615

Chapter 20
An Accurate and Efficient
Analytical Method to Extract
the Parameters of the
Single and Double Diode
Photovoltaic Cells Models
Radouane Majdoul Mohamed Aboulfatah
Hassan 1st University, Morocco Hassan 1st University, Morocco

Elhassane Abdelmounim Abd Elwahed Touati

Hassan 1st University, Morocco Hassan 1st University, Morocco
Ahmed Moutabir
Hassan 1st University, Morocco

ABSTRACT
For Photovoltaic systems designers and manufacturers, it is very important to develop suitable models
to closely emulate the characteristics of PV cells, predict their behavior and evaluate their efficiency. So
the main contribution of this chapter is to propose an improved and accurate method for identifying and
determining the equivalent circuit elements values of photovoltaic module using only exact analytical
equations and four manufacture’s data reference, i.e., the open-circuit voltage (VOC), the short-circuit
current (ISC), the current and the voltage at the maximum power point (IM, VM). In order to extract the
five-parameter Single or Double-Diode models of photovoltaic module, the authors try initially to de-
termine analytically all parameters according to RS (the value of the series resistance). Thus, all these
parameters are calculated once RS is determined. Rapid and iterative algorithm is then designed to solve
a strongly nonlinear equation in order to extract the value of RS in a precise manner and without any
mathematical simplification used usually by many other authors.

DOI: 10.4018/978-1-5225-0773-4.ch020

Copyright © 2017, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
 
An Accurate and Efficient Analytical Method to Extract the Parameters

INTRODUCTION

The most powerful natural energy resource is the sun and the solar technologies offer clean and sustainable
options for generating electrical energy without pollution. Photovoltaic (PV) technologies have distinct
environmental advantages for generating electricity over conventional technologies. The operation of
photovoltaic systems does not produce any noise, toxic-gas emissions, or greenhouse gases. Photovoltaic
electricity generation, regardless of which technology is used, is a zero-emissions process.
The basic unit for converting solar energy into useful electrical energy is the solar cell. Grouped cells
form photovoltaic (PV) modules with the aim of increasing energy production and make the process more
practical. However, due to the high investment cost on PV modules, optimal utilization of the available
solar energy has to be ensured. This necessitates a precise and reliable simulation of the designed PV
systems prior to installation.
For Photovoltaic systems designers and manufacturers, it is very important to develop suitable
models to closely emulate the characteristics of PV cells, predicting their behavior and evaluating their
efficiency (Majdoul et al., 2015). It can be used also to study the interaction between the power con-
verter and the PV arrays. Climate and solar radiation affect both on PV system supply side issues and
on system demand side issues. Designers need both solar data and temperature data. The modeling tool
must allow the analysis of the behavior of electrical characteristics in accordance with environmental
changes such as temperature and irradiance. It is verify that these extrinsic factors influence strongly
the photovoltaic efficiency.
In practice, in order to describe the current-voltage (I-V) relationship for PV simulators, the most
popular approach is to use the electrical equivalent circuit with both linear and non-linear components
(Ishaque et al., 2011). According to what has been said, it is clear that these components have to be
adjusted automatically when the operating conditions change. Over the years, many models have been
proposed, but two main equivalent circuit models have been widely used: the single diode model also
called simple exponential model and the double diode model or double exponential model. These models
differ in the accuracy and number of parameters involved in the calculation of PV current-voltage char-
acteristics. To use these models in the simulation and evaluation of PV systems, one needs to determine
the models parameters. However, parameter identification of such models is a challenging problem,
since the derived equations for the estimation of a PV model parameters are implicit and nonlinear and
may not be analytically solved (Hejri et al., 2014).
This chapter aims to give at first a summary overview of many aspects of photovoltaic cells model-
ling used and the different parameters estimation methods explained and promoted by many researcher
authors. In a second phase, the authors proposes and develops an improved and accurate method for
identifying and determining the equivalent circuit elements values of photovoltaic module. They use
in this approach, only exact analytical equations and four manufacture’s data reference, i.e., the data of
three remarkable points: the open-circuit voltage (VOC), the short-circuit current (ISC) and the current
and the voltage at the maximum power point (IM, VM). This approach is presented in a way that it can be
easily accessed by the expert and the non-specialist of photovoltaic systems.
This promising alternative would be to compute the parameters PV cells model using the hybrid
approach by combining analytical method and numeric computing. Thus, the authors try initially to
determine analytically all parameters according to RS (the value of the series resistance). Thus, all these
parameters are calculated once the series resistance RS is determined. Then a rapid and iterative algorithm
is designed to solve a strongly nonlinear equation in order to extract the value of RS in a precise manner

616

An Accurate and Efficient Analytical Method to Extract the Parameters

and without any mathematical simplification used usually by other authors. Finally, this accurate and
combined method is used in order to extract the five-parameter Single-Diode and Double-Diode models
of photovoltaic modules without any mathematical simplifications or other physical assumptions used
by many scientist papers authors. The accuracy of this combined approach and the efficiency of the
two models are evaluated and validated by package Matlab/Simulink environment, through comparison
of simulation results of the model to the data provided by product’s manufacturer. This chapter has the
merit to present in detail the calculation steps and the analytical equations used to extract the PV model
parameters at standard test condition (STC) or at any different operating conditions. This modeling can
be used to implement or compare different algorithms of maximum power point tracking in order to
achieve the best MPPT technique. It can also be used to determine the correct and precise estimation
of the extracted power based on a real profile of environmental conditions of a specific location. The
simulator based on the equivalent electrical circuit of PV cells model, also allows implementing control
laws and regulation techniques in order to connect efficiently the photovoltaic systems to the grid.
The remainder of this chapter is organized as follows. Section 2 deliberates the state of art of the
parameter estimation methods and discusses the literature of the important works on the modelling and
parameters extract of photovoltaic (PV) cells for PV simulation. It provides the concepts, the advantages
and drawbacks of the most popular methods used recently. Section 3 presents the equivalent electrical
circuit-based models widely used to simulate PV cells. In the section 3, a brief review of some artificial
intelligence methods of parameters estimation is presented. In the section 4, one presents the proposed
hybrid method in detail, followed by the experimental results and analysis in section Solutions and
Recommendations. This section illustrates the results by applying the proposed method to different PV
modules, and proves the validity of the suggested parameter identification scheme in comparison with the
experimental data. In the end of the chapter, a conclusion and summary of the future trend and direction
of research that is relevant to this topic is given.

BACKGROUND

The number of scholarly publications related to PV modelling and parameters estimation methods is
very high. Many researches are dispersed in numerous journals, conference proceedings, standards and
other classified/non-classified documents (Chin et al., 2015). Generally, in the literature, there are two
types of approaches for the purpose of parameter extraction of solar cell models: analytical approaches
and numerical approaches. Since the parameter extraction of solar cell models is a non-linear, multi-
variable, and multi-modal problem with many local optima, the use of pure analytical techniques may
be difficult to extract the parameters without many mathematical simplifications or other physical as-
sumptions used by several author’s papers.
The authors Ishaque et al (2011a), Cubas et al (2014), and Hejri et al (2014) propose analytical methods
but with several mathematical simplifications, making it easy to solve equations. Many authors, Villalva
et al (2004) and others, use equations or formulas based on physical assumptions for a fast parameter
extraction. Easwarakhanthan et al (1986), Chegaar et al (2004) and Ortiz-Conde et al (2000) discuss in
their research an alternative approach to obtain the model parameters extracting them by means of I–V
curve fitting using non-linear optimization techniques. They conclude that the accuracy of the extracted
parameters depends on the type of algorithm, the objective function and the initial conditions. It is rec-
ognized that, inappropriate choice of initial parameters values may lead to non-convergence. Further-

