2017 International Conference on Advanced Mechatronics, Intelligent Manufacture, and Industrial Automation (ICAMIMIA)

Comparison Method of MPSO, FPA, and GWO

Algorithm in MPPT Sepic Converter Under
Dynamic Partial Shading Condition
Farid Dwi Murdianto Moh. Zaenal Efendi
Electrical Engineering Department Electrical Engineering Department
Electronics Engineering Polytechnic Institute of Surabaya Electronics Engineering Polytechnic Institute of Surabaya
[email protected] [email protected]

Rangga Eka Setiawan Alfis Syah Laili Hermawan

Electrical Engineering Department Electrical Engineering Department
Electronics Engineering Polytechnic Institute of Surabaya Electronics Engineering Polytechnic Institute of Surabaya
[email protected] [email protected]

Abstract --- Solar energy can be utilized through emergence of more than one peak point of maximum power
photovoltaic (PV) into electrical energy. The output power of is GMPP (Global Maximum Power Point) and LMPP (Local
PV can reach maximum point when the irradiation of the Maximum Power Point) [2]. If trapped in LMPP then the
sun is exposed to the surface without any object being power obtained is not the maximum power of PV. Under
obstructed. Unfortunately, In the actual implementation of
normal circumstances without unobstructed objects, the
irradiation received by PV it is often blocked by a moving
object or a shadow of a non-moving object. This problem characteristic PV curve will bring up a peak of maximum
resulted in the emergence of more than one maximum power power point [3]. To overcome the problem is not ideal due to
point on the P-V characteristic curve. The condition can’t the emergence of LMPP then comes the partial shading
be solved by using the usual MPPT Algorithm so that MPPT algorithm. Conventional partial shading algorithms such as
Partial shading algorithm is required. Partial shading Perturb and Observe (P&O) [9,11] and Incremental
algorithm has been developed from conventional methods to Conductance [10] to artificial intelligence methods such as
artificial intelligence methods. Hence, in this paper proposed Particle Swarm Optimization (PSO) [17], Flower Pollination
comparison method of MPSO, FPA, and GWO in MPPT Algorithm (FPA), Grey Wolf Optimization [4,23]. One
SEPIC converter under dynamic partial shading condition.
such method is the PSO developed into Modified Particle
The level of accuracy of all three methods is validated
through simulation. The simulation results show that the Swarm Optimization (MPSO) but it takes a little longer in
three methods are able to overcome the problem of partial computation [5]. In this paper, we will simulate the
shading very well with a high level of accuracy. The performance of the three proposed methods of MPSO, FPA,
simulation results also show that the FPA method has and GWO to determine the switching process in DC-DC
advantages compared to other proposed methods. The Converter [19-20] specially Sepic converter [18]
simulation results also show that the MPSO method has
more advantages compared to other proposed methods in II. SOLAR PV MODULES UNDER PARTIAL
tracking accuracy and the FPA method is faster to reach SHADING CONDITIONS
convergence situation by other proposed method.
The solar PV module has a function to convert the
intensity of sunlight into electrical energy. Output power
Keywords --- Maximum Power Point Tracking (MPPT);
Partial Shading; Grey Wolf Optimization (GWO); Modified from PV influenced by three factors there are irradiance,
Particle Swarm Optimisation (MPSO); Flower Pollination temperature, and shading. Generally, each solar PV module
Algorithm (FPA); Sepic converter; Photovoltaic (PV) has 36 or 72 cells. Each cell has an equivalent circuit
consisting of diode in parallel with a current source as shown
I. INTRODUCTION in Fig. 1 [6]. The mathematical are described in equations
Solar energy is a very large energy that can be (1) and (2).
utilized from the morning until the afternoon. Solar energy
in addition to large energies that can be utilized is also
environmentally friendly or pollution-free [1]. Solar energy
can be absorbed and utilized by Photovoltaic (PV) to be a
source of electrical energy. PV has a problem when the solar
irradiation received by the surface can’t be absorbed
optimally because it is blocked by an object. The intensity of
sunlight that is partially blocked by object results in the Fig. 1. Equivalent circuit of solar PV modules

