A Neural Network Based MPPT Technique Controller for Photovoltaic Pumping System

International Journal of Power Electronics and Drive System (IJPEDS)
Vol. 4, No. 2, June 2014, pp. 241~255
ISSN: 2088-8694  241
Journal homepage: http://iaesjournal.com/online/index.php/IJPEDS
A Neural Network Based MPPT Technique Controller for
Photovoltaic Pumping System
Mohammed Yaichi*, Mohammed-Karim Fellah**, Abdelkrim Mammeri*
* Photovoltaic Pumping Team, Research Unit in Renewable Energies in the Saharan Medium URER/MS-Adrar, CDER
** Intelligent Control and Electrical Power Systems Laboratory, Djillali Liabes University of Sidi-Bel-Abbes
Article Info ABSTRACT
Article history:
Received Jan 27, 2014
Revised Mar 20, 2014
Accepted Apr 8, 2014
The article proposes a novel method using the artificial neural network
(ANN) for the improvement of the performances of a photovoltaic system
composed of a photovoltaic (PV) array, an inverter, a motor asynchronous
and a centrifugal pump. For this type of system, different optimization
strategies have been proposed to improve the over of the PV system
efficiency, i.e. the PV generator is forced to operate at its maximum power
point “MPPT”, generally, by the insertion of DC/DC boost converter
between the photovoltaic array and the inverter. In this work we propose an
approach, where optimization is realized without need adding a DC/DC
converter to the chain, using field-oriented control through the monitoring of
the voltage-fed inverter frequency. The motor is also ensured in all insolation
conditions. A multilayer feed forward perception type NN is proposed for
MPPT control, and the back-propagation algorithm is used for training. The
performances of the drive with ANN-based MPPT are excellent. The
maximum power point (MPP) can be easily obtained to frequency-controlled
drive.
Keyword:
PV Pumping System
Field-oriented control
MPPT
Neural Network
Performances
Copyright © 2014 Institute of Advanced Engineering and Science.
All rights reserved.
Corresponding Author:
Mohammed Yaichi,
Photovoltaic Pumping Team, Photovoltaic Conversion Division,
Research Unit in Renewable Energies in the Saharan Medium URER/MS-Adrar, CDER, Algeria,
Email: yaichi_mohammed@yahoo.fr
1. INTRODUCTION
The Saharan medium by its arid nature and the availability of solar immense can return the
application of the water pumping via photovoltaic (PV) solar pumping, like a very important and the most
attractive operation in the agricultural and industrial domains.
During the day, the speed of the motor asynchronous depends on the temperature and the quantity of
the solar radiation that is fallen on the photovoltaic panels to extract the maximum power. This latter, is
obtained by the proper adjustment of the inverter frequency (by increase or decreasing) instead of the MPPT
circuit (maximum power point tracker), inducing a total improvement of the efficiency of the system without
need adding a DC/DC converter to the chain. On the other hand, the flow daily and the efficiency of the
motor-pump for a total head (HMT) depend on the speed (related to the stator frequency) if we considered
that the number of stages is fixed (i.e Standard Centrifugal Pump, SCP) [1], [2]. For example, for a given
speed, the pump functions at an operation point. If irradiance decreases, that involves a reduction speed; the
system operating point is determined by the intersection point of the (Q, H) characteristics of the motor-
pump.
The temperature and the radiation have random variation nature. By applying the technique of
MPPT, the efficiency of the system rises whatever is the solar radiation value and the temperature of the
environment. This technique was applied by different means [3]. In this study, we execute this technique with
the use of artificial neural network (ANN). The application of ANN is recently growing in photovoltaic
system. A feedforward ANN implements nonlinear input-output mapping. A back propagation type