617
 
An Accurate and Efficient Analytical Method to Extract the Parameters

more, as the number of parameters increases, the optimizer loses its ability to produce accurate results.
Therefore, in the later numerical approaches, the use of artificial intelligence techniques for solar cell
parameter extraction has received considerable attention. Several research projects were exploring the
use of Genetic Algorithms GA (Jervase et al., 2001; Sellami et al., 2011; Moldovan et al., 2009). Other
works focus on Particle Swarm Optimization PSO (Huang et al., 2011; Hengsi et al., 2011; Ye et al.,
2009) and Artificial Bee Swarm Optimization ABSO (Askarzadeh et al., 2013). Also be distinguished
Differential Evolution DE (Ishaque et al., 2011b; Da Costa et al., 2010), Simulated Annealing SA (El-
Naggar et al., 2012), Bacteria Foraging Algorithm BFA (Rajasekar et al., 2013), Harmony Search SA
(Askarzadeh et al., 2012), Pattern Search optimization PS (Alhajri et al., 2012) and so on. It is noteworthy
that many of these algorithms remain inaccessible and difficult to use by photovoltaic professionals.
The artificial intelligence algorithms are particularly attractive due to their global research capability,
and their effectiveness in handling nonlinear functions without gradient information. Despite these ad-
vantages, soft computing is not employed in PV simulators due to its stochastic nature, which could not
guarantee consistency in computational and convergence speed. To draw a comparison between several
different methods of estimation and calculation, the authors Chin et al., (2015) discuss and suggest in
their review paper important works in the modeling and parameter estimation of PV cells. In the critical
evaluation, Chin et al., (2015) summarized the modelling methods based on several relevant criteria.
To overcome various drawbacks, the chapter presents an hybrid method that combines an analytical
approach and another numerical. This method is based mainly on a large of mathematical manipula-
tions, allowing to calculate or to express the majority of parameters in a purely analytical manner. The
parameters photocurrent IPH, the saturation currents IS1, IS2 and the shunt resistance RSH are expressed
according to RS the series resistance. This analytical method is no longer appropriate to solve a strongly
nonlinear equation to extract the value of RS. Therefore, a fast computing algorithm is designed for the
final step calculating. This combined method is particularly interesting because it is based only on ana-
lytical expressions without mathematical simplifications, and the soft algorithm does not require initial
values and converges very quickly.

MAIN FOCUS OF THE CHAPTER

Equivalent Electrical Circuit of PV Cells Models

The equivalent electrical circuit-based models are most widely used to simulate PV cells. The main
advantage of using the electrical circuit model is the availability of the standard electrical software such
as the package MATLAB/Simulink and PSpice, where the PV model can be seamlessly integrated into
a larger PV systems comprising of power static converter, grid connectivity, etc. In practice, a photo-
voltaic module presents nonlinear I-V characteristic curves that depend mainly on the cell irradiance
and temperature. Based on the physical interpretation of a PV cell and its arrangements to form a PV
module, two equivalent electrical models can be built: the single diode and the double diode PV models.

Description of the Single Diode PV Model

Different models based on the I-V characteristic curve of a PN junction are used to describe the behavior
of PV cells. Among various models, the simplest is the single diode PV model. This model is the most

618

An Accurate and Efficient Analytical Method to Extract the Parameters

widely used in the technical literature, since it offers a reasonable relationship between simplicity and
accuracy. Based on this model, PV module equivalent circuit, shown in figure 1, consists of four compo-
nents. One distinguishes a photogenerated current source IPH associated to the generation of electron-hole
pairs, a parallel diode D representing recombination current that accounts for diffusion of electrons and
holes across the junction and two resistor components. The losses due to the contact resistance between
the silicon and electrodes surfaces, the current flow resistance in the silicon material and the resistance
of the electrodes need to be considered (Villalva et al., 2004). One way to represent these losses is by
incorporating a series resistance RS to the ideal model. The presence of the shunt resistance RSH cor-
responds to the leakage current in the PN junction. Its value varies with different fabrication methods.
It should be noted that the PV modules are manufactured from the series and parallels connection
of PV cells to achieve high current and voltage levels. Moreover, some bypass diodes are installed in
PV modules to avoid avalanche breakdown and hot spots during partial shading. The characteristics of
the connected cells inside a module are nearly identical (Ishaque et al., 2011c). As a result, a PV panel
model is considered as a single cell with some multipliers factors that are incorporated in the cell model
dependent upon the number of series and parallels connected cells in the module. According to the
equivalent circuit given in Figure 1, the mathematical equation that describes the behavior of the solar
cell in a specified illumination and temperature is given in relationship (1).
The single diode PV model shown in Figure 1 can be represented mathematically by the following
nonlinear transcendental I-V equation:

 V + R I   V + R I
I = I PH − I S exp  S 
 − 1 −
S
(1)
  a.N S .VT   RSH

where

I PH : Photogenerated current (in A).

I S : Saturation current due to diffusion mechanism (in A);
RS : The series resistor (in Ω);

Figure 1. Electric equivalent circuit for a single diode PV model with four components

619
 
An Accurate and Efficient Analytical Method to Extract the Parameters

RSH : The shunt resistor (in Ω);

a : Diode ideality factor for diffusion current;
VT : Thermal voltage defined as VT =k*T/q (in V)
k : Boltzmann’s constant, (k = 1.38 e-23 J/K);
q : The elementary charge of electron, (q = 1.6 e-19 C);
T: PN junction temperature in Kelvin.
I denotes the output current generate by the PV module, V the output voltage.

Considering the equivalent electrical circuit for the single diode PV model, four parameters to be
extracted are IPH, IS, RS and RSH. In the relationship (1), is also distinguished a fifth unknown parameter:
a non-dimensional constant noted " a " is added to the term of the recombination current in the p- and
n-sides. This constant is called the ideality or quality factor (or sometimes emission coefficient), and it
takes into account the deviation of the diodes from the Shockley diffusion theory (the value of this fac-
tor, a, is assumed to be constant and between 1 and 1.5 for one-junction cells. In this work, one takes
the value of the ideality factor equal to 1.3 ( a = 1.3), it was chosen to be the same value selected by
many paper authors (Villalva et al., 2004; Cubas et al., 2014). In the equation (1), the unknown param-
eters are mutually coupled. For this reason, it is difficult to determine those parameters using simple
analytical method. One notes that only four parameters IPH, IS, RS and RSH are to extract in this case.

Description of the Double-Diode PV Model

The single diode models discussed thus far inherently neglect the effect of recombination current loss
in the depletion region. Consideration of this loss, particularly at low voltage, leads to a more precise
solution, known as the double diode model. The electric equivalent circuit based on this model is shown
in Figure 2.
The expression of its current–voltage relationship is given as following:

 V + R I    
I = I PH − I S 1 exp  S 
− 1 − I exp V + RS I  − 1 −V + RS I (2)
   a .N .V  
  a1.N S .VT   S 2   2 S T  RSH

Figure 2. Electrical equivalent circuit for a double diode PV model with five components

620

An Accurate and Efficient Analytical Method to Extract the Parameters

Clearly, the trade-off is the need to consider two new parameters, i.e. the saturation current IS2 and
the ideality factor for the second diode " a2 " . This is in addition to the existing five parameters, namely
IPH, IS1, a1 , RS, RSH, thus increasing the number of unknown parameters to seven. Furthermore, the pres-
ence of the two exponential terms poses a considerable challenge to solve and extract seven parameters
analytically. Nevertheless, the double-diode model remains an attractive option due to its superior ac-
curacy at low irradiance conditions (Chin et al., 2015).
To reduce the number of unknown parameters to five, the values of the ideality factors a1 and a2
are commonly approximated respectively as 1 and 2 with a reasonable accuracy. This assumption is
widely used in the literature especially it has been used by Hejri et al., (2014) in their paper; this makes
it possible the analytical resolution of the equations.

Brief Review of Parameters Estimation Methods

using Artificial Intelligence Techniques

Genetic Algorithm (GA)

Inspired by the process that drives biological evolution, the genetic algorithm at each step employs a
random selection of individuals, known as parents or chromosomes, from a current population of solu-
tions (Hernanz et al., 2010, Kumari et al., 2012, Suthar et al., 2013) . It mimics the process of natural
selection i.e. survival of the fittest. The parental solutions produce the children for the next step of the
algorithm by introducing changes or mutations into some of the current solutions, or by combining two
current solutions to form newly improved ones (crossover). The underlying idea is that over successive
generations, the initial population evolves toward an optimal solution.
Ismail et al., (2013) applies Genetic Algorithm to find the global optimum parameter values of a PV
module, i.e. values which result in the least average absolute error over a sampled range of G and T. Once
these values are determined, they are assumed as constants to predict the output of the PV modules at
different operating conditions. Jervase et al., (2001) employed GA to extract the parameters of double
diode model from a synthetic I–V curve. Moldovan et al., (2009) and Zagrouba et al., (2010) used this
type of algorithms to identify the parameters of the single diode PV model from the measured I–V
curves of several PV cells and modules. Although the extracted parameters were considerably accurate,
the extraction process was admittedly slow.

