery charger for isolated photovoltaic systems. It
uses a boost converter acting as maximum power point tracker
(MPPT). This technique is compared with the traditional
strategy based on proportional-integral (PI) linear controllers.
The dynamic of the controlled system during variations in the
solar radiation and the influence of the inductor parameters
uncertainty in the response of the battery charger are analyzed.
I ndex Terms Battery, Boost Converter, Passivity Based
Control, Maximum Power Point Tracker, Photovoltaic Panel.
I. INTRODUCTION
The generation of electricity through alternative sources
has been widely studied, due to countries wanting to diversify
its energy matrix focusing on power generation without
pollution and with as low environmental degradation as
possible. In this context, the solar photovoltaic energy has
been growth considerably in the last years [1].
Photovoltaic systems can be classified into three main
categories: hybrid, grid-connected and isolated. The use of
each is a function of the application and availability of energy
resources. The isolated system can be applied in rural areas of
difficult access. This system is easily adapted to the buildings
and roofs of houses [2].
Until 2012 had been installed in the world 102 GW in
photovoltaic systems, as shown in Fig. 1, and more than 95%
of these are connected to the grid. This represents an increase
of 31 GW in the last year. With this power level is possible to
save 53 million tons of

[1].


Fig. 1 Installed power on solar photovoltaic systems until 2012 [1].
The battery bank is an important element of the isolated
system and can represent until 15% of initial costs for
installation, reaching to 46% if considered the maintenance
costs [3]. This is explained by the fact that batterys life cycle
is smaller than the others components of the photovoltaic
system. Furthermore, the batteries are subjected to the most
diverse operating conditions, due to nonlinear behavior of
photovoltaic panels, [4] , [5].
Nickel-metal hydride batteries have as advantages the
following points [6], [7]:
It can be loaded several times a day;
reduced memory effect;
long life cycle;
low maintenance requirements ;
it is less toxic than others.

However, this type of battery has a high cost, high risk of
damage to overload and high self-discharge [6], [7].
In order to increase the lifetime of the batteries, the
charging method needs to consider the charge state of the
battery and its rated values. When the battery is discharged the
system can works at maximum power point if the rated current
of the battery is not exceeded. If the rated current is exceeded
or the battery reaches full charge, the MPPT algorithm is
maintained in standby mode. Therefore, the battery is
protected against overload.
Many works in literature study the control of battery
chargers. A traditional methodology is the proportional-
integral (PI) control technique [8]. Several authors have been
proposed nonlinear techniques like the passivity-based control
(PBC), due to the existence of large disturbances in the
generated power [9], [10].
The PBC is based on energy functions and finds a state of
operation in which the plant stores the minimum energy. This
method of control has some advantages over traditional
techniques, because it is not necessary to linearize the system
around the operation point. This characteristic can improve the
response of non-minimum phase systems, like switching
converters. Using PBC it is possible to obtain a robust battery
charger using a boost converter.
This work proposes the use of passivity-based control in a
battery charger for a photovoltaic system. It is used a boost
converter acting as maximum power point tracker (MPPT).
5,3
6,9
9,5
16,2
23,6
40,6
71,1
102,1
2005 2006 2007 2008 2009 2010 2011 2012
I
n
s
t
a
l
l
e
d

P
o
w
e
r

(
M
W
p
)

Year
The authors would like to thank CNPq, FAPEMIG and CAPES by their
financial support.
This technique is compared with the traditional linear PI
strategy. The topology of isolated photovoltaic system is
presented in Fig. 2.


Fig. 2 Battery charger topology.
II. MODELING OF THE SYSTEM
A. Maximum Power Point Tracker (MPPT)
The MPPT is an algorithm that maintains the photovoltaic
panel delivering maximum power to the system at various
levels of solar radiation. It was used the incremental
conductance algorithm. The incremental conductance
algorithm is considered better than other techniques based on
the principle of perturbation and observation, because it has a
faster response to changes in solar radiation.
The operating principle of the algorithm is the incremental
calculation of the derivative of the curve of power once this
value is zero at the maximum power point [11]. Fig. 3 shows
the algorithm.

Fig. 3 Incremental conductance algorithm [8].
B. Solar panel modeling
The electrical model of the solar panel used in this work is
shown in Fig. 4 and is presented in details by [12]. The
resistances represent the voltage drop and losses for both the
current flowing through the load (

) and the reverse leakage

), respectively.

is a controlled DC
current source. In Table I the parameters of a solar panel
SM48KSM, manufactured by Kyocera are presented.

