Optimization photovoltaic pumping system

based BLDC using

Fuzzy logic MPPT control

Nabil Taibi
Mahdi Ouada* ,Mohamed Salah Meridjet Department of mechanical,university 20 august55
Department of Electromechanical,University Badji Skikda. Algeria
Mokhtar Annaba.,(UBMA),Algeria nabiUalbi@yahoo.fr
ouadamehdi@gmail.com,meridsa@yahoo.fr

Abstract- In this paper, we present an approach for efficiency of PY installations by giving the maximum power
modeling , control and optimization of a photovoltaic point tracking (MPPT) algorithm the ability to track
pumping system which contains a pv generator , DC-DC effectively the maximum power point of a photovoltaic system
boost converter, MPPT controller, DC-AC inverter and a
under variable irradiation conditions.
BLDC motor. We use an intelligent control method for
In this paper, an intelligent control technique using fuzzy
searching the maximum power point (MPP). This method uses
logic control is associated to an MPPT controller in order to
a fuzzy logic controller applied to a drive a DC-DC converter to
an optimal operating point using PV panel's measured variables.
improve energy conversion efficiency of a PY standalone water
The PWM signals are generated by the interaction of the motor
pumping system.
speed closed-loop system and the current hysteresis. The
motor reference current is compared with the motor speed II. DESIGN OF PUMPING SYSTEM
feedback signal.
The following figure describes elements constituting the
The considered model has been implemented in Matlab Isimulink
environment. The results show the effectiveness of the
water pumping system:
proposed method to increase the performance of the water
pumping system.

Keywords- Pumping system; photovoltaic; MPPT; BLDC;

Fuzzy logic

INTRODUCTION

Solar energy is the most important,most effective and least

expensive over other renewable energy source.
Solar energy conversion can be achieved using either by
thermal or photovoltaic effects. Many applications can use
Fig. I Block diagram of the BLOC water pumping system
such renewable source of energy such as: water pumping, air
conditioning, light sources electric vehicles, refrigeration
systems. Standalone photovoltaic (PY) systems [2] are widely A. P V array
used in military and space applications [1,2]. This is the most important element since it provides the
electric power needed from the water pumping system; we
The evolution of life has been possible thanks to the
chose the PY panel Kyocera 200 GT [3].
presence of water. Using photovoltaic generators to operate
the water pumps is now a technology in development that is
characterized by gradual decrease in cost. Since the first B. DC-DC boost converter
installation of photovoltaic pumping system in the late We used a boost DC-DC converter in this system Fig .2. It
seventies, these systems provide human domestic needs, consists of a boost inductor, controlled switch, diode and a
livestock and irrigation water in rural areas, and have filtering capacitor [4].
gained considerable acceptance in terms of reliability and
performance and today they are considered to be the most
C. Search a/the maximum power point tracking (MPPT)
significant applications of photovoltaic energy conversion.
On the other hand, fuzzy logic is an intelligent control Many method have been proposed to track the maximum
method that has been used recently for improving the power point such as Perturb & Observe (P&O) method,

978-1-4673-6374-7/13/$31.00 ©2013 IEEE

Incremental Conductance (IncCond) algorithm, Parasitic Y Fuzzification.
Capacitance, Voltage Based Peak Power Tracking, Current
Y Inference.
Based peak power Tracking [5,6]. Perturb and observe have
many advantages like, simplicity, easy to implement and less Y Defuzzification.
acquired parameters, have simple structure [5, 6, 7], its main
disadvantage being the oscillation of the operating point around
the maximum power point [5, 6]. To solve this problem we
1) Fuzzification
propose using a fuzzy logic based control algorithm . The variable E and CE are expressed as follows:
iL
� Pph (k) Pph (k -1)

Ivo '
L ... _ -

E(k) (1)
i ph (k) - i ph (k 1)
t
+ -
VL

I
iC
is
C
S
dECk) = E(k) - E(k -1) (2)

where Pph(k) and Iph(k) are the power and current of the
PV array, respectively.

Therefore,E(k) is zero at the maxImum power point of a

Fig.2 Boost converter PV array.
Fuzzy logic was introduced in 19 65 with the work of The input E(k) shows if the operating point of the load is
L .Zadeh [12]. He has formalized the representation and the situated to the left or right of the maximum power point of the
processing of knowledge with imprecise or approximate PV curve. If this value is positive, then the operating point is
variables to solve high complexity systems. to the left of MPP,otherwise,the operating point is to the right
The objective of the fuzzy logic control method is to track of MPP. The second input variable dE(k) shows the direction
the maximum power point of a photovoltaic generator for and allows us to estimate the speed of convergence to the
different irradiations. The maximum power that corresponds point MPP operating point.
to the optimal operating point is determined for different level Knowing these two inputs we can decide what will be the
of irradiation. change we must impose to the duty cycle given to a boost
chopper. To increase the voltage operating point, D must be
increased and vice versa.
D. MPPTfuzzy controller design