Particle Swarm Optimization (PSO)

The algorithm is inspired by the swarm intelligence of bird flocking or fish schooling (Kennedy et al.,
2010). It is initialized with a population of random solutions, called particles, within a predefined solu-
tion search space. At each iteration, the position and velocity of each particle are adjusted in light of
the personal best experience of the particle and that of the entire swarm. The weights of the two factors
(i.e. position and velocity) are determined by a system of randomized correction coefficients (learning
factors) (Chin et al., 2015).
In this context, Ye et al., (2009), Macabebe et al., (2011), Sandrolini et al., (2011) and other authors
applied PSO to extract the cell parameters of the single and two-diode models from the I–V curves. In

621
 
An Accurate and Efficient Analytical Method to Extract the Parameters

comparison to Genetic Algorithm, PSO was found to be more accurate with better computational speed.
The extracted parameters were studied to identify the different mechanisms affecting PV performance.

Differential Evolution (DE)

Differential Evolution is a variant of Genetic Algorithm, which is introduced by Storn et al., (1997). Unlike
Genetic Algorithm, which relies on crossover operation, DE utilizes mainly on the mutation operation
to explore prospective regions in the search space. The algorithm starts with a population of random
candidate solutions, known as target vectors. For each target vector, a mutant vector is produced through
mutation, by adding the weighted difference between two random vectors to a third vector; the procedure
varies depending on the chosen strategy. In the crossover operation, a population of trial vectors is created
by mixing the elements from the mutant vectors and target vectors. Finally, a one-to-one tournament is
employed in the selection stage. If a trial vector is found to yield better fitness value it will replace the
existing target vector. Otherwise, the original target vector carries forward to the next generation. The
iteration continues until either a satisfactory fitness value has been achieved or a maximum number of
generations has been produced. In the PV context, Da Costa et al., (2010) employed DE to identify the
modules parameters at different operating conditions based on digitized datasheet I–V curves. Kashif
et al., (2011) evaluated the speed and accuracy of DE in extracting parameters of the double PV diode
model for different types of modules. Two variations of DE namely, the boundary based DE (B-DE) and
the penalty based DE (P-DE), are proposed to avoid non-physical values of the parameters. P-DE was
used to extract the parameters of the double diode PV model (Chin et al., 2015).

Bacteria Foraging Algorithm (BFA)

The Bacteria Foraging Algorithm is inspired by social foraging strategy of Escherichia coli bacteria (Das
et al., (2009). Each bacterium is analogous to a candidate solution moving on the functional surface to
locate the global optimum. Based on four processes: chemotaxis, swarming, reproduction, and elimina-
tion- dispersion. In chemotaxis, each bacterium can, via the rotation of its flagella, swim in a certain
direction or tumble. The bacteria react to surrounding nutrient gradient concentration (objective value)
by alternate between these two modes. Once a bacterium has located a high nutrient concentration, it
releases a cell-to-cell attractant, informing other bacteria to form a homocentric group with high bacte-
rial density to move toward the region. This process is known as swarming. During reproduction, the
least healthy bacteria eventually expire while healthier bacteria asexually split into two at their present
locations, keeping the swarm population constant.
Rajasekar et al., (2013) developed a Bacteria Foraging Algorithm modelling to compute all parameters
of the single diode PV model at different operating irradiance G and temperature T levels. By utilizing
only information available on the manufacturer’s datasheet, IPH and IS current values were analytically
computed, while the series resistance RS and shunt resistance RSH values were obtained by optimizing
the slope equation at the maximum power point.

Estimation of Model Parameters using Analytical Methods

To use the models described previously, the values of their equivalent electrical circuit components,
i.e. the parameters, need to be calculated. The practical parameters extract method is to estimate them

622

An Accurate and Efficient Analytical Method to Extract the Parameters

using the available information in the manufacturer datasheet. The equations are derived from the three
remarkable points on the I–V curve, namely, short circuit, open circuit and maximum power points. On
the other hand, beside these essential points, the derivatives functions may also be utilized to derive
one or two new equations required to complete the parameter extraction procedure. For instance, many
proposals extract approaches used reciprocal of slopes at the open circuit and short circuit conditions
to estimate RS and RSH. In some analytical approaches, the knowledge that slope of the power curve at
MPP equals to zero can serve as a reference to adjust the values of several parameters. The analytical
expressions for open circuit temperature coefficient Kv and short circuit temperature coefficient Ki are
used to provide information required to compute the parameters at various operating conditions different
of the STC. For a more realistic and accurate values of parameters, the effects different environmental
conditions must be take into account. Manufacturer datasheet provides the important data references at
STC. Some manufacturers provide I–V curves measured at conditions other than STC, but such infor-
mation is normally quite limited.

Proposed Hybrid Analytical and Soft Approach to Extract

the Four Parameters Single Diode PV Model

The stages of the analytical resolution of equations:

It is a common practice to characterize PV cells as a function of four datasheet information’s. They
correspond to the three operational point conditions:

• Short-Circuit Point: V = 0 and I = I SC ;

• Open Circuit Point: V = VOC and I = 0 ;
• Maximum Power Point: V = VM and I = I M .

This information is always provided with reference to the nominal condition or standard test condition
(STC) of temperature and solar irradiation: 1000W/m2 radiation level, AM1.5 air mass at the module
temperature of 25 °C (298 °K).
It is recalled that the four parameters to extract are the photocurrent IPH, the saturation current IS, the
series resistance RS and the shunt resistance RSH. It is clear that this analytical parameter extract approach
requires four equations to solve.
The early stages of this analytical approach consist of evaluating the equation (1) at the three remark-
able points of the I-V curve of the PV module:

• The open circuit point (VOC, 0): One writes the first simplified equation to be solved. The output
current is zero:

  V   V
0 = I PH − I S exp  OC  − 1 − OC (3)
 a. N S .VT   RSH

• The short circuit point (0, ISC): In the following equation to be solved, the output voltage is zero:

623
 
An Accurate and Efficient Analytical Method to Extract the Parameters

  R I   R I
I SC = I PH − I S exp  S SC  − 1 − S SC (4)
 a. N S .VT   RSH

• The maximum power point MPP (VM, IM):

 V + R I   V + R I
I M = I PH − I S exp  M  − 1 −
S M  M S M
(5)
  a. N S .VT   RSH

Using the relationship (3), the following expression of the photocurrent can be deduced:

  V   V
I PH = I S exp  OC  − 1 + OC (6)
 a. N S .VT   RSH

By substituting the relationship (6) in the equations (4) and (5), the two following expressions (7)
and (8) are derived with only three unknown parameters to be extract (IS, RS and RSH)

  V   R I  V − R I
I SC = I S exp  OC  − exp  S SC  + OC S SC
(7)
 a. N S .VT  a. N S .VT  RSH

  V  V + R I  V −V − R I
I M = I S exp  OC  − exp  M S M 
 +
OC M S M
(8)
 a.N S .VT   a.N S .VT  RSH

To clarify the equations and subsequently simplify their handling, one poses:

 V 
XOC = exp  OC  (9)
a. N S .VT 

V + R I 
X M = exp  M S M 
 (10)
 a. N S .VT 

 R I 
X S = exp  S SC  (11)
a. N S .VT 

624

An Accurate and Efficient Analytical Method to Extract the Parameters

At the standard test condition, XM and XS can be calculated if only the value of RS is determined. The
equations (7) and (8) can be written as follows, in a simpler manner:

 R  V
I SC 1 + S  =I S (XOC − X S ) + OC (12)
 RSH  RSH

 R  V −VM
I M 1 + S  =I S (XOC − X M ) + OC (13)
 RSH  RSH

Using different approaches, many research works authors calculate and find that RS values is of the
order of a few tenths of Ohm and RSH values is of the order of a few hundred ohms. Considering their
typical values, the approximation (14) is widely justified.