Fig. 4 Photovoltaic panel model.
TABLE I. MAIN PARAMETERS OF A SOLAR PANEL.
Parameter Symbol Value
Maximum Power ()

48
Maximum Power Voltage ()

18.6
Maximum Power Current ()

2.59
Open Vircuit Voltage ()

22.1
Short Circuit Current ()

2.89
Temperature Coefficient of

( )

-0.07
Temperature Coefficient of

( )

0.00166

C. Boost converter modeling and PI control
The modeling of the boost converter assumes that the
battery voltage is constant. This is a good approximation
because the battery voltage variation during the charging
process is very small. Besides, the curve of the solar
panel is linearized around the maximum power point, like is
shown in Fig. 5 [8]. In this situation, the solar panel can be
modeled as a voltage source in series with a resistance. Thus,
the battery charger equivalent circuit is shown in Fig. 6.

Fig. 5 I x V curve of the solar panel SM48KSM and its linearization around
the maximum power point.

Fig. 6 Battery charger equivalent circuit.






0 5 10 15 20 25
0
0.5
1
1.5
2
2.5
3
3.5
Voltage (V)
C
u
r
r
e
n
t

(
A
)


I x V curve
Linearization
The average model of the converter is expressed by:

(1)
Where:
[

)
];

];

] [

] [

].
denotes the average value of and is the duty
cycle.
The purpose of this modeling is to find the converter
transfer function that relates the small-signal input voltage of
the converter (

) with the control variable

. One
important detail is that an increase in the duty cycle reduces
the input voltage of the converter.
According to reference [13], a small-signal model is
obtained considering that each variable of (1) can be
represented by a value in steady state with a small disturbance,
like shown in (2).
{

(2)
Where

, and .
Developing (1) and applying the Laplace transform, it can
be obtained:
( (

( (3)
Where:
(

[ ]

]
In a matrix form:
[

] [

(
] [

] (4)
The transfer function

( relates the capacitor voltage

(5)
The small signal model of (5) is:

(6)
The multiplication

( by (6) results in:

(
(7)
It is used a PI compensator, given by:
(

(8)
Fig. 7 shows the Bode diagram of

(, and the
compensated transfer function (

(. It can be observed
in open loop, a resonance frequency and a small phase margin.


Fig. 7 Bode diagram of open-loop system Gvd (s) and the compensated
G(s)Gvd (s).
D. Boost converter modeling and PBC control
Considering the equivalent circuit presented in Fig. 6 and
neglecting the effect of the capacitor resistance, the Euler-
Lagrange model of the battery charger is given as [14]:

(9)
Where:

[


]

[


]

] and [

]
represents the duty cycle and

and

represent the
inductor current and capacitor voltage, respectively.
The desired values for the average inductor current and
average capacitor voltage are respectively the panel current
and voltage on maximum power point. The vector of averaged
dynamic error is defined by [15]:
( [

(
] [

(
] (10)
As ( (

( with

( [

(]

the desired state, it can be obtained that:

]
(11)
The design of the PBC consists in modifying the system
energy by adding damping through the dissipative structure
[15]. This modification is accomplished through the addition,
in closed loop, of a dissipative term that emulates a resistor
connected in series with the inductor, denoted by

. This
strategy is denominated indirect control, or series control. The
dissipative term added is:

] (12)
-100
-50
0
50
100
M
a
g
n
i
t
u
d
e

(
d
B
)
10
-4
10
-2
10
0
10
2
10
4
-180
-135
-90
-45
0
P
h
a
s
e

(
d
e
g
)


Bode Diagram
Frequency (Hz)
Open loop
Compensated
And the new dissipative structure is given as:

] (13)
Given a desired

, it is possible to verify
the following change in the dynamic averaged error equation
(14):

]
(14)
The energy adjustment of the system is obtained doing:

] (15)
In this circumstance, the error dynamic equation is:

(16)
The desired energy in terms of the error can be modeled
by

(17)

is a Lyapunov function candidate for (16). The time

( ( (18)
Where is strictly positive and constant. The condition
(18) is ensured for (16), and satisfied if:

] (19)
Doing the matrix products of (19), the result is:
{

(20)
The equations (20) contains the expressions of the control
law. To avoid the influence of parasite elements, reference
[16] proposed an integral action, as:

(21)
Equation (21) gives the duty cycle of the converter for the
control of the input voltage. The variables

and

are
parameters of the controller. Where

and

.
E. Simulation
It was simulated in Matlab/Simulink a photovoltaic system
of 48. The solar array consists in a single panel model SM
48KSM whose parameters are shown in TABLE I. The
parameters of the boost converter applied in photovoltaic
system are showed in TABLE II. Finally the parameters of the
batteries used in the simulation are showed in TABLE III.
The algorithm of incremental conductance (MPPT) uses a
sampling frequency of 10 and a step voltage of 1 . It
is necessary calculate the maximum power point current of the
panel for the PBC technique. As this algorithm is only
obtained maximum output voltage of the panel, this value is
obtained using (22).

(22)
Where

is the maximum power point current of the

is the
voltage calculated by the algorithm.
TABLE II. PARAMETERS OF THE CONVERTER.
Boost Converter
Inductor 8.0
Capacitor in Panel

1.5
Capacitor in Battery

15
Output Voltage

36
Frequency Switching

20
Inductor Resistance

0.1
Capacitor Resistance

0.05
Diode Voltage

0.8
IGBT Voltage

1.0
TABLE III. PARAMETERS OF THE BATTERY.
Battery (Nickel-Metal-Hydride)
Nominal Voltage 36
Nominal Capacity 60
Initial Stage of Charge 20 %
Maximum Capacity 64.61
Fully Charged Voltage 42.4
Nominal Discharge Current 12
Internal Resistance 0.006

The Fig. 8 shows the solar radiation profile used to test and
compare the performance of the two control techniques. This
profile consists in ramp variations and will impact in the
maximum power point voltage of the panel.
In order to analyze the impact of parameters uncertain in
the system response, variations of in the inductance
and in the resistance of the inductor are simulated.


Fig. 8 Solar radiation profile.
III. RESULTS
A. Performance during solar radiation variation
The Fig. 9 shows the electrical variables of the solar panel
for both PI and PBC technique. The panel voltage and current
follows the maximum power at all levels of radiation.
0 2 4 6 8 10 12
0
100
200
300
400
500
600
700
800
900
1000
1100
Time (s)
S
o
l
a
r

i
r
r
a
d
i
a
n
c
e

(
W
/
m
2
)
During solar radiation variations, the transient response of
each controller is different. As can be seen in Fig. 10, the
PBC technique follows the maximum power point voltage
with a smaller current overshoot and is faster than PI
technique.
The battery current ripple reduces when the PBC technique
is used, like is shown in Fig. 11. This fact can be justified by
the duty cycle calculated for each technique. In the same
operation point there is a larger oscillation in the duty cycle
calculated by the PI technique. In this case, the PBC control
has the advantages of improving the battery current and a
reduction of the switch stress.

Fig. 9 Electrical parameters of the photovoltaic panel for PI and PBC
techniques.

Fig. 10 Details in the voltage and current response for PI and PBC
techniques.
.
Fig. 11 Battery current and duty cycle of converter for PI and PBC
techniques.
B. Performance during parameter variation
Variations in internal resistance and inductance of the
inductor influence in the dynamic of both control techniques,
like shown in Fig. 12 and Fig. 13. However, in the PI
technique there is an increase in the response time, which
does not happen in the PBC technique.
The behavior of the current in the battery is presented in
Fig. 14 and Fig. 15. Variations in the inductor did not impact
in the current dynamic response. It can be observed that the
inductance has an impact in the current ripple. On the other
hand, the increase of the resistance impacts in the efficiency of
the converter, reducing the average current in the battery.