 ISSN: 2088-8694
IJPEDS Vol. 4, No. 2, June 2014 : 241 – 255
242
feedforward ANN is trained with the databases generated by a procedure that evaluated on a field-oriented
controlled induction motor drive, it explains in detail in §3. The back propagation training algorithm needs
only inputs and the desired output to adapt the weight [4].
2. SOLAR PUMPING SYSTEMS
The photovoltaic system is constituted of a self-piloted asynchronous motor operating a centrifugal
load. The unit is fed by solar cells through an inverter. Pumping without intermediate power storage enabled
us to have a simpler photovoltaic system, more reliable; maintenance-fee is less expensive than a system with
battery [5].
The system to be investigated is an immersed centrifugal motor-pump SP5A7 (nominal speed
2860rpm for an industrial frequency equalizes with 50Hz).
Figure 1 shows the block diagram of a closed loop field-oriented controlled (FOC) induction motor
drive incorporating the proposed ANN-based MPPT controller. The command voltage , is generated
from the frequency or speed command (irradiance G and temperature Ta dependents).
Figure 1. Field-oriented control of induction motor showing neural network based implementation
2.1. Photovoltaic array model
Some authors have proposed more sophisticated models that present better accuracy and serve for
different purposes. The basic equation from the theory of semiconductors that mathematically describes the
I–V characteristic of the PV array, equation (1) describes the single-diode model presented in Figure 2 [6],
[7].
Figure 2. Single-diode model of the theoretical PV cell and equivalent circuit of a practical PV device
including the series and parallel resistances
(1)
1 ∆ (2)
exp 1 (3)
(4)
When:
(5)
phI
DI
D
shI
s
sh
p
N
R
N
s
s
p
N
R
N V
I

IJPEDS ISSN: 2088-8694 
A Neural Network Based MPPT Technique Controller for Photovoltaic Pumping System (Mohammed Yaichi)
243
(6)
Where and are the photovoltaic (PV) and saturation currents, respectively, of the array and
cells connected in series per module. is the equivalent series resistance of the array and is the
equivalent parallel resistance. Modules connected in parallel increase the current and modules connected in
series provide greater output voltages. and are the number of modules connected in series and in
parallel respectively of the entire array. e is the electron charge 1.6 ∙ 10 C , is the Boltzmann constant
1.38 ∙ 10 J/K , Kelvin is the temperature of the p-n junction, and is the diode ideality constant.
(7)
Where is the bandgap energy of the semiconductor ≅ 1.12 for the polycrystalline Si at
25 and is the reference temperature.
(8)
Then (9)
∆ (10)
The cell temperature can be determined from ambient temperature and with the help of some
standard test information:
(11)
2.2. Inverter Model
Figure 3 shows a series-connection of single phase inverters. In this asymmetrical multilevel
topology, each H-bridge must be fed by an individual DC-voltage and is used to increase the multilevel
converter performances, without adding any complexity in the power circuit [8], [9].
d 1
K
'
d 1K
d 2
K
'
d 2K
a
2di
d 3K
'
d 3K
d 4K
'
d 4K









o
d 1v
d 2v
1di
d 1
U
d 2
U
Figure 3. Structure asymmetrical 5, 7 or 9 level cascaded inverter
7 51.9 , 2 , at standard test conditions (STC)
Each couple of switches , , ∈ 2 1,2 ∙∙ is controlled by a couple of switching
functions , such that:

 ISSN: 2088-8694
244
, ∈ 0,1
1
(12)
The conversion of the switch commutations in to a voltage is described by a conversion function,
such that:
⇒ 1,0,1 (13)
The output voltage of each cell is given by:
∙ , 0, (14)
With the input voltage of each cell.
The output voltage of a phase is given by:
(15)
With: s, s a, b or c ; d, d 1,2 or 3: represent the number of the phase (leg).
2 1 1
1 2 1
1 1 2
(16)
is the array input .
To improve the output voltage for such inverters, many different modulation strategies have been
developed. Among these strategies, the SVM (Space Vector Modulation). This technique provide the nearest
switching vectors sequence to the reference vector that is depends the modulation index r without involving
trigonometric functions and provide the additional advantages of superior harmonic quality. The generalized
algorithm being used to determine, for the hexagonal structure, the exact position of the vector of reference
(detection of nearest three vectors and duty cycles computation) was developed and studied in detail in [9].
2.3. Motor Asynchronous Model
The mathematical dynamic model of the asynchronous motor is described in [10], [11] by the
following equation:
(17)
With:

0
0
;
0 0 0
0 0 0
;
; 1 ; ;
, : axes corresponding to the asynchronous reference axes in Park model. , , , are d-
axis stator current, rotor flux and q-axis stator current, rotor flux respectively. , , , and are: stator
and rotor main inductances, resistances and mutual inductance respectively. σ dispersion factor, and
are the angular speed of the rotating magnetic and electric fields respectively.
The induction motor develops an electromagnetic torque expressed as follows:
(18)

245
is the number of pole pairs.
2.4. Centrifugal Pump Model
The variation of the pump’s speed can give us numerous charts Q-H. The use of a centrifugal pump
needs a preliminary study of the most important charts that characterize it, where efficiency will be optimum
with the total head and the speed envisaged by control the pumped water quantity to a desirable head. In
addition, they are related to dimensions, kinds and speed of the pump.
The chart of the water quantity-head Q-H (Figure 4) explains the different variations in the head of
pumping, according to water quantity which forms bent charts. The code then proceeds to the generation of
the entire curve. Successively, the program generates a series of characteristics for a fixed range of shaft
rotational speeds (generally 35–56 Hz), related to the same pump. It all is done referring to the affinity laws
at variable speed. The affinity laws are valid only under conditions of constant efficiency [1], [12]-[13].
Ω Ω⁄ ;
Ω Ω⁄ ; Ω Ω⁄ at (19)
Where , , , Ω and are the actual volumetric flowrate, head, the pump shaft input
power, rotational speed respectively and the pump efficiency, whereas , , , Ω and
are the same variables referred to the design speed.
Once the characteristic and power demand curves are defined, the pump efficiency may be
calculated as follows:
(20)
Where,
∙ ∙ (21)
is the power output in terms of pumped water [kW] (i.e. hydraulic output power),
∙ : 9.81 ∙ the constant of gravity, 10 ∙ water volumic mass the constant and
is the load constant.
Figure 4. Characteristic
Using the relations (19) and the pump datasheet, a set of curves giving the flow versus the head and
parameterized by speed can be then obtained, but this procedure is large and in order to use these results
during calculations this set of points should be fitted to obtain an algebraic equation, this is done by the use of
a two-variables third order polynomial function in order to achieve the best possible regression coefficient
[13]:
, ∙ ∙ ∙ ∙ ∙ ∙ ∙
∙ ∙ ∙ ∙ ∙ (22)

 ISSN: 2088-8694
246
For the determination of the pump operating point it is required to know both the pump and pipeline
characteristics. The piping system deals with the total head that must be overcome by the pump. The Q-H
characteristic of the pipe network is given as a function of the geodetic head and head losses (as function of
the flow-rate) [14]. Thus, it should at least equal the head corresponding the flow computed by the pump
flow-head equation. It comes:
(23)
The constant relates to the bead loss caused by fluid friction. : Geodetic head.
2.5. Vectorial Command Approach
The vector control is based on the field-oriented control (FOC) method. For the regulation of the
main variables (current, flux, speed) to their reference values, regulators were used of the type PI [10], [11].
So we have as interest to keep only the axis flux component, and that means to oblige the axis
flux component to be zero.
0 and (24)
The flux and torque will be reduced respectively then to:
(25)
I (26)
Therefore the and axis statoric voltages equations will be:
(27)
The compensation has as objective to uncouple the two and axis voltages and currents. Under
such conditions, the system becomes linear like in case of DC motor. Thus we have:
(28)
Where,
(29)
(30)
Are the compensation voltages.
The transfer functions of the plant for the controllers of the vector-controlled induction motor drives
can be derived as shown in [10], [11].
Speed, and current controllers:
(31)
Where is the Laplace operator.