RS
1+ ≅ 1 (14)
RSH

So, considering this assumption the relationships (12) and (13) become relatively easier to solve:

VOC
I SC = I S (XOC − X S ) + (15)
RSH

VOC −VM
I M = I S (XOC − X M ) + (16)
RSH

Basis of the equations developed so far, one tries to express IS and RSH with only one unknown parameter:
series resistance RS. Then the expressions of saturation current IS and shunt resistance RSH are deduced:

VOC * (I SC − I M ) −VM * I SC
IS = (17)
VOC * (X M − X S ) −VM * (XOC − X S )

I SC * (X M − XOC ) + I M * (XOC − X S )
RSH −1 = (18)
VOC * (X M − X S ) −VM * (XOC − X S )

625
 
An Accurate and Efficient Analytical Method to Extract the Parameters

In the two last expressions, only the two terms XM and XS are unknown and depend only on RS and
the values of IS and RSH are calculated once RS value is determined. For this reason, one more equation
is required necessarily.
The electric power delivered by photovoltaic module is expressed as follows:

P =V * I (19)

Next, the derivative of the power expression with respect to the voltage at the maximum power point
(MPP) is zero. Considering this notion, the following equation can be deduced easily:

dP  dI 
=  V + I = 0 (20)
dV (V ,I ) dV 
M M

Which leads to the following relationship giving the value of the variation of the I-V curve at MPP:

dI I 
= −  M  (21)
dV (V VM 
M M)
,I

The equation (22) is obtained by taking the derivative of the equation (1) with respect to voltage V
at the MPP:

     
dI
=−
IS 1 − RS .I M  .exp VM + RS I M  − 1 1 − RS .I M  (22)
   a.N .V  R  
dV (V a.N S .VM  VM     VM 
M M)
,I S T SH

Considering the three equations (17), (18) and (22), the following relationships are derived:

IM  R .I   I S 1 
= 1 − S M   XM + (23)
VM  VM   a.N S .VT RSH 

Therefore, after some mathematical manipulations, one obtains the relationship that allows comput-
ing the value of the series resistance RS.CAL

VM 1
RS .CAL = − (24)
IM  I 1 
 
 a.N .V X M + R 
S

 S T SH 

One notes that the equations (6), (17) and (18) found analytically, allow the calculation of the pa-
rameters IPH, IS and RSH if RS value is calculated. It is obvious that the latest relationship (24) is strongly

626

An Accurate and Efficient Analytical Method to Extract the Parameters

nonlinear. Thus for its resolution, the authors consider more interesting to calculate numerically the value
of RS with a simple and accurate iterative algorithm and this imposes no difficulty.

Proposed Fast Numerical Algorithm to Extract RS

The simplified flowchart of the iterative algorithm is showed in Figure 3.

In this algorithm, the series resistance takes an initial value and increases gradually. For each value
of series resistance, IS and RSH are calculated using the relationships (17) and (18). The values of these
parameters will be validated if only the RS.CAL value calculated using the equation (24) coincides closely
to the value of RS taken at the beginning of each iteration. Unlike other algorithms developed and used
by other authors (Newton – Raphson, genetic algorithm, pattern search optimization, bacteria foraging
algorithm…), this algorithm is very simple, turn very fast and needs to extract only one parameter, thus
justifying its strength and power. The iterative method computes the solution RS = 0.2307Ω for the Kyo-
cera KC200GT solar panel. The others parameters are calculated at the same time, using the analytical
equations previously developed.

RESULTS AND DISCUSSIONS

Table 1 shows the four manufacturer’s data reference for the Kyocera KC200GT solar panel used in
this work for computing the four-parameter single diode PV model, i.e., the open-circuit voltage (VOC),
the short-circuit current (ISC), the current and the voltage at the maximum power point (IM, VM). These
data references concern the solar panel working at the Standard Test Condition (STC), i.e., 25 °C - A.M

Figure 3. The flowchart of the iterative matching algorithm in determining the final value of RS

627
 
An Accurate and Efficient Analytical Method to Extract the Parameters

Table 1. Manufacturer’s data reference for the KC200GT solar array at STC

Manufacturer’s Data Reference Rating Values

Voltage Open Circuit VOC 32.9 V
Current Short Circuit ISC 8.21 A
Voltage at MPP VM 26.3 V
Current at MPP IM 7.61 A
Maximal Power PM 200.143 W
Temperature coeff of open circuit voltage VOC KV -0.1230 V/K
Temperature coeff of short circuit current ISC KI 0.0032 A/K
Number of PV cells per module NS 54

1.5 and 1000 W/m2. In the table 1 one distinguishes, also other technical manufacture’s data: KV the
temperature coefficient of open circuit voltage VOC, KI Temperature coefficient of short circuit current
ISC and NS the number of PV solar cells per module.
The Kyocera KC200GT solar panel has been widely used by different authors to check different
parameters extraction methods: numerical (Villalva et al., 2004) or analytical (Cubas et al., 2014). In
table 2, the authors present the values of the four parameters calculated using this combined analytical
numerical method and then compare the results to what has been found by the previously cited authors.
VOC.CAL, ISC.CAL, VM.CAL, IM.CAL and PM.CAL are the recalculated data reference using the parameters values
determined by the present method and Villalva and Cubas method.
Villalva’s approach method intends to adjust RS and RSH based on the fact that there is only one pair
(RS, RSH), which guarantees that: PMP,MODEL = PMP, CURVE = VMIM at the Maximum Power Point (VMP, IMP)
of the I–V curve, i.e., the maximum power based on the simulated electrical model (PMP,MODEL) is equal
to the maximum power obtained from the datasheet curve (P MP, CURVE) at VM . First, the authors assume

Table 2. Comparative calculated parameters and data reference for the Kyocera KC200GT solar array
at STC

Parameters and Recalculated Data Present Calculated Values Cubas Calculated Villalva Calculated Values
Reference Values
RS (Ω) 0.2307 0.2308 0.221
RSH (Ω) 603.83 597.39 415.41
IPH (A) 8.2100 8.2132 8.214
IS1 (nA) 98.656 97.631 98.25
a 1.3 1.3 1.3
VOC.CAL (V) 32.9 32.9 32.9
ISC.CAL (A) 8.21 8.21 8.21
VM.CAL (V) 26.3 26.3 26.3
IM.CAL (A) 7.61 7.61 7.61
PM.CAL (W) 200.143 200.143 200.143

628

An Accurate and Efficient Analytical Method to Extract the Parameters

that the ideality factor a can be chosen arbitrarily, assuming a constant value between 1 and 1.5 (a is
equal to 1.3), and then, it can be changed to better fit the I–V curve, if necessary. Villalva uses many
simplifications in order to extract the initial value of shunt resistance RSH for its algorithm.

Proposed Hybrid Analytical and Numerical Approach to

Extract the Five-Parameter Double Diode PV Model

The stages of the analytical resolution of equations:

In its original form, the double diode model is not attractive for PV simulation due to its complexity.
It needs to solve seven model parameters. In order to make this model analytically manageable, the
seven unknown parameters will be reduced to five by assuming a1 = 1, a2 = 2 ; this assumption is
widely used in the literature.
It is recalled that the five parameters to extract are the photocurrent IPH, the saturation currents (IS1,
IS2), the series resistance RS and the shunt resistance RSH. It is clear that this analytical parameter extract
approach requires five equations to solve.
Considering once again the equation (2) at the three remarkable points of the I-V curve of the PV
module:

• The open circuit point (VOC, 0):

  V     V   V
0 = I PH − I S 1 exp  OC  − 1 − I S 2 exp  OC  − 1 − OC (25)
  N S .VT     2. N S .VT   RSH

• The short circuit point (0, ISC):

  R I     R I   R I
I SC = I PH − I S 1 exp  S SC  − 1 − I S 2 exp  S SC  − 1 − S SC (26)
  N S .VT     2. N S .VT   RSH

• The maximum power point MPP (VM, IM):

 V + R I    
I M = I PH − I S 1 exp  M S M 
− 1 − I exp VM + RS I M  − 1 −VM + RS I M (27)

 
 N S .VT   S   2.N .V   RSH
 S T  

Using the equation (25), one deduces the expression (28) of photocurrent:

  V     V   V
I PH = I S 1 exp  OC  − 1 + I S 2 exp  OC  − 1 + OC (28)
  N S .VT     2. N S .VT   RSH

By substituting the expression (28) in (26) and (27), one obtains the two following equations (29)
and (30) derived with four unknown parameters to be extract (IS1, IS2, RS and RSH)