Fig. 12 Voltage response of the PI and PBC techniques during variations in
the inductance.
IV. CONCLUSION
This work presented a battery charger supplied by a 48
photovoltaic panel connected to a boost converter. The
incremental conductance algorithm was used to find the
maximum power from the panel. It was proposed the PBC
technique and this was compared with the traditional PI
technique.
The PBC control had a faster response in the maximum
power point tracker and a smaller ripple in the duty cycle.
Both components lifetime and charger process are improved
using this technique. Besides, the PBC strategy was more
stable during parameters variation than PI technique.
0 2 4 6 8 10 12
10
15
20
25
Time (s)
V
o
l
t
a
g
e

(
V
)


PI
PBC
0 2 4 6 8 10 12
0
1
2
3
Time (s)
C
u
r
r
e
n
t

(
A
)


PI
PBC
0 2 4 6 8 10 12
0
10
20
30
40
50
Time (s)
P
o
w
e
r

(
W
)


PI
PBC
3.5 4 4.5 5 5.5 6 6.5
1
1.5
2
2.5
3
Time (s)
C
u
r
r
e
n
t

(
A
)


PI
PBC
3.5 4 4.5 5 5.5 6 6.5
15
16
17
18
19
20
Time (s)
V
o
l
t
a
g
e

(
V
)


PI
PBC
0 2 4 6 8 10 12
0
50
100
Time (s)
D
u
t
y

C
y
c
l
e

(
%
)


PI
PBC
0 2 4 6 8 10 12
0
0.5
1
Time (s)
C
u
r
r
e
n
t

i
n

t
h
e

b
a
t
t
e
r
i
e
s

(
A
)


PI
PBC
3.5 4 4.5 5 5.5 6 6.5
16.5
17
17.5
Time (s)
V
o
l
t
a
g
e

(
V
)
PI


L L +20% L -20%
3.5 4 4.5 5 5.5 6 6.5
17.2
17.4
17.6
17.8
Time (s)
V
o
l
t
a
g
e

(
V
)
PBC


L L +20% L -20%


Fig. 13 Voltage response for PI and PBC techniques during variations in the
resistance.

Fig. 14 Battery current response for PI and PBC techniques during
variations in the inductance.

Fig. 15 Battery current response for PI and PBC techniques during
variations in the resistance.


REFERENCES
[1] EUROPEAN PHOTOVOLTAIC INDUSTRY ASSOCIATION.
Global Market Outlook for Photovoltaics 2013-2017. European
Photovoltaic Industry Association. [S.l.], p. 60. 2012.
[2] CASTRO, R. M. G. Energias Renovveis e Produo
Descentralizada. Universidade Tcnica de Lisboa. Lisboa. 2002.
[3] ENSLIN, J. H. R.; WOLF, M. S.; SNYMAN, D. B. Integrated
photovoltaic maximum power point tracking converter. IEEE
Transactions on Industrial Electronics, 44, 1997. 769-773.
[4] IEA. Management of Storage Batteries used in Stand-Alone
Photovoltaic Power Systems - Report_IEA_PVPS_T3-10:2002.
International Energy Agency (IEA). [S.l.]. 2002.
[5] SOUSA, J. M. N. D. Sistema bidirecional de carga de baterias
para o FEUP VEC. Universidade do Porto. Porto, p. 111. 2013.
[6] LINDEN, D.; REDDY, T. B. Handbook of Batteries. 3. ed.
[S.l.]: McGraw- Hill, 2002.
[7] AMBROSIO, R. C.; TICIANELLI, E. A. Baterias de nquel-
hidreto metlico, uma alternativa para as baterias de nquel-
cdmio. Scielo, 2001. ISSN ISSN 0100-4042. Disponivel em:
<http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0100-
40422001000200015&lng=en&nrm=iso>. Acesso em: 28 jan.
2014.
[8] VILLALVA, M. G.; SIQUEIRA, T. G. D.; FILHO, E. R.
Voltage regulation of photovoltaic arrays: small-signal analysis
and control design. IET Transactions on Power Electronics, v. 3,
p. 869-880, 2010.
[9] BECHERIF, M.; AYAD, M. Y.; ABOUBOU, A. Hybridization
of Solar Panel and Batteries for Street Lighting by Passity Based
Control. IEEE International Energy Conference, Al Manamah,
p. 664-669, 2010.
[10] MU, K.; MA, X.; ZHU, D. A New Nonlinear Control Strategy
for Three-Phase Photovoltaic Grid-Connected Inverter.
International Conference on Eletronic & Mechanical
Engineering and Information Technology, Harbin, p. 4611-4614,
2011.
[11] ALMEIDA, P. M. D. Modelagem e Controle de Conversores
Estticos Fonte de Tenso utilizados em Sistemas de Gerao
Fotovoltaicos Conectados Rede Eltrica de Distribuio.
UFJF. Juiz de Fora, p. 190. 2011. (Master thesis).
[12] VILLALVA, M. G.; GAZOLI, J. R.; FILHO, E. R.
Comprehensive Approach to Modeling and Simulation of
Photovoltaic Arrays. IEEE Transactions on Power Electronics,
v. 24, n. 1, p. 1198-1208, March 2009.
[13] ERICKSON, R. W.; MAKSIMOVIC, D. Fundamentals of
Power Eletronics. 2. ed. New York: Klumer Academic
Publishers, 2004.
[14] CUPERTINO, A. F. et al. A Grid Connected Photovoltaic
System with a Maximum Power Point Tracker using Passivity
Based Control applied in a Boost Converter. UFV. Fortaleza, p.
8. 2012.
[15] JELTSEMA, D.; SCHERPEN, J. M. A. Tuning of Passivity-
Preserving Controllers for Switched Mode Power Converters.
IEEE Transactions on Automatic Control, v. 49, p. 1333-1334,
August 2004.
[16] LEYVA, R. et al. Passivity-based integral control of a boost
converter for large-signal stability. IEE Proceedings. Control
Theory and Applications, v. 153, p. 139-146, March 2006.

