Effect of Measurement Factors on Photovoltaic Cell Parameters Extracting

International Journal of Electrical and Computer Engineering (IJECE)
Vol. 7, No. 1, February 2017, pp. 50~57
ISSN: 2088-8708, DOI: 10.11591/ijece.v7i1.pp50-57  50
Journal homepage: http://iaesjournal.com/online/index.php/IJECE
Effect of Measurement Factors on Photovoltaic Cell Parameters
Extracting
El Hadi Chahid1
, Mohammed Idali Oumhand2
, M’barek Feddaoui3
, Mohammed Erritali4
,
Abdessamad Malaoui5
1,2,3
Laboratory of Energy Engineering, Materials and systems, National School of Applied Sciences,
Ibn Zohr University, B.P. 1136 Agadir, Morocco
1,5
Laboratoire Interdisciplinaire de Recherche en Sciences et Techniques (LIRST), Faculté Polydisciplinaire,
Université Sultan Moulay Slimane, B.P 592, Béni Mellal, Morocco
4
Laboratoire TIAD, Université Sultan Moulay Slimane, B.P 592, Béni Mellal, Morocco
Article Info ABSTRACT
Article history:
Received Sep 27, 2016
Revised Nov 27, 2016
Accepted Dec 20, 2016
In this paper, we study the influence of external factors on the measurement
for the current–voltage (I-V) characteristic of the photovoltaic cell. These
factors are the size of the number of measurements, the range of the cell
generated voltage and the influence of measures step and mode combination
of photovoltaic cells (parallel, serial, or hybrid). The main extracted
parameters solar cell are the photocurrent Iph, the reverse diode saturation
current I0, the ideality factor of diode n, the series resistance Rs and the shunt
resistance Rsh. A method for finding these parameters, according to the
single-diode model, was developed by Newton-Raphson’s method using
Matlab. To assess the accuracy of this method, measured and calculated I–V
characteristics were compared with provided data by the manufacturer at
standard test condition (STC). The measurement results showed that these
parameters are highly dependent on these four factors.
Keyword:
Electrical parameters
Extracting photovoltaic cell
Measurement factors
Newton-raphson method
Single-diode model
Copyright © 2017 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
El Hadi Chahid,
Laboratory of Energy Engineering, Materials and systems,
National School of Applied Sciences, Ibn Zohr University,
B.P. 1136 Agadir, Morocco.
Email: chahid2016@yahoo.com
1. INTRODUCTION
The photovoltaic solar energy is one of the most used and promising resources in the renewable
energy domain. The conversion of photovoltaic energy is performed by cells based on semiconductors such
as silicon, which exploits the photovoltaic effect. The operation of these cells is described generally by the I-
V characteristic which provides the information on the mechanisms of intern electric transport and the
technological steps imperfections of their fabrications [1], [2]. This characteristic depends on many intrinsic
electrical parameters, that are the photocurrent Iph, the saturation current I0, the ideality factor n, the Rs series
and Rsh shunt resistances [3], [4].
The accuracy determination of these electrical parameters allows to understand and explain certain
physical phenomena like the degradation of the layers that constitute the photovoltaic cells (PVC), the
lifetime of minority charge carriers, the energy efficiency, ... etc. [5], [6]. Normally, several methods have
been proposed for the extraction of these parameters using the single-diode model which is very sufficient to
describe the electrical functioning of the cell [7], [8].
The purpose of this paper is to study the impact of extrinsic factors on the I-V characteristic
measurements. That is, to study the influence of the size of the measurements basis, the distribution and the
ranges of the output voltage of the solar cell and the types of the combination of the different cells that form

IJECE ISSN: 2088-8708 
Effect of Measurement Factors on Photovoltaic Cell Parameters Extracting (El Hadi Chahid)
51
the solar modules. In general, these factors are often neglected by researchers in the field of electrical
modeling and identification of the photovoltaic electrical parameters, especially in the phase of the I-V
electrical characteristic measurements, which is the main source of electrical and physical information related
to PVC [9], [10].
2. ELECTRICAL MODELING
The mathematical model that describes the photovoltaic cell is determined by that of PN junction.
We add to it the photocurrent Iph which is proportional to the irradiation and also a term that modulate
electric intern phenomena. The given current I from the cell is written as follows [11].
s s
ph 0
th sh
V+R I V+R I
I=I -I exp -1 -
nV R
  