247
Based on those equations and after some algebra (Imposition of the poles method) we obtain PI
parameters for speed and current (Table 1), were J is the rotor inertia moment, is the viscosity coefficient
of the induction motor. After is obtained upon calculation of the transfer function of a closed loop,
parameters and are identified from imposition of the poles method. A wise choice the and was
made by changing the value of .
Table 1. PI Regulator Coefficients
Coefficients
Current
Speed ω
3. LOCATION OF MAXIMUM POWER POINT
At the maximum power point we have:
∙
0 ⇒ 0 (32)
With the current described by equation (01), the voltage may be expressed as:
∆
(33)
Back substitution of the partial derivative of with respect to and using _ for
and _ for gives:
∆
1
_ ∆
_ _ 0 (34)
With _ and _ the voltage and current at maximum power.
Numerical method is applied to solve for _ using an initial guess given by:
, 1 ∆ or , ∆ (35)
We obtain a relation linear which can be expressed by:
_ 0.9325 0.0142 (36)
Once _ is found, _ may be calculated using equation (33) and thus the current and voltage
at the maximum power point is determined as a consequence the maximum power.
_ 0.94 1 ∆
. ∆
.
. ∆
1 (37)
At optimal regime, for calculation the speed reference, the assessment of power of asynchronous
motor is given by:

 ISSN: 2088-8694
248
_ 3 _ _ (38)
The relations governing the operation of the motor-pump [15].
(39)
∙ ∙ (40)
≅ ∙ (41)
(42)
By substituting equations and into Equation (41).
∙ (43)
∙ (44)
Thus the relations (41) then (38) becomes:
∙ ∙ (45)
_ ∙ (46)
By rearranging the Equation (45) gives:
0 (47)
With,
,
0.94 1 ∆
. ∆
.
. ∆
1
For each value of radiation and the temperature, i.e. the value of . By solving the Equation (47),
its roots, give the speed of instruction corresponding to the optimal operation of the generator PV. In
addition, this speed depends on the radiation and the ambient temperature. Results examples are summarized
in Figure 7 (red stars).
4. NEURAL NETWORK BASED MPPT
The ANN theory, in general, has been well-discussed in the literature [16]-[18], and a number of
authors [3], [4], [19] have described its operation in the PV system and for speed control of an induction
motor. Neural network architecture is specified through finding the appropriate solution for the non-linear
and complex systems or the random variable ones. Among its types, the more widespread, important and
useful back propagation network. The function and results of an artificial neural network are determined by
its architecture that has different kinds. The simpler architecture contains three layers as shown in Figure 6.
The input layer receives the extern data. The second layer, hidden layer, contains several hidden neurons
which receive data from the input layer and send them to the third layer, output layer. This later responds to
the system.

249
We can conclude unlimited neural network architectures. The more several hidden layers and
neurons in each layer are added; the more complex they become. The network is fully connected which a bias
signal is coupled to all neurons in the hidden layer through a weight . The realization of the back
propagation network is based on two main points: learning and knowledge. This research was applied by the
use of sigmoid function as an activation function in order to calculate the hidden layer output and the linear
function to calculate the output.
In our study of the effect of network architecture on learning, we have chosen architecture with one
hidden layer. We have changed the number of neurons in hidden layer; this architecture is summarized in
Table 2. By 5 neurons in hidden layer, the regression coefficient R-square (R2
) increases to 0.9999.
Table 2. Effect of number of hidden layers on learning
Number of neurons in hidden layer R2
3 0.9984
5 0.9999
8 0.9991
9 0.9989
11 0.9991
Figure 5 shows the proposed ANN topology, it use multilayer perceptron type network consists of
an input layer, one hidden layer and an output layer with sigmoidal type transfer function. The composite
network uses two neurons at the input, one for the radiation and the other for the ambient temperature, which
is normalized and then pulse neurons in the hidden layer. The output layer has only one neuron for the
control DC/AC frequency.
The procedure described in the previous section will be utilized to generate training data for ANN-
based MPPT.
Figure 5. Neural network topology (2-5-1) for speed generated
The outputs are calculated by the following equation:
∑ ∙ with: 1 ∙∙ 5 (48)
(49)
∑ ∙ (50)
The update of is done according to the rule of delta:

 ISSN: 2088-8694
250
Δ and Δ (51)
Where: , Δ and Δ (52)
With is the step of training and,
∙ , (53)
∙ , ∙ ∙ (54)
Algorithm architecture of the back propagation implies the following steps as shown in Figure 6.
The tests results of ANN-based MPPT at the five operating case (proposed learning examples) were
found that the output speed response of the motor-pump with analytical results is higher than the inputs
radiation and ambient temperature as shown in Figure 7 (a)-(e) which shows the output of ANN at five case
and shows the corresponding data with the analytical results. Also, the MPPT method was validated for the
case 37 and compared with the data, they were found to correlate very well (Test of generalization in
Figure 8). This means that the network can be operated at any conditions climatic.
Figure 6. Architecture of the back propagation
1: 1 ;∙∙∙∙∙∙∙∙∙;
1 ;∙∙∙∙;
0.004, 0.01,
1, :
2: 1 ;∙∙∙∙∙∙∙∙∙;
And the example given one by one
To initialize the weights of the network by
random values 0; 2 5 2 ; 3 1 5
Begin
Yes
No

∆ . 51
1
2
End
, , , ,
calculate the output

Of
the synaptic
weights anew
0, 1
. 53 , . 54
1

251
(a) (b)
(c) (d)
(e)
Figure 7. Test of validation
Figure 8. Test of generalization

 ISSN: 2088-8694
252
5. RESULTS AND DISCUSSION
The Model field-oriented control using MPPT technique is developed and studied in detail in
another work [1].
Figure 9. The block diagram of a field-oriented control induction motor coupled to a centrifugal pump,
fed by solar cells through an inverter
Figure 10. Behavior of the pump by considering radiation variations
Figure 9 gives a diagrammatic sight of a drive at variable speed using the principle of the field-
oriented control. For the simulation results used in this paper, the numerical values for the PV array and the
pump SP5A7 are summarized in Table 3-4. The numerical values of the simulation model are obtained either
by measurements or identification from laboratory experiments.
In order to verify the proposed model, digital simulations were carried out for four insulation level.
At the beginning, radiation is worth G=540W/m2
, then radiation is brought to G=1200W/m2
in 1s then in
G=940W/m2
in 3s and finally, radiation reached G=150W/m2
in 6s.

253
Table 3. PV array specifications [7]
isofotón Array; : 1.224 0.545
9 , : 11.2%, 74.4%
Cell type
Peak power 75 10%
Optimal voltage 17.3
Optimal current 4.34
Open circuit Voltage 21.6
Short Circuit current 4.67
Cell number 36
Temperature Coef iciency of %/ 0.02
Reference ambient temperature 20
Diode ideality constant 1.2
Photo current 4.6756
Equivalent series resistance of the panel Ω 0.2402
Equivalent parallel resistance Ω 199.4843
0.0047
1.9976 ∙ 10
3.7042
PV array: Number of modules connected in series and in parallel 9 Ns × 1 Np
Table 4. Motor-pump specifications [1, 20]
Subsystem Specification
Pump load Constante k N ∙ m ∙ s/rad 1.524 ∙ 10
Asynchronous motor
Type GRUNDFOS MS 402
RMS motor voltage V Volt 65∆
Maximal Courrent I A 8.5
Speed Ω rms 2860
Power P kW 0.55
Motor frequency fre Hz 50
Stator resistance per phase R Ω 0.6667
Motor power factor cos φ 0.8
Rotot resistance R Ω 1.1965
Stator leakage inductance L H 0.0895
Rotor leakage inductance L H 0.0895
Mutual Inductance M H 0.087
Pair pole number p 1
Inertie Torque J kg ∙ m 0.003
Friction coef icient fr N ∙ m ∙ s/rad 2.905 ∙ 10
Ef iciency η % 0.78
Flux Q Wb 0.28 0.3
The response of the system using MPPT controller obtained by simulation is shown in Figure 10.
All test results show that the robust tracking speed of induction motor drive with the proposed MPPT control
strategy is very effective in tracking the selected tracks at all times, it is observed that, in every case. This
controller provides a very useful tool for ensuring optimal efficiency, which essentially keeps its state
optimal. On the other hand, the results show the output voltage responses of the PV array to the change of
radiation and temperature. For example, at 3s the radiation stepping stepping-down from 1200 to 940W/m2
,
the PV array voltage stepping-up from 118 to 122volt. This system is able to adapt the operating point
according to variations of the radiation and the temperature (random variation).
The PI controller brings the measured speed to the desired value smoothly and without the
overshoot. We observed the flow estimation and also the resistive torque imposed by the quantity of pumped
water. The electromagnetic torque compensates for instantaneously this request of the resistive torque.
The figures show that the flow and the couple are decoupled. In addition, flow is constant in permanent mode
0.28 . The statoric voltage by frequency ratio is always kept constant with a / value of 1.3
during operation, the motor is thus, managed by the / constant law, which shows the effectiveness of
the loop of speed regulation.
6. CONCLUSION
An optimal operation of a photovoltaic pumping system based on an induction motor was described.
A neural network based MPPT has been investigated that operates very well in the PV pumping system. The
proposed structure allows at the same time the field oriented control and the maximum power tracking of the
photovoltaic array.