629
 
An Accurate and Efficient Analytical Method to Extract the Parameters

  V   R I 
I SC = I S 1 exp  OC  − exp  S SC  +
  N S .VT   N S .VT 
(29)
  V   R I  V − R I
 
I S 2 exp  
 − exp  
 2.N .V  + 
OC S SC OC S SC

  2.N S .VT   S T 
RSH

  V  V + R I 
I M = I S 1 exp  OC  − exp  M S M 

  N S .VT   N S .VT 

  V  V + R I  V −V − R I
+I S 2 exp  OC  − exp  M S M 
 +
OC M S M
(30)
  2. N S .VT   2. N S .VT  RSH

To clarify the equations and subsequently simplify their handling, one poses:

 V   V 
X1OC = exp  OC  and X 2OC = exp  OC  (31)
 N S .VT   2.N S .VT 

V + R I  V + R I 
X1M = exp  M S M 
 and X 2M = exp  M S M 
 (32)
 N S .VT   2.N S .VT 

R I   R I 
X1S = exp  S SC  and X 2S = exp  S SC  (33)
 N S .VT   2. N S .VT 

At the standard test condition (STC), the four terms X1,2M and X1,2S can be calculated if only RS value
is determined. The equations (29) and (30) can be written as follows:

 R  V
I SC 1 + S  − OC =I S 1 (X1OC − X1S ) + I S 2 (X 2OC − X 2S ) (34)
 RSH  RSH

 R  V
(I SC − I M ) 1 + S  − M = I S 1 (X1M − X1S ) + I S 2 (X 2M − X 2S ) (35)
 RSH  RSH

630

An Accurate and Efficient Analytical Method to Extract the Parameters

From these equations, one can try to express IS1 and IS2 with only two unknown parameters: series
and shunt resistances RS and RSH. Then we deduce the expression of IS1 and IS2:

1
IS1 = A + *B (36)
RSH

1
I S 2 = A′ + * B′ (37)
RSH

with

X DP = (X1OC − X1S ) * (X 2M − X 2S ) − (X 2OC − X 2S ) * (X1M − X1S ) .

A= (X 2M − X 2S ) I SC − (X 2OC − X 2S )(I SC − I M ) / X DP .

 

A′ = (X1OC − X1S )(I SC − I M ) − (X1M − X1S ) I SC  / X DP .

 

( )
B =  VM − RS (I SC − I M ) (X 2OC − X 2S ) − (VOC − RS I SC )(X 2M − X 2S ) / X DP .
 

( )
B ′ = (VOC − RS I SC )(X1M − X1S ) − VM − RS (I SC − I M ) (X1OC − X1S ) /X DP .
 

There are two expressions with four unknown parameters IS1, IS2, RS and RSH. Two more equations
are required for solving and computing the values of these four parameters. The fifth parameter IPH can
be computed thereafter.
For this reason, the expression of derivative of the delivered electrical power with respect to the volt-
age is considered. At the Maximum Power Point (MPP), this expression is equal to zero:

dP  dI 
=  V + I = 0
dV (V ,I ) dV 
M M

Which leads to the following relationship ( RM represents fictive resistance):

dI I  1
= −  M  = − (38)
dV (V ,I )
VM  RM
M M

631
 
An Accurate and Efficient Analytical Method to Extract the Parameters

   

dI
= −
IS1 1 − RS .I M  .exp VM + RS I M  −
   
dV (V ,I M )
N S .VT  VM   N S .VT 
M

I S 2  RS .I M  V + R I 

1 −  .exp  M S M 
2.N S .VT  VM   2.N .V 
 S T 

 
−
1 1 − RS .I M  (39)
 
RSH  VM 

Considering the two latest equations, the two following relationships are deduced:

IM  R  I IS 2 1 
= 1 − S   S 1 X1M + X 2M + (40)
VM  RM   N S .VT 2.N S .VT RSH 

1 1 I IS 2
= − S 1 X1M − X (41)
RSH RM − RS N S .VT 2.N S .VT 2M

By using the expression (41) in the equations (36) and (37), the parameters IS1 and IS2 can be expressed
with only one unknown parameter RS. RSH is substituted by its expression developed in equation (41).

 1 I IS 2 
I S 1 = A +  − S 1 X1M − X 2M  * B (42)
 RM − RS N S .VT 2.N S .VT 

 1 I IS 2 
I S 2 = A′ +  − S 1 X1M − X 2M  * B ′ (43)
 RM − RS N S .VT 2.N S .VT 

The two equations (42) and (43) can be written otherwise

B  B.X1M   B.X 
A+ = I S 1 * 1 +  + I S 2  2M 
 2.N .V  (44)
RM − RS  N S .VT   S T 

B′  B ′.X   B ′.X 2M 
A′ + = I S 1   1M  
 + I S 2 1 +  (45)
RM − RS  N S .VT   2.N S .VT 

632

An Accurate and Efficient Analytical Method to Extract the Parameters

Solving the set of equations (44) and (45) with respect to the unknown variables IS1 and IS2 yields:

 B   B ′.X 2M   B '   B.X 2M 

A +  * 1 +  − A '+
+
 * 
 RM − RS   2.N S .VT   RM − RS   2.N S .VT 
IS1 = (46)
 
1 + B.X1M + B ′.X 2M 
 
 N S .VT 2.N S .VT 

 B '   B.X1M   B   B '.X 

A '+ *  1 +  − A +  *  1M 
       N .V 
 R − RS  
N V
.
S T   R − RS   S T 
IS 2 = M M
(47)
 ′ 
1 + B.X1M + B .X 2M 
 
 N S .VT 2.N S .VT 

After some mathematical manipulations, one obtains the relationships (48) and (49) that allows us
to compute the values of IS1 and IS2 for any value of the series resistance RS.

 B  (A.B ′ − A′.B ).X 2M

A +  +
 RM − RS  2.N S .VT

IS1 = (48)
 
1 + B.X1M + B ′.X 2M 
 
 N S .VT 2.N S .VT 


A '+ B '  (A '.B − A.B ').X1M
 +
 RM − RS  N S .VT
IS 2 = (49)
 
1 + B.X1M + B ′.X 2M 
 
 N S .VT 2.N S .VT 

Therefore, the value of RSH can be calculated using the following expression:

B
RSH = (50)
I S 1 − A

One notes that the three independent equations (48), (49) and (50) found analytically allow calculation
of the three parameters IS1, IS2 and RSH if only the value of series resistance RS is known. Then in order
to calculate or validate the value of RS, one more equation is required. In this context, one considers the
equation (51) obtained by taking the derivative of the equation (2) with respect to output voltage V at
the short-circuit point at the I-V curve:

633
 
An Accurate and Efficient Analytical Method to Extract the Parameters

dI 1
=− (51)
dV (V =0,I =I ) RSH 0
SC

     
dI I
= − S1 1 − RS  .X − I S 2 1 − RS  .X − 1 1 − RS  (52)
     
dV (V =0,I =I ) N S .VM  RSH 0  1S 2.N S .VM  RSH 0  2S RSH  RSH 0 
SC

Therefore, one obtains the followings equation:

1  I IS 2 1 
=  S 1 X1S + X 2S +  (53)
RSH 0 − RS  N S .VT 2.N S .VT RSH 

Many paper authors consider that RSH 0 ≈ RSH . This can be justified by the fact that the series resis-
IS1 IS 2
tance is very small compared to RSH and the terms X1S + X are very small compared
N S .VT 2.N S .VT 2S
1
to . Therefore, RSH0 is replaced by RSH in the equation (54).
RSH
 I IS 2 
RS = RSH (RSH − RS ) S 1 X1S + X 2S  (54)
 N S .VT 2.N S .VT 

With two exponential terms, the latest equation is highly nonlinear.

Proposed Fast Numerical Algorithm to Extract RS

The latest relationship (54) is strongly nonlinear and do not allows us to calculate the value of RS analyti-
cally. The same soft algorithm developed for the single diode model, is proposed but with much more
calculations. The numerical approach is to make several iterations for different values of RS and compute
IS1, IS2, RSH and RSCAL using the equations (42), (43), (44), and (48) until the two values RSH and RSCAL are
coincided, thereby validating the value taken for RS.