MPPT fuzzy controller was designed and simulated using Input output
the Simulink Fuzzy Logic Simulink Toolbox represented in
(Fig. 3): E
Fuzzification Inference Defuzzification dD
dE
II �I Editor. ElN 1""",1.@�
File Edit VifjN
numerical values fuzzy domain i numerical values

lu� Fig. 4 Basic structure of the fuzzy controller MPPT

�
ELN

2) Inference:
•

{rnarrdanij

luY de
0
The used inference method is "Mamdani", that is the most
commonly used inference method. It uses the MIN operation
1 FISN""" ELN mType: mOOldari

1
for the "AND" operator and MAX for the "OR". Inference
rules can make the right decision for output to dD from the
And method
rril
3 CurrenfVariabie
values of the inputs E and dE. In our work we chose the rules
max
� Ie I presented in the (Table 1): variable E and dE are expressed as
Or melhOO �me

�
Implication
rril
T)1le inp.ll
follows:
Range 1·3591
Aggregation
max F.i TABLE I
Defuzzificalion
centroKt
9 I! He\>
I "Cbs�11 INFERENCE TABLE OF THE FUZZY CONTROLLER
1 System 'UN": 2 ""�s, 1 0JIJ>.l, and 25 rOe,
1 ET dE NG NP ZE PG PP
NG PG PG PG PG PG
NP PG PP PP PP ZE
Fig. 3 General diagram of fuzzy controller MPPT MATLAB ZE PP PP ZE NP NP
PG NP NP NP NP NP
A Fuzzy Logic Controller (FLC) consists of three blocks PP NG NP NP NP ZE
(FigA):
 3) Defuzzification The expression of electromagnetic torque produced by the
The defuzzification method used in this work is the FLC stator windings is
center of gravity. dw
J-=T,e -TL - Bw (7)
dt
Where:
E. Inverter Te: magnetic torque (N.m)
Tr.: load torque (N.m)
The inverter provides three-phase system voltages
B: damping factor (N.m.s/rad) ;
variable in amplitude and frequency to operate with
w: motor speed (rad / s)
variable loads and frequency (from 0.1 up to 1 time the
J: motor moment of inertia (kg/m).

r j �r jr j
rated frequency) [8]. The current is modulated sinusoidally to
obtain a high efficiency. The pulse frequency is maximal G. Centrifugal pump model
2kHz. The phase voltage can be expressed as follows [8]:
The centrifugal pump applies a load torque proportional
van 2 -1 -l C1 to the square of the rotational speed of the motor[8].
Vbn = -1 2 -1 C2 (3) Centrifugal pump is the most commonly employed type of
Vcn -1 -1 2 C3 pumps [13, 14], it has a relatively high efficiency, and it is
capable of pumping a high volume of water [[4].
F. BLDC The performances (Q " H' and P ') are given in terms of the
The Brushless DC (BLDC) motors present many speed using the following relationships:

Q '= Q (jj'
advantage like having better mechanical characteristics, high
(9)
efficiency,high dynamic response,small size construction [9]. OJ
[n the last decade, brushless dc motors have begun to replace
brushed dc motors and induction motors for small scale H'= H(oJ')2 (10)
OJ
pumping applications [[[].
P'= p((jj')3 (11)
OJ
To simulate the BLOC, we have selected the three phases
model. It is defined in [10] by the following equations of stator III. ANALYSIS OF RESULTS
and rotor voltages:
Fig. 5 show that output voltage DC/DC Boost converter we
note that the system follows the
variations of irradiation.
During this time there was an increase 0-0.57 of the voltage
versus time is the regime transition. At t 0.57 is the steady =

state affected the voltage remains substantially constant up to

t [.5 and then decreases according to the irradiation.
=

Where:
1000 -...... -"""----'-....---.
... .,

ua,ub,uc: Stator winding phase voltage (V) 900

Ra, Rh, Rc : Stator winding resistance (Q); 800

ia, ib,i c : Stator winding phase current (A);
� 700
ea, eh' ec: Stator winding back EMF (V); � I I
-;:; 6oo� � � �
I � � �I � � � � �
I
-------------------
I
I
La' Lb, Lc : Self-inductance (H); o
-- I I I
� 500
Mab,Mac,Mch: Mutual-inductor (H); "E
400

300

200

0.2 OA 0.6 0.8 11 A lB 1.8

times(s)
Fig. 5 Variation of irradiance levels to the pv array while keeping the ambiant
temperature constant
 45 ,-----�--�-- 2.5 -------�-

2
40
1.5- ------ --
35

30
is 0.5 - - - - .i __ .i -
�
'"
25 c:
�
:J 0
� 1 u
20 -------1-------
o 5 - 0.5 - - - -
> 1U
1 U)
15 ---------------
1 -1- ----- - - -

_______ J _______
1 -1.5

-2- ------- -

-
-2.5c--------- - -----------'--------- - ---------
0 0.5 1.5 2
0 .5 1.5 2
time (5) time(s)

Fig. 6 boost converter output voltage Fig. 9 Stator current phase a (A)