 ISSN: 2088-8694
254
This control has given very good results; where it assured a total decoupling between the system
parameters. Moreover, it assured a good tracking of the speed with a minimum response time. The regression
coefficient R-square (R2
) arises in high values through the neural network with one hidden layer, with 5
neurons. The proposed algorithm in neural networks was simulated; the simulation results show that this
system is able to adapt the maximum operating point according to variations in external disturbances. In
addition, the proposed approach does not need two converters; it uses only a DC/DC converter. Also,
increase of both the daily pumped quantity and array PV efficiency are reached by the proposed approach.
REFERENCES
[1] M Yaichi, MK Fellah. Centrifugal Motor-Pump System Model intended for the Photovoltaic Pumping applications.
Przegląd Elektrotechniczny. 2013; 89: 100-105.
[2] D Fiaschi, R Graniglia, G Manfrida. Improving the effectiveness of solar pumping systems by using modular
centrifugal pumps with variable rotational speed. Solar Energy. 2005; 79: 234–244.
[3] A Saadi, A Moussi. Neural network use in the MPPT of photovoltaic pumping system. Review of renewable
energies ICPWE. 2003: 39-45.
[4] S Belakehal, H Benalla, A Betounsi. Performance investigation of a solar pumping system using the neural networks
MPPT approach. The International Conference on Electrical Engineering and its Applications ICEEA'08. Djillali
Liabes University of Sidi-Bel-Abbes, Algeria. 2008.
[5] J Royer, T Djiako, E Schiller, BS Sy. Pompage Photovoltaïque. IEPF, University of Ottawa, EIER, CREPA. ISBN
2-89481-006-7, Canada, 1998.
[6] MG Villalva, JR Gazoli, ER.Filho. Comprehensive Approach to Modeling and Simulation of Photovoltaic Arrays.
IEEE transactions on power electronics. 2009; 24(5).
[7] M Yaichi, A Neçaibia, M Sadok. Etude, Caractérisation et Estimation des Paramètres du Module Photovoltaïque
Isofotón-75W. 2nd
International Conference on Energy and Sustainable Development. University of Adrar and
URER/MS, Adrar. 2013.
[8] J Song-Manguelle, A Rufer. Asymmetrical Multilevel Inverter for large induction machine drives. International
conference for Electrical drives and Power Electronics, EDPE’01, Slovakia. 2001.
[9] M Yaichi, MK Fellah, A Neçaibia, A Mammeri. Structure des Onduleurs Multiniveaux Asymétriques Applications
aux Systèmes Photovoltaïques. International Conference on Energy and Sustainable Development. University of
Adrar, Algeri. 2011.
[10] A Meroufel. Commande découplée d’une machine asynchrone sans capteur mécanique. Thesis of doctorate,
University of Sidi-Bel-Abbes. 2004.
[11] M Makhlouf, F Messai, H Benalla. Vectorial command of induction motor pumping system supplied by a
photovoltaic generator. Journal of Electrical Engineering. 2011; 62(1).
[12] B Ben Ghanem. Performance of submersible PV water pumping systems in Tunisia. Energy for Sustainable
Development. 2012. http://dx.doi.org/10.1016/j.esd.2012.10.003.
[13] T Martiré, C Glaize, C Joubert, B Rouvière. A simplified but accurate prevision method for along the sun PV
pumping systems. Solar Energy. 2008; 82: 1009-1020.
[14] A BETKA. Perspectives for the sake of photovoltaic pumping development in the south. Thesis of doctorate,
University of Batna. 2006.
[15] MN Mansouri, N Ghanmi, MF Mimouni. Commande et analyse des performances d’une station de pompage
photovoltaïque fonctionnant en régime optimal. Revue des énergies renouvelables. 2008 ; 11(1).
[16] RS Dhekekar, NV Srikanth. ANN Controlled VSC STATCOM with Harmonic Reduction for VAR Compensation.
International Journal of Power Electronics and Drive System (IJPEDS). 2012; 2(1): 76~84.
[17] Laxmi Devi Sahu, Satya Prakash Dubey. ANN based Hybrid Active Power Filter for Harmonics Elimination with
Distorted Mains. International Journal of Power Electronics and Drive System (IJPEDS). 2012; 2(3): 241~248.
[18] Eric Davalo, P Naim. Des réseaux de neurones. Edition EYROLLE. 1993.
[19] Ashutosh Mishra, Prashant Choudhary. Artificial Neural Network Based Controller for Speed Control of an
Induction Motor using Indirect Vector Control Method. International Journal of Power Electronics and Drive
System (IJPEDS). 2012; 2(4): 402~408.
[20] M Yaichi, A Mehdaoui, A Mammeri. Research activity report. Research Unit in Renewable Energies in the Saharan
Medium URER/MS-Adrar, Algeria. 2012.