   
  
(1)
with:
: Photocurrent current (or a generated current from irradiation).
: Diode saturation current.
: Series resistance.
: Shunt resistance.
n: Ideality Factor of diode.
: Thermal voltage given by , .
: Boltzmann’s constant .
q: Electrical charge of the electron .
T: Temperature of the cell expressed by °K.
Under illumination, a solar cell can be modulated by the electrical circuit with a single diode,
represented in the Figure 1 [12]:
Figure 1. Equivalent Electric Circuit of a PVC
3. IDENTIFICATION OF PVC PARAMETERS
In order to understand the physicals mechanisms produced within the interior of the solar cell; there
are various parameters that specify these phenomena. Many methods have been proposed for the
identification of these parameters, not only for increasing the efficiency of the photovoltaic cell, but also to
be able to simulate its comportment and optimize their different parameters.
In our work, we use the model of a single exponential of the PN junction for the PV cell; formulated
in the equation (1). This model depends on five electrical parameters such as, the photocurrent ( ), the
reverse saturation current of diode ( ), the series resistance ( ), the shunt resistance ( ) and the ideality
factor (n).
The most used methods to extract electrical parameters are:
a. The method of least square [13].
b. Newton’s method [14].
c. The Levenberg Marquardt’s method [15].
The identification method that we use is based on the Newton's algorithm to find the final values of
these parameters. This algorithm does not normally converge when the initial parameters (Rs, Rsh, Iph, I0
and n) are not well chosen. Indeed, to estimate these parameters, we used the bouzid’s method [16] at the
beginning of this algorithm.
V
Rsh
Rs
I
I
ph
Id

 ISSN: 2088-8708
IJECE Vol. 7, No. 1, February 2017 : 50 – 57
52
Newton's method use an algorithm which search for the approximations of the roots of function
(x)=0. This method develope the I-V model using Taylor’s formula in the order 1 around an optimal point
f(V,I) = 0 as the following expression:
s s
ph 0
th sh
V+R I V+R I
f(I,V)=I-I +I exp -1 +
nV R
  
   
  
(2)
For a given voltage V the current I is given by the root of equation (2), where “I” is expected to be
the single root. In order to determine it, we start with the initial value Ii, the following iteration is executed:
'
i+1 i i iI =I -f(I )/f (I ) (3)
The iteration stops if a suitable condition is met. In the case above the absolute difference of two
succeeding values of the iteration is used :
i+1 iI -I <ε (4)
Using the fitting technique of an experimental curve with a theoretical curve modulating the
equation (1), we are able to determine intrinsic parameters of a solar cell. This technique focuses on
minimizing the sum of the squares of the errors used in the objective function , defined as:
      
N N
exp 2
i i i i
i=1 i=1
χ φ = I -I V ,φ = R φ  (5)
a. φ is a number of unknown parameters: ph 0 s shφ(I ,I ,n,R ,R )
b. Ri is a Residual error between experimental values exp
iI and calculated values i iI (V ,φ) .
c. N is a Number of the measurement points.
d. Ii and Vi are respectively the measured current and voltage at the Ni
point of measured data.
We have used Newton’s algorithm for the minimization of the objective function χ . The five
optimal values of the parameters φ are evaluated by Newton’s iterations:
   
-1
j j-1 j-1 j-1φ =φ - J φ f φ 
  (6)
 J φ : Jacobian matrix and their elements are defined as the derivatives of the current concerning the five
parameters mentioned before
iφ=φ
I
φ
 
 
 