3.5 4 4.5 5 5.5 6 6.5
16.5
17
17.5
Time (s)
V
o
l
t
a
g
e

(
V
)
PI


RL RL +20% RL -20%
3.5 4 4.5 5 5.5 6 6.5
17.2
17.4
17.6
17.8
Time (s)
V
o
l
t
a
g
e

(
V
)
PBC


RL RL +20% RL -20%
3.5 4 4.5 5 5.5 6 6.5
0.4
0.6
0.8
1
Time (s)
C
u
r
r
e
n
t

(
A
)
PI


L +20% L L -20%
3.5 4 4.5 5 5.5 6 6.5
0.5
0.6
0.7
0.8
0.9
Time (s)
C
u
r
r
e
n
t

(
A
)
PBC


L +20% L L -20%
3.5 4 4.5 5 5.5 6 6.5
0.4
0.6
0.8
1
Time (s)
C
u
r
r
e
n
t

(
A
)
PI


RL -20% RL RL +20%
3.5 4 4.5 5 5.5 6 6.5
0.5
0.6
0.7
0.8
0.9
Time (s)
C
u
r
r
e
n
t

(
A
)
PBC


RL -20% RL RL +20%


BIOGRAPHIES


Valentim Ernandes Neto was born in Aimors,
Brazil. He is student of Electrical Engineering at
Federal University of Viosa (UFV), Viosa, Brazil,
since 2012. Currently is integrant of GESEP, where
develop works about power electronics applied in
renewable energy systems. His research interests
include photovoltaic energy and control applied in
power converters.



Filipe Perez was born in Uberaba-MG, 1990. He is
student of Electrical Engineering at Federal
University of Viosa (UFV), Viosa, Brazil, since
2008. Currently is integrant of GESEP, where
develop works in the area of power systems and
renewable sources, especially solar energy. Currently
working with control converters for photovoltaic
panels. His research interests include power systems,
automation and control.



Andr Gomes Trres received the B.S. degree, the
M.S. and the Ph.D. degree in electrical engineering
from the Federal University of Minas Gerais
(UFMG), Belo Horizonte, Brazil, in 1998, 2000 and
2004 respectively. Since 2005 he has been with the
Department of Electric Engineering, UFV, Brazil. His
research interests include power electronics, electrical
drives and process automation.




































Allan Fagner Cupertino received the B.S. degree in
electrical engineering from the Federal University of
Viosa (UFV), Viosa, Brazil, in 2013. He is
integrant of GESEP, where developed works about
power electronics applied in renewable energy
systems. Currently he is Master student from Federal
University of Minas Gerais (UFMG), Belo Horizonte,
Brazil. His research interests include solar
photovoltaic, wind energy, control applied in power
electronics and grid integration of dispersed generation systems.


Heverton Augusto Pereira received the B.S. degree
in electrical engineering from the Federal University
of Viosa (UFV), Viosa, Brazil, in 2007, the M.S.
degree in electrical engineering from the State
University of Campinas (UNICAMP), Campinas,
Brazil, in 2009. Currently he is Ph.D. student from
the Federal University of Minas Gerais (UFMG),
Belo Horizonte, Brazil. Since 2009 he has been with
the Department of Electric Engineering, UFV, Brazil.
His research interests are wind power, solar energy and power quality.