255
BIOGRAPHIES OF AUTHORS
Mohammed Yaichi was born on 1980 in Adrar, Algeria. He received the engineer degree in
electrical engineering from the the University of Bechar, Bechar, Algeria, in 2003. And the
magister degree from Djillali Liabes University, Sidi-Bel-Abbes, Algeria in 2006. He is
currently working toward the doctorate degree in the Power Electronics, and the Photovoltaic
Pumping System, Djillali Liabes University of Sidi-Bel-Abbes, Algeria. Since 2009, he is with
the Photovoltaic Pumping Team, Research Unit in Renewable Energies in The Saharan Medium
(URER/MS) Adrar, Algeria. His research interests include a study on performance improvement
of a stand-alone photovoltaic pumping systeme, variable-speed AC motor drives, and different
multilevel inverter circuit topologies thus its technique of control PWM.
Mohammed-Karim Fellah was born in Oran, Algeria, in 1963. He received the Eng. degree in
Electrical Engineering from University of Sciences and Technology (USTO), Oran, Algeria, in
1986, and The Ph.D. degree from National Polytechnic Institute of Lorraine (Nancy, France) in
1991. Since 1992, he is Professor at the University of Sidi Bel-Abbes (Algeria) and Member of
the Intelligent Control and Electrical Power. His current research interest includes Power
Electronics, HVDC links, and Drives.
Mammeri Abdelkrim was born on 1979 in Bechar, Algeria. He received the engineer degree in
Hydraulic from University of Sciences and Technology (USTO), Oran, Algeria, in 2006, and
now studying for master degree. Since 06/11/2006, he is with the Photovoltaic Pumping Team,
Research Unit in Renewable Energies in The Saharan Medium (URER/MS) Adrar, Algeria. His
research interests include a study on performance improvement of a stand-alone photovoltaic
pumping system.

A Neural Network Based MPPT Technique Controller for Photovoltaic Pumping System

More Related Content

What's hot (20)

Similar to A Neural Network Based MPPT Technique Controller for Photovoltaic Pumping System (20)

More from IJPEDS-IAES (20)

Recently uploaded (20)

A Neural Network Based MPPT Technique Controller for Photovoltaic Pumping System