 I IS 2 
RSCAL = RSH (RSH − RS ) S 1 X1S + X 2S  (55)
 N S .VT 2.N S .VT 

The iterative method gives the solution RS = 0.318 Ω for the KC200GT PV solar panel. The others
parameters are calculated at the same time.
It will be very interesting to draw a comparative between this hybrid approach and another method
recently developed in the same context by other leading scientists. The authors Hejri et al., (2014) has
developed an analytical method based on several mathematical approximations. Their analytical results

634

An Accurate and Efficient Analytical Method to Extract the Parameters

found were approximate. Then they used these results as initial values for a second numerical method
based on Newton-Raphson method much slower and requiring much calculation and with high possibil-
ity of the numerical divergence due to high sensitivity of the accuracy of the initial values. Ishaque et
al., (2011c) has developed a simple and fast approach but with many approximations and mathematical
simplifications.
In Table 3, are presented the final values of the five parameters calculated using the proposed hybrid
analytical numerical method and compared with the results that has been found by the previously cited
authors.
In order to validate the results of this combined approach, the four data reference are recalculated
using the parameter values obtained and the different equations developed. The calculated values VOC.CAL,
ISC.CAL, VM.CAL, VM.CAL and PM.CAL, are exactly similar to the manufacture’s datasheet reference. It represents
the first successful stage of verification and validation of the calculated model.

SOLUTIONS AND RECOMMENDATIONS

Validating the PV Model with the Calculated Parameters

To test even more the validity of this method and its efficiency and accuracy, the other experimental
(I, V) points extracted from the datasheet will be compared with the I-V curve obtained using the PV
calculation model circuit shown in Figure 4.
Once the appropriate model and its computational model have been identified, a complete PV system
simulation can be developed. A good PV simulation package should fulfill the two following criteria:

• It should be fast but can accurately predict the I-V and P-V characteristic curves, including special
conditions such as partial shading;

Table 3. Comparative calculated parameters and recalculated data reference for the KC200GT Solar
array at STC

Parameters and Data Hybrid Method Hejri Calculated Hejri Calculated Ishaque et al. Calculated
Reference Calculated Values Analytical Values Numerical Values Values
RS (Ω) 0.31814 0.2280 0.3181 0.32
RSH (Ω) 278.83 135.52 278.92 160.5
IPH (A) 8.2194 8.2100 8.2193 8.21
IS1 (nA) 0.3795 0.2915 0.3795 0.4704
IS2 (µA) 4.4324 17.558 4.4330 0.4704 e-3
VOC.CAL (V) 32.9
ISC.CAL (A) 8.21
IM.CAL (A) 7.61
PM.CAL (W) 200.143

635
 
An Accurate and Efficient Analytical Method to Extract the Parameters

Figure 4. The flowchart of the iterative matching algorithm in determining the final value of RS

• It should be a comprehensive tool to develop and validate the PV system design inclusive of the
power converter and MPPT control.

In this section, using the package MATLAB/SIMULINK environment, the electrical equivalent circuit
based on the double diode PV model is achieved. In this simulated circuit as shown in the Figure 4, one
distinguishes a controlled current source, which replace the photocurrent source and the two diodes.
The two resistances are also included in the circuit. The value of the controlled current IC = IPH - ID1 - ID2
is generated by a computational block that has the output voltage V, output current I, temperature and
irradiation levels (T, G) and the different calculated parameters as inputs. This block is implemented
in order to evaluate the relationship (2). As charge, is connected a controlled voltage source where the
voltage can take several values between 0 and VOC. The controlled voltage source can represent a boost
or buck-boost DC-DC converter used for the MPPT control in grid-connect PV system or stand-alone
PV supply.
As it can be seen from Figures 6 and 7, the values calculated and plotted of P-V and I-V points are
very close to the real values extracted from manufacturer’s datasheet, and again, this confirms the ac-
curacy of the five parameters calculated of the double-diode PV cell model developed and presented
in this chapter. This hybrid analytical and numerical approach, the five parameters values are extracted
using only four analytical expressions and a very fast and simple algorithm.

Dependence of the Parameters on the Temperature and Irradiation Levels

In order to complete this study, the dependence of the five parameters on the temperature and irradia-
tion levels is considered in this chapter. The four manufacturer’s data reference refers to the standard

636

An Accurate and Efficient Analytical Method to Extract the Parameters

Figure 5. The electrical equivalent circuit of the double-diode PV model implemented in Matlab/Simulink

Figure 6. Comparative between I-V curves extract from datasheet and obtained using the calculated PV
model, for KC200GT Solar Panel at STC

test condition (STC), and therefore, the five calculated parameters are determined only at STC. The
dependence of the PV model parameters on T and G can be inserted in the mathematical model via a
set of suitable transactional formulas that Hejri et al., (2014) use in their work:

GSTC
RS = RS ,STC , RSH = RSH ,STC (56)
G

637
 
An Accurate and Efficient Analytical Method to Extract the Parameters

Figure 7. Comparative between P-V curves extract from datasheet and obtained using the calculated PV
model, for KC200GT Solar Panel at STC

 T 
3
 E E 
I S 1 = I S 1,STC   exp  1  g ,STC − g  (57)
T    
 STC   k  TSTC T 


5
 T 2  E E 
I S 2 = I S 2,STC   exp  1  g ,STC − g  (58)
T    
 STC   2k  TSTC T 


E g = E g ,STC (1 − 0, 0002677 (T −TSTC ) (59)

G 
I PH = I + K I (T −TSTC ) (60)
GSTC  PH ,STC 

where the subscript STC stands for the values of different parameters in Standard Test Condition. Eg is
the material band gap energy and its value in TSTC = 25°C is Eg, STC =1.121 eV for silicon cells and the
expression of Eg as function of T is given in (59).

638

An Accurate and Efficient Analytical Method to Extract the Parameters

Figure 8. P-V curves obtained for KC200GT solar panel at 25°C and at many levels of irradiance, using
the calculated PV model

Figure 9. I-V curves obtained for KC200GT solar panel at 25°C and at many levels of irradiance, using
the calculated PV model

FUTURE RESEARCH DIRECTIONS

Notwithstanding their known limitations at low levels of irradiance, the hybrid approach for parameter
extract of the single diode PV model is still interesting for PV systems analysis and simulation. This is
mainly due to the practical fact that the four manufacture’s data reference is sufficient to solve four analyti-
cal equations developed without any physical assumptions or mathematical simplifications. However, its

639
 
An Accurate and Efficient Analytical Method to Extract the Parameters

Figure 10. P-V curve obtained for KC200GT solar panel at 1000W/m2 and at many temperatures, using
the calculated PV model

Figure 11. I-V curve obtained for KC200GT solar panel at 1000W/m2 and at many temperatures, using
the calculated PV model

accuracy deteriorates at low irradiance and with the increasing interest in the study of the characteristics
of PV array under partial shading conditions, there is a need for a high-precision simulation especially
at low irradiance. The double diode PV model is known to have better accuracy under low irradiance,
allowing more accurate prediction of PV system performance during partial shading condition. The
proposed model supports a large simulation array and it can be interfaced with MPPT algorithms and
power electronic converters (Chin et al., 2015).

640

An Accurate and Efficient Analytical Method to Extract the Parameters

Based on these developments, elaborate models such as the two or three diode PV model developed
using more accurate approaches is becoming more relevant. As can be seen in this chapter, the promis-
ing technique is to compute the parameters model using the hybrid approach, i.e. by combining the fast
numeric computing with transparent analytical methods.
The double diode PV simulator based on this approach is much more powerful and can be integrated
in more complex PV systems for general overall simulation: MPPT, Partial shading and power injection
into the electrical grid.
Given the increase in the computational power of computers, numerical methods are expected to
be widely used. Despite this, they are currently not preferred due to their slow convergence speed and
their complexity. Therefore, it is advisable to combine the soft computing with the analytical methods
especially for more complicated models: seven-parameter double-diode models or triple-diode models
with several parameters to be determined. Furthermore, with more data reference given by manufacturer
and consequently with more analytical equations developed, it leads to a fast and precise parameters
extraction method. An example of additional data that the manufacturer should provide is the slope of the
current-voltage characteristic in the vicinity of the open circuit point for different levels of temperature
and irradiance.
On another note, the thin film solar cell modules market has been expanding in recent years. However,
to date, the most photovoltaic models are still predominantly based on monocrystalline and polycrystal-
line silicon structure. In their work, Laudani et al., (2013) note that the amorphous thin film solar cell
have particularly high ideality factor, due to their low fill factor. Yet, currently most models assume the
ideality factor is within the range between 1 and 2 (1 < a < 2). Furthermore, there are few dedicated
simulation works concerning the multi-junction PV cells, organic PV cells and solar concentrators.
These are growing areas of interests and apparently, there are quite a number of specific issues in their
model that need to be resolved. Finally there are concerns on the cell deterioration as a function time
(ageing) and weather conditions. Additional coefficients should be added to the equivalent circuit model
to mimic the effects of degradation for different types of PV modules. Such work would provide a better
understanding of cell performance over a long time. That can be very beneficial in estimating in precise
manner, the payback period for the consumers.