4. MEASURES AND USED TECHNIQUES
4.1. Experimental assembly
The measuring device comprises an electronic card based on a microcontroller (Arduino μC), a PC
computer and an interface electronic card connected to the PV module. The computer is equipped with the
LabView software for acquisition of current I and voltage V of the module through the Arduino analog inputs
A0 and A1. Electromechanical relays are controlled by the μC, and allow changing PV load through binary /
decimal decoders.
Measurement voltage V of PVC is performed by a voltage divider created by resistors, and
operational amplifiers which provide impedance matching. In general, the maximum voltage at the output of
the PV module is about 20 volts, so we use a ratio of ¼ to the voltage divider to get the maximum 5 volts on
the analog input of μC.
To increase the accuracy of the measurement of the current, an amplifier gain of 2 is adjusted and
used at the output of this sensor. Electrical measurements are taken using an automatic solar tracker, which

IJECE ISSN: 2088-8708 
53
we developed to ensure a relatively constant light intensity for these measures [16]. Figure 2 is shows used
experimental assembly.
Figure 2. Used Experimental Assembly
4.2. Measurement technique
The test experiments realized are performed according to an assembly of measures achieved from
solar panel constituted from silicon polycrystalline “Solarex MSX60”. This module is composed of 36 solar
cells which are combined parallel at the Standard Test Condition STC of the irradiation and the temperature.
Our research group has already worked on optimizing these measures by improving methods and
instruments of the optimal measuring I-V of a solar panel [17].
This paper highlights the effects of the impact of the four major factors on the accuracy of the
intrinsic parameters of the electric model (Rs, Rsh, Iph, I0 and ideality factor n). These factors are:
1. The size of the measured data base on the I-V characteristic, which means the number (N) of
measurement points for measuring the I–V characteristics of solar cells. The number (N) of the
measurements realized during the experience of I-V.
2. The nature and the values of the step of the measures (Ij, Vj) in a characteristic.
3. The distribution of the measures on the different zones of the voltage V of the characteristic. Three zones
are precised; the zone 1, for low voltages, the zone 2 for medium voltages and the zone 3 is related to the
high voltages.
4. The association of the connection of the PV cells in the interior of a PV module.
In our case, we have studied the effect of two types association of the cells. In the first side, the
series association of two PVC is examined, followed by a test on another parallel association of the cells. The
validation and the interpretation of the results discovered are made through a comparing between these
values with a reference [17]. Taking into account that during the steps of these measures, we have respected
the acquisition technique and the filtering without noise which is developed by our research group [10], [18].
After determining the optimal factors that give a better measure, a deep study is currently underway by our
research group to determine the relation between of the minorities carriers lifetime of electrons and holes
and respectively and intrinsic electrical parameters.
5. RESULTS AND DISCUSSION
Before studying the impact of these factors on the reliability and the accuracy of the PVC electrical
parameters, the device and the developed program were tested and their results were compared with a
mathematic model (equation 1). The simulation and measured results of I-V characteristics are compared and
shown in Figure 3.
From the first sight, this curve provides an enough confidence on the tools of taken measures during
this work. It is noted that, there is a correlation between the measured data and the electrical model of the
single diode, within the whole range of the applied voltage which is changed from 0V to Voc =21.1V in 40
measures data base. Indeed, to have more details about of the influence of measured numbers (N) and divided
step () of these measures, tests are carried on the same photovoltaic cell by varying these two parameters.

 ISSN: 2088-8708
54
Figure 3. Experimental and Calculate of PVC Characteristics
5.1. Effect of the Measurements Number (N)
5.1.1. Measures with a Constant Step
A constant step of taken measures (Ij, Vj) are chosen in this level. The measurements were carried
from the whole range of voltages [0V- Voc], through a step for each size of the measured data base.
Eelectrical parameters are calculated by the developed program [16], and the results are shown in the
Figure 4. The calculated error R as a function of the number of the measures (N) is presented in the Table 1.
Table 1. Measurement Error with Number (N)
N 20 40 60 90 110 150 160 180 200 220
R( 0.67 0.06 8.7 2.2 0.63 2.8 11.5 10.4 3.52 2.0
Figure 4. Variation of PVC Electrical Parameters with N
0 2 4 6 8 10 12 14 16 18 20 22
0,0
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
Current(A)
Voltage (V)
Experimental values
Calculated values
Zone 1 Zone 2 Zone 3