40
500 ,-----�---�-------

30
�- -T--------

I
400 I: 20

i :1

J
10

E 300 ---- � -- I ------- I ---------------- '>

E / '"
(j)
0

"'-
-10
CO>

� 200 -
j -------

-20

-30
100
-40
0 0 .5 1.5 2
time(s)
o
o 0.5 1.5
Fig. 10 Stator back EMF e_a (V)
time(s)
6
Fig. 7 Rotor speed Wm (rad/s)
"J1-IJUU-.lUU,HJ'i l'-t�I"t�
: :
--------�-- 5 - - - - - - - - I- - - - - - - - - - - '�"" - - - 1- - ------
' 4.1�� ,
r I
f ",,-,�" ..
Jtuuf n
�.. ..I,�TJ. f.
3.5
4 1

_______________ L ______ _

1
---------------�-------
o

-1
o 0.5 1.5 2
time(s)
o ------� ----------------------

Fig.11 Pump flow

-0.5 ________�______________________________.J
o 0.5 1.5 2
The curves of figures 7, 8, 9, 10, II shows that there is an
time(s)
almost linear increase. The speed increase until at the time
Fig. 8 Electromagnetic torque Ie (N*m) instant t 0.57s and it remains constant until t 1.5 then fall
= =

because of reduced sunlight. In starting regime, There is a

strongly pulsating torque and in steady state the
electromagnetic torque follows the resistant torque of pump.
 IV. CONCLUSION
Photovoltaic energy conversion is one of the
alternatives of renewable energy sources,since the advent of
major applications in the world , and has shown its
flexibility and its ability to operate in several
environments. This is an interesting solution to conventional
means of production.
To ensure the operation of a photovoltaic generator at its
maximum power point, MPPT controllers are often used.
These controllers are intended to effectively track the MPP
and thus minimize the error between the operating power and
the maximum power which is the reference variable as a
function of the load and the climatic conditions.
The optimal choice of the method of tracking depends
mainly on the given specifications.
In this article we were interested in a MPPT fuzzy logic
control that gave in general good perfonnances. The studied
method can be redesigned quickly and present a robustness to
variations of the solar irradiation.

V. REFERENCES
[1] Roger Messenger, Gerard G Ventre, Jerry Ventre " Photovoltaic
Systems Engineering "',2nd edition ,CRC PRESS, 2005.
[2] Mohamad A. S. Masoum, Seyed Mahdi Mousavi Badejani, Ewald F.
Fuchs, " Microprocessor- Controlled New Class of Optimal Battery
Chargers for Photovoltaic Applications,"' IEEE Transactions on energy
conversion, vol. 19, no.3, September2004 .
[3] "KC200 GT data sheet,"' high efficiency multicrystal photovoltaic
module
[4] Dariusz Czarkowski, "Power electronics handbook "' ,Florida ,US,
academic press , USA, 200 l,pp211 -223.
[5] Satarupa Bal, Chitti Babu, " Comparative Study between P&O and
Current Compensation Method for MPPT of PV Energy
System,IEEE (SCES) India,16-18 March2012 .
[6] Vikrant.A.Chaudhari, "Automatic peak power tracker for solar pv
modules using dspace software'". Eng Thesis , maulana azad national
institute of technology, India,luly -2005.
[ 7] Hairul Nissah Zainudin, Saad Mekhilef "Comparison Study of
Maximum Power Point Tracker Techniques for PV Systems'"
Proceedings of the 14 th International Middle East Power Systems
Conference (MEPCON'IO), Cairo University, Egypt, December 19-21 ,
2010 .
[8] N. HAMROUNI, M. JRAID!, A. CHERIF, A. DHOUIB,
"Measurements and Simulation of a PV Pumping Systems Parameters
Using MPPT and PWM Control Strategies, IEEE MELECON 2006
IEEE MELECON 2006, May 16-19, Benalmadena (Malaga), Spain pp
885- 888 .
[9] <;:. Gencer , M. Gedikpinar "Modeling and Simulation of BLDC Using
MATLAB/SIMULINK, Jornal of Applied Scineces vol 6 no 3 ,2006.
[10] Yiming. Li, Jun .Rong,"The Study of New Modeling Method for
Permanent Magnet Brushless DC motor"' International Conference on
Electronic & Mechanical Engineering and Information 12 -14 August
china, 2011 , pp -4 7134 716.
[11] M. A. Elgendy, B. Zahawi, D. 1. Atkinson, D. Giaouris "Dynamic
Behaviour of DC Motor- Based Photovoltaic Pumping Systems under
Searching MPPT Algorithms "' POWERENG 2009, Lisbon, Portugal,
March18 -20 ,2009 pp413-418 .
[12] L.A. Zadeh, "Fuzzy sets'", Information and Control, vol 8 , pp. 338 -
353,1965.
[13] S.G. Malla, C.N. Bhende , S. Mishra "Photovoltaic based Water
Pumping System" Energy, Automation, and Signal (ICEAS), 2011
International Conference on28 -30 Dec.2011 india.
[14] Thesis- Akihiro Oi,'"Design and simulation of photovoltaic water
pumping system", California Polytechnic State University ,USA
September2005