CONCLUSION

This chapter has studied the extraction of PV model parameters that are based only on analytical transpar-
ent equations without any mathematical simplifications or physical assumptions used by many authors.
The main problem that has been evoked and investigated is the accuracy of the determined parameter
values calculated using the improved hybrid method. The two main models of PV cell-namely single
diode and the double diode models are elaborated to highlight their advantages and drawbacks. In this
combined analytical and numerical parameters extraction method, the PV current-voltage relationship
evaluated at the three remarkable points, offers three analytical equations. One more equation can be
developed considering the derivative of the PV power with respect to the voltage made equal to zero at
the MPP. The fifth equation was obtained by taking the derivative of the current with respect to voltage
at the short-circuit point in the I-V curve. The resulting solutions lead to determine analytically and ac-
curately the parameters IS1, IS2, IPH and RSH as expression of RS. The fact that the latest equation was highly
nonlinear, its resolution was done by developing a simple and fast iterative algorithm. At the last stage

641
 
An Accurate and Efficient Analytical Method to Extract the Parameters

of this approach, the value of the series resistance RS is computing without considering any simplifica-
tion. The validity of this approach and the accuracy of the resulting calculated equivalent circuit of the
single-diode and double-diode PV model has been evaluated by comparing the calculated values of the
three remarkable points with the three references datasheet and it is shown that the different simulated
I-V curves obtained based on experimental data are sufficiently precise and accurate. Finally, this work
deliberates on the prospective areas of research for the future PV simulators. This approach can be
extended to multi-diode model. In this case, the soft computing can be more developed and complex.
This chapter has the merit to present in detail the calculation steps and equations used to extract
the PV model parameters at STC or at any different operating conditions. This modeling can be used
to implement or compare different MPPT algorithms. It can also be used to implement control laws
and regulation techniques in the grid connected PV systems. This parameters extract method is very
attractive in industrial photovoltaic applications because of its speed and need for few data from the I-V
characteristic curves which are commonly available in the photovoltaic manufacturer catalogues.

REFERENCES

AlHajri, M. F., El-Naggar, K. M., AlRashidi, M. R., & Al-Othman, A. K. (2012). Optimal extraction of solar
cell parameters using pattern search. Renewable Energy, 44, 238–245. doi:10.1016/j.renene.2012.01.082
Askarzadeh, A., & Rezazadeh, A. (2011). A new artificial bee swarm algorithm for optimization of proton
exchange membrane fuel cell model parameters. Journal of Zhejiang University Science Computer &
Electronics, Springer, 12(8), 638–646. doi:10.1631/jzus.C1000355
Askarzadeh, A., & Rezazadeh, A. (2013). Artificial bee swarm optimization algorithm for parameters
identification of solar cell models. Applied Energy, 102, 943–949. doi:10.1016/j.apenergy.2012.09.052
Bãnos, R., Manzano-Agugliaro, F., Montoya, F., Gil, C., Alcayde, A., & Gomez, J. (2011). Optimiza-
tion methods applied to renewable and sustainable energy: A review. Renewable & Sustainable Energy
Reviews, 15(4), 1753–1766. doi:10.1016/j.rser.2010.12.008
Chan, D., & Phang, J. (1987). Analytical methods for the extraction of solar-cell single and double-diode
model parameters from I-V characteristics. IEEE Transactions on Electron Devices, 34(2), 286–293.
doi:10.1109/T-ED.1987.22920
Chegaar, M., Ouennoughi, Z., & Guechi, F. (2004). Extracting dc parameters of solar cells under il-
lumination. Vacuum, 75(4), 367–372. doi:10.1016/j.vacuum.2004.05.001
Chen, Y., Wang, X., Li, D., Hong, R., & Shen, H. (2011). Parameters extraction from commercial solar
cells I-V characteristics and shunt analysis. Applied Energy, 88(6), 2239–2244. doi:10.1016/j.apen-
ergy.2010.12.048
Chin, V. J., Ishaque, K., & Salam, Z. (2015). Cell modelling and model parameters estimation tech-
niques for photovoltaic simulator application: A review. Applied Energy, 154, 500–519. doi:10.1016/j.
apenergy.2015.05.035
Cubas, J., Pindado, S., & Victoria, M. (2014). On the analytical approach for modeling photovoltaic
systems behavior. Journal of Power Sources, 247, 467–474. doi:10.1016/j.jpowsour.2013.09.008

642

An Accurate and Efficient Analytical Method to Extract the Parameters

Da Costa, W., Fardin, J., Simonetti, D., & Neto, L. V. (2010). Identification of photovoltaic model pa-
rameters by differential evolution. In International Conference on Industrial Technology, (pp. 931-936).
doi:10.1109/ICIT.2010.5472557
Das, S., Biswas, A., Dasgupta, S., & Abraham, A. (2009). Bacterial foraging optimization algorithm:
Theoretical foundations, analysis, and applications. Foundations of Computational Intelligence, Springer,
3, 23–55.
Easwarakhanthan, T., Bottin, J., Bouhouch, I., & Boutrit, C. (1986). Nonlinear minimization algorithm
for determining the solar cell parameters with microcomputers. International Journal Solar Energy,
4(1), 1–12. doi:10.1080/01425918608909835
El-Naggar, K. M., AlRashidi, M. R., AlHajri, M. F., & Al-Othman, A. K. (2012). Simulated anneal-
ing algorithm for photovoltaic parameters identification. Solar Energy, 86(1), 266–340. doi:10.1016/j.
solener.2011.09.032
Hejri, M., Mokhtari, H., Azizian, M. R., Ghandhari, M., & Soder, L. (2014). On the Parameter Extrac-
tion of a Five-Parameter Double-Diode Model of Photovoltaic Cells and Modules. IEEE Journal of
Photovoltaics, 4(3), 915–923. doi:10.1109/JPHOTOV.2014.2307161
Hernanz, R. J., Campayo, M. J., Zamora, B. I., Larranaga, L. J., Zulueta, G. E., & Puelles, P. E. (2010).
Modelling of photovoltaic module. In International Conference on Renewable Energies and Power Quality.
Huang, W., Jiang, C., Xue, L., & Song, D. (2011). Extracting solar cell model parameters based on chaos
particle swarm algorithm. In International Conference on Electric Information and Control Engineering,
(pp. 398-402). doi:10.1109/ICEICE.2011.5777246
Ishaque, K., & Salam, Z. (2011b). An improved modeling method to determine the model parameters
of photovoltaic (PV) modules using differential evolution (DE). Solar Energy, 85(9), 2349–2359.
doi:10.1016/j.solener.2011.06.025
Ishaque, K., Salam, Z., & Taheri, H. (2011a). Simple, fast and accurate two-diode model for photovoltaic
modules. Solar Energy Materials and Solar Cells, 95(2), 586–594. doi:10.1016/j.solmat.2010.09.023
Ishaque, K., Salam, Z., Taheri, H., & Shamsudin, A. (2011d). A critical evaluation of EA computational
methods for photovoltaic cell parameter extraction based on two diode model. Solar Energy, 85(9),
1768–1779. doi:10.1016/j.solener.2011.04.015
Ishaque, K., Salam, Z., Taheri, H., & Syafaruddin, . (2011c). Modeling and simulation of photovoltaic
(PV) system during partial shading based on a two-diode model. Simulation Modelling Practice and
Theory, 19(7), 1613–1626. doi:10.1016/j.simpat.2011.04.005
Jervase, J. A., Bourdoucen, H., & Al-Lawati, A. (2001). Solar cell parameter extraction using genetic
algorithms. Measurement Science & Technology, 12(11), 1922–1925. doi:10.1088/0957-0233/12/11/322
Kennedy, J. (2010). Particle swarm optimization. In Encyclopedia of machine learning (pp. 760–766).
Springer.
Kumari, J. S., & Babu, C. S. (2012). Mathematical modeling and simulation of photovoltaic cell using
Matlab-Simulink environment. Iranian Journal of Electrical and Computer Engineering, 2(1), 26–34.