IJECE ISSN: 2088-8708 
55
This table shows a minimum value of the error R which fits with a number N = 40 of measurements.
The Figure 3 presents a minimum at N = 40 concerning the three parameters Rsh, I0 and n, whereas the other
parameters Rsh and Iph, show a maximum at N = 40.
Meanwhile, this result guides us to minimize the number of I-V measurements, in order to save time
and energy while measuring. Furthermore, it allows us to save the memory space for digital devices of
acquisition. Indeed, in contrary to what we believe that a big number of measurements are frequently
required. A reduced number of measurements is largely sufficient and it is sometimes optimal in order to
discover specific results of the electrical parameters. This is explained by the reduction of the fluctuations
and the minimization of the errors produced through a large data base of measurements. On the other hand,
concerning the values (N) which are greater than 40, the accuracy of PVC electrical parameters are not linear.
So, instead of using larger sizes of the measures of the I-V characteristic, we take a minimum number of
measurement (N = 40) as in our case.
5.1.2. Measures with a variable step
The measuring of the I-V characteristic at this level was examined by taking a variable step ().
Indeed, the three zones have been fixed:
1 = 0.4 V: Low voltages domain, as (0 <V< 5/8 Voc),
2 = 0.01 V: Medium voltages domain, as (5/8 Voc <V< 7/8 Voc),
3 = 0.2 V: High voltages domain, as (7/8 Voc < V < Voc).
The electrical parameters extracting (Rs, Rsh, n, Iph and I0) are calculated and presented in Table 2.
Table 2. Comparison of the Intrinsic Parameters with the Variable and Constant Step
Electrical parameters Rs(Ω) Rsh(Ω) Iph(A) I0(A) n
Variable step () 0.179 360.123 3.7999 1.85 x 10-7
1.325
Constant step () =0.5V 0.180 360.020 3.8000 1.85 x 10-7
1.325
Reference [18] 0.180 360.002 3.8119 1.85 x 10-7
1.360
From these values, it is found that the variation of the step () of the voltage does not have too much
effect on the electrical parameters of the cell. Therefore, it is recommended to work with a regular step to
optimize the electrical measurement chain. A variable step () requires additional time to the processor
which rotates the calculation algorithm. Indeed, the program is slower and requires more storage space for
the iterations. Therefore, the calculation is more complicated and makes the possibility of divergence of the
program.
5.2. Effect of the bias voltage ranges:
We took the best measures which correspond to the 40 measures in Table 3.
Table 3. Intrinsic Electrical Parameters Depending on the Variable Voltage Range
Voltage Range (V) Rs(Ω) Rsh(Ω) Iph(A) I0(A) n
Low [0-5/8Voc] 54.681 304.732 4.4796 1.04 x 10-108
0.966
Averages [5/8Voc-7/8Voc] 1.895 27.681 4.4964 1.02 x 10-23
0.504
High [7/8Voc -Voc] 0.330 0.534 40.1492 2.33 x 10-52
0.193
Reference [18] 0.180 360.002 3.8119 1.85 x 10-7
1.360
At a low voltage domain, the obtained electrical parameters show that the optimal parameters are
those of Rsh, Iph and n according to the output voltage ranges of the solar cell module. At a medium voltage
domain, the optimal parameter is that the saturation current I0. At high values of voltage, the optimal value is
that of the series resistance Rs.
5.3. Effect of PVC association:
The used photovoltaic cell in this work is that of silicon [17]. The effect of the mounting type of
these solar cells was studied in the case of a series connection, and then in parallel. The results are shown in
Table 4.

 ISSN: 2088-8708
56
Table 4. Intrinsic Electrical Parameters Depending on the Type of Association
solar cell mounting Rs(Ω) Rsh(Ω) Iph(A) I0(A) n
One cell 0.106 872.398 0.1023 6.30 x 10-8
1.416
Two series of cells 0.213 1744.843 0.1023 6.30 x 10-8
2.833
Two paralleled cells 0.038 503.329 0.2045 1.96 x 10-7
1.460
Reference [18] 0.180 360.002 3.8119 1.85 x 10-7
1.360
Based on the extracted values from the Newton’s method, we found that in the case of a series
connection, the electrical parameters Rs, Rsh and n having values that are doubled, contrary to the
photocurrent Iph and the saturation current I0 remain unchanged. In the case of a parallel association, the
parameter Iph is doubled, contrary to the ideality factor n that remains constant and the saturation current I0
are significantly increased, whereas the resistances Rs and Rsh are decreased in a low manner.
6. CONCLUSION
In this work, we have studied the sensitivity of the intrinsic electricl parameters of a photovoltaic
solar cell (the photocurrent Iph, the saturation current I0, the ideality factor n, the series resistance Rs and the
shunt resistance Rsh) according, first, to the number of measures (N), second, to the step of the I-V
characteristics voltages, third, to the range of voltages in this characteristic and the type of the different cells
connection. So, the obtained electrical parameters are relatively depending on the previous factors. Based on
what has preceded, we deduce that the minimum number of measurements (N=40 in our case) is largely
sufficient and it is recommended to work with a regular step to optimize the accuracy of the extracted
parameters. Relatively to the range of voltages, we must choose the low output voltages in order to achieve
an optimal measure of the following parameters Rsh, Iph and n. However, the optimal value of the extracted
series resistance Rs must be made in high bias voltages; while the extracted saturation current I0 should be
performed in the average voltages. Concerning the influence of different cells connection mode, we observe
that the values of electrical parameters Rs, Rsh and n, are approximately doubled in the PVC series
combination; whereas, the photocurrent Iph also doubled in a parallel association in this study. These last
results are in accordance with the theoretical results, while the other results are unexpected and can be
exploited to increase the performance of the PVC and their installations.
REFERENCES
[1] K. Yamaguchi and H. Kodera, “Optimum Design of Triode-Like JFET’s by Two-Dimensional Computer
Simulation,” IEEE Transaction on Electron Devices, vol/issue: ED-24(8), pp. 1069, 1977.
[2] S. M. Sze, “Physics of Semiconductor Devices,” John Wiley, New York, 1981.
[3] A. Schenk and U. Krumbein, “Coupled Defect-Level Recombination: Theory and Application to Anamalous Diode
Characteristics,” Journal of Applied Physics, vol/issue: 78(5), pp. 3185-3192, 1995.
[4] S. Yadir, et al., “Extraction of Solar Cell Physical Parameters Model with Double Exponential from Illuminated I-
V Experimental Curve,” Global Journal of Physical Chemistry, pp 236-240, 2011.
[5] N. F. Mott, “Metal-Insultor Transitions,” Second Edition, Taylor & Francis, London, 1990.
[6] J. S. Kumari and C. S. Babu, “Mathematical Modeling and Simulation of Photovoltaic Cell using MatlabSimulink
Environment,” International Journal of Electrical and Computer Engineering, vol/issue: 2(1), pp. 26-34, 2012.
[7] A. P. Alivisatos, “Semiconductor Clusters, Nanocrystals, and Quantum Dots,” Science, vol. 271, pp. 933-937,
1996.
[8] O. Breitenstein, et al., “Material-Induced Shunts In Multicrystalline silicon solar Cells,” Semiconductors, vol/issue:
41(4), pp. 440-443, 2007.
[9] A. Malaoui, et al., “implementation a new approach for modeling and determining the electrical parameters of solar
cells,” International Journal of Innovation and Applied Studies, vol/issue: 15(2), pp. 329–338, 2016.
[10] A. Malaoui, “New method for improving the quality of electrical measurements: application to the extraction of the
intrinsic parameters of the photovoltaic cells,” International Journal of Innovation and Applied Studies, vol/issue:
15(2), pp. 375-386, 2016.
[11] M. Louzazni, et al., “Modeling and Simulation of A Solar Power Source for a Clean Energy without Pollution,”
International Journal of Electrical and Computer Engineering, vol. 3, pp. 568-576, 2013.
[12] T. Ikegami, et al., “Estimation of equivalent circuit parameters of PV module and its application to optimal
operation of PV system,” Solar Energy Materials & Solar Cells, vol. 67, pp. 389-395, 2001.
[13] W. H. Press, et al., “Numerical Recipes in C: The Art of Scientific Computing,” 2nd Ed. Cambridge University
Press, vol. 9, pp. 362 – 367, 2002.
[14] T. Easwarakhanthan, et al., “Nonlinear Minimization Algorithm for Determining the Solar Cell Parameter with
Microcomputers,” International Journal of Sustainable Energy, vol. 4, pp. 1-12, 1986.