643
 
An Accurate and Efficient Analytical Method to Extract the Parameters

Laudani, A., Mancilla-David, F., Riganti-Fulginei, F., & Salvini, A. (2013). Reduced-form of the pho-
tovoltaic five-parameter model for efficient computation of parameters. Solar Energy, 97, 122–129.
doi:10.1016/j.solener.2013.07.031
Macabebe, E., Sheppard, C. J., & Van Dyk, E. E. (2011). Parameter extraction from I–V characteristics
of PV devices. Solar Energy, 85(1), 12–18. doi:10.1016/j.solener.2010.11.005
Majdoul, R., Abdelmounim, E., Aboulfatah, M., Touati, A., & Moutabir, A. (2015). Combined analytical
and numerical approach to determine the four parameters of the photovoltaic cells models. In International
Conference on Electrical and Information Technologies, (pp. 263-268). doi:10.1109/EITech.2015.7162977
Mellit, A., & Kalogirou, S. A. (2008). Artificial intelligence techniques for photovoltaic applications: A
review. Progress in Energy and Combustion Science, 34(5), 574–632. doi:10.1016/j.pecs.2008.01.001
Moldovan, N., Picos, R., & Garcia-Moreno, E. (2009). Parameter extraction of a solar cell compact
model using genetic algorithms. In Spanish conference on electron devices, (pp. 379-382).
Ortiz-Conde, A., Sanchez, F. J. G., & Muci, J. (2006). New method to extract the model parameters
of solar cells from the explicit analytic solutions of their illuminated I-V characteristics. Solar Energy
Materials and Solar Cells, 90(3), 352–361. doi:10.1016/j.solmat.2005.04.023
Rajasekar, N., Kumar, N. K., & Venugopalan, R. (2013). Bacterial foraging algorithm based solar PV
parameter estimation. Solar Energy, 97, 255–265. doi:10.1016/j.solener.2013.08.019
Saleem, H., & Karmalkar, S. (2009). An analytical method to extract the physical parameters of a so-
lar cell from four points on the illuminated J-V curve. IEEE Electron Device Letters, 30(4), 349–352.
doi:10.1109/LED.2009.2013882
Sandrolini, L., Artioli, M., & Reggiani, U. (2010). Numerical method for the extraction of photovoltaic
module double–diode model parameters through cluster analysis. Applied Energy, 87(2), 442–451.
doi:10.1016/j.apenergy.2009.07.022
Sellami, A., & Bouaicha, M. (2011). Application of the genetic algorithms for identifying the electrical
parameters of PV solar generators. In K. Leonid (Ed.), Solar cells-silicon wafer-based technologies (pp.
359–364). Croatia: InTech. doi:10.5772/22714
Sera, D., Teodorescu, R., & Rodriguez, P. (2007). PV panel model based on datasheet values.IEEE
International Symposium on Industrial Electronics, (pp. 2392–2396). doi:10.1109/ISIE.2007.4374981
Storn, R., & Price, K. (1997). Differential Evolution - A simple and efficient heuristic for global optimization
over continuous spaces. Journal of Global Optimization, 11(4), 341–359. doi:10.1023/A:1008202821328
Suthar, M., Singh, G. K., & Saini, R. P. (2013). Comparison of mathematical models of photovoltaic
(PV) module and effect of various parameters on its performance. In International Conference on Energy
Efficient Technologies for Sustainability, (pp. 1354–1359). doi:10.1109/ICEETS.2013.6533584
Villalva, M., Gazoli, J., & Filho, E. (2009). Comprehensive approach to modeling and simulation
of photovoltaic arrays. IEEE Transactions on Power Electronics, 24(5), 1198–1208. doi:10.1109/
TPEL.2009.2013862

644

An Accurate and Efficient Analytical Method to Extract the Parameters

Wei, H., Cong, J., Lingyun, X., & Deyun, S. (2011). Extracting solar cell model parameters based on
chaos particle swarm algorithm. In International Conference on Electric Information and Control En-
gineering. (pp. 398–402).
Wenyin, G., & Zhihua, C. (2013). Parameter extraction of solar cell models using repaired adaptive dif-
ferential evolution. Solar Energy, 94, 209–220. doi:10.1016/j.solener.2013.05.007
Ye, M., Wang, X., & Xu, Y. (2009). Parameter extraction of solar cells using particle swarm optimiza-
tion. Journal of Applied Physics, 105(9), 094502. doi:10.1063/1.3122082
Zagrouba, M., Sellami, A., Bouaicha, M., & Ksouri, M. (2010). Identification of PV solar cells and
modules parameters using the genetic algorithms: Application to maximum power extraction. Solar
Energy, 84(5), 860–866. doi:10.1016/j.solener.2010.02.012

KEY TERMS AND DEFINITIONS

Air Mass: It defines the direct optical path length through the Earth’s atmosphere, expressed as a
ratio relative to the path length vertically upwards, i.e. at the zenith. The air mass coefficient can be used
to help characterize the solar spectrum after solar radiation has traveled through the atmosphere. The
air mass coefficient is commonly used to characterize the performance of solar cells under standardized
conditions, and is often referred to using the syntax “AM” followed by a number. “AM1.5” is almost
universal when characterizing terrestrial power-generating panels.
Buck Converter: A type of DC-DC converter that has an output voltage magnitude less than the
input voltage magnitude.
Fill Factor: Abbreviated FF, is a parameter which characterizes the non-linear electrical behavior of
the solar cell. Fill factor is defined as the ratio of the maximum power from the solar cell to the product
of Open Circuit Voltage and Short-Circuit Current.
Grid-Connect PV System: Or grid-connected photovoltaic power system is an electricity generating
solar PV system that is connected to the utility grid. It consists of solar panels, one or several inverters,
a power-conditioning unit and grid connection equipment.
Multi-Junction Solar Cell: Solar cells with multiple (three or four) P-N junctions made of differ-
ent semiconductor materials. Each material’s P-N junction will produce electric current in response to
different wavelengths of light. The use of multiple semiconducting materials allows the absorbance of
a broader range of wavelengths, improving the cell’s sunlight to electrical energy conversion efficiency.
Partial Shading: Shading effects is due to dusts, clouds, leafs, branches of trees and buildings caus-
ing shading on part of cells, modules or panels.
Photovoltaic System: Also: solar PV power system, or PV system, is a power system designed to
supply usable solar power by means of photovoltaics. It consists of an arrangement of several compo-
nents, including solar panels to absorb and convert sunlight into electricity, a solar inverter to change the
electric current from DC to AC, as well as mounting, cabling and other electrical accessories to set up
a working system. It may also use a solar tracking system to improve the system’s overall performance
and include an integrated battery solution, as prices for storage devices are expected to decline. Strictly
speaking, a solar array only encompasses the ensemble of solar panels, the visible part of the PV system,

645
 
An Accurate and Efficient Analytical Method to Extract the Parameters

and does not include all the other hardware, often summarized as balance of system (BOS). Moreover,
PV systems convert light directly into electricity and should not be confused with other technologies,
such as concentrated solar power solar thermal, used for heating and cooling.
Stand-Alone Power System: (SAPS or SPS), also known as remote area power supply (RAPS), is
an off-the-grid electricity system for locations that are not fitted with an electricity distribution system.
Typical SAPS include one or more methods of electricity generation, energy storage, and regulation.
Electricity is typically generated by one or more of the following methods: Photovoltaic system using
solar panels, Wind turbine or Diesel or bio fuel generator.
Thin Film Solar Cell: A second generation solar cell that is made by depositing one or more thin
layers or thin film (TF) of photovoltaic material on a substrate, such as glass, plastic or metal.

646