IJECE ISSN: 2088-8708 
57
[15] D. Marquardt, “An Algorithm for Least Squares Estimation of Non Linear Parameters,” SIAM Journal Of Applied
Math, vol. 11, pp. 431-441, 1963.
[16] K. Bouzidi, et al., “Solar cell parameters evaluation considering the series and shunt resistance,” Solar Energy
Materials & Solar Cells, vol. 9, pp. 1647-1651, 2007.
[17] D. Bonkoungou, et al., “Modelling and Simulation of photovoltaic module considering single-diode equivalent
circuit model in MATLAB,” IJETAE International Journal of Emerging Technology and Advanced Engineering,
vol/issue: 3(3), 2013.
[18] A. Malaoui, “Implementation and tests of an automatic system to improve electrical energy in photovoltaic
installations,” International Journal of Innovation and Applied Studies, vol. 8, pp. 328–340, 2014.
BIOGRAPHIES OF AUTHORS
E. Chahid was born in El Jadida, Morocco in 1973. Currently is an Assistant Professor at
Regional Center for Careers of Education and Training in Beni Mellal. His area of interest is
Photovoltaic system and Electronics Engineering.
E-mail : chahid2016@yahoo.com
M. Idali Oumhand was born in Casablanca, Morocco in 1966. Currently is an Assistant
Professor at Regional Center for Careers of Education and Training in Inzegan. His area of interest
is Photovoltaic system and Electronics Engineering. Member at LGEMS Laboratory of Energy
Engineering, Materials and systems, National School of Applied Sciences ENSA, Ibn Zohr
University, Agadir, Morocco.
E-mail : idali4@yahoo.fr
M. Feddaoui is a professor of mechanical engineering at National School of applied Sciences, Ibn
Zohr University. He is head of Energy Engineering, Materials and Systems. He received his
professor (HDR) degree from Ibn Zohr University in 2011. He has published more than 30
research papers in international journals and more than 50 papers in edited proceedings and actes.
He has been working in several fields, such as multiphase flow, phase change, heat and mass
transfer, nanofluids.
E-mail : m.feddaoui@uiz.ac.ma
M. Erritali Obtained a Phd in 2013 at the faculty of sciences, Mohamed V Agdal University
Rabat. His current interests include developing specification and design techniques for use within
Intelligent Network, data mining, image processing, cryptography and Electrical systems. He is
currently a professor at the Faculty of Science and Technology, University Sultan Moulay
Slimane.
E-mail : mederritali@yahoo.fr
Dr. Abdessamad Malaoui is a Prof. at Polydisciplinary Faculty (Sultan Moulay Slimane
University of Béni Mellal). He obtained his PhD. doctorat in electronics & computer sciences,
from University of Provence - France in 2005. He obtained his Habilitation thesis in electronics,
computer sciences in 2011. His research interests are in electronics, Industrial computer, µP & µC,
renewable energy, PV system performance monitoring.
E-mail : a.malaoui@usms.ma

Effect of Measurement Factors on Photovoltaic Cell Parameters Extracting

More Related Content

What's hot (16)

Similar to Effect of Measurement Factors on Photovoltaic Cell Parameters Extracting (20)

More from Yayah Zakaria (20)

Recently uploaded (20)

Effect of Measurement Factors on Photovoltaic Cell Parameters Extracting