978-1-5386-2729-7/17/$31.00 ©2017 IEEE 315

 V + IRs [18] so this converter suitable using for MPPT. The circuit
I = Isc – Io ( e(q(V+IRs))/kT-1) - ( ) (1)
Rp of Sepic converter is shown in Fig. 2. and the parameters of
kT Isc it can be calculated using equation (3) until equation (7).
Voc = ln ( +1) (2)
q Io

Where:
I : the cell’s current (A);
V : the cell’s voltage (V);
Isc : the cell short-circuit current (A);
Voc : the cell open-circuit voltage (V);
Io : the reverse saturation current (A); Fig. 2. Sepic converter circuit
q : the electron charge (C);
k : Boltzmann’s constant; D
T : the module temperature (K); Vout = Vin (3)
1− D
Rs and Rp : the resistances used to consider the impact of Vin × D
shading and losses. L1 = (4)
ΔiL1 × fs
Output power PV was decreased when PV under Vin × D
L2 = (5)
partial shading. The output power generated by the shaded ΔiL 2 × fs
cells is lower than unshaded cells. If this condition happens Vout × D
for a long duration, PV may can’t work properly because in C1 = (6)
R × ΔVo × fs
PV surface cause hot spot [7]. In this paper using three PV Vout × D
modules with the total output power is 150 Watt. The C2 = (7)
R × ΔVo × fs
specification of PV module proposed in Table I.

TABLE I. SOLAR PV MODULES SPESIFICATION [14] Switching frequency of converter using 40 KHz and
POLYCRYSTALLINE
by using high frequency can reduce the value of inductance
SP 50 W (P) 12 V L1 and L2. The output signal from Sepic converter will be
Silicon Solar PV Module smoother if using high frequency than low frequency [13].
The calculation of the Sepic converter is almost same as the
Pmax (Maximum power) 50 W
Buck-Boost converter [12]. The parameters of Sepic
Voc (Open circuit voltage) 21,1 V converter are shown in Table II.
Isc (Short circuit current) 4,23 A
TABLE II. SEPIC SPECIFICATION AND DESIGN PARAMETERS
Vmp (Voltage @ Pmax) 16,5 V
Parameter Value Unit
Imp (Current @ Pmax) 3,34 A
Input Voltage (Vin) 49.5 V
Max System V 1000 V
775 x 680 x Switching Frequency (fs) 40 KHz
Dimension mm
28
AM 1,5 1000 W/m2 Output Voltage (Vo) 14.4 V
Test Condition
25 oC
Current Ripple (∆IL) 20% A
In partial shading condition, the Power-Voltage (P-V)
Voltage Ripple (∆Vo) 0.1% V
characteristic curve has more maximum power point. Some
patterns of shading have been set to get more different P-V Inductor 1 (L1) 0.4168 mH
characteristic curve.
Inductor 2 (L2) 0.4168 mH
III. MODELING OF SEPIC CONVERTER
Sepic converter is a type of non-isolated DC-DC Capacitor 1 (C1) 3.587 mF
converter that developed from the Buck-Boost converter.
The function of it is to lowering and increasing the output Capacitor 2 (C2) 3.587 mF
voltage. In this paper, the converter will receive the input
voltage from solar PV module. The difference is polarity
output voltage of Sepic converter same with input, no
polarity reversal [12, 15, 16] unfortunately Buck-Boost is
reversal. The other features, it has lower current ripple input

316
 IV. MPPT PARTIAL SHADING ALGORITHMS b) Grey Wolf Optimization
Many algorithms were developed for tracking Grey Wolf Optimization is one of metaheuristic
maximum power point in P-V characteristic curve. However, algorithm which inspired the leadership hierarchy and
only Artificial Intelligent (AI) methods are capable to hunting mechanism of grey wolf. Four types of grey wolf
resolve partial shading conditions. In this paper, using three such as alpha (α), beta (β), delta (δ), and omega (ω) are
AI methods there are FPA, GWO, and MPSO. The pattern of proposed for simulating the leadership hierarchy. The best
shading in this paper already fixed are shown in Fig. 3. solution was named by alpha (a). The second and third best
solution was named by beta (β) and delta (δ). The left
solution is named by omega (ω) [4].
The formula of GWO method [23] can be defined
using the equations (11) until (14).

D =| C. X p (t ) − X p (t ) | (11)
A = 2a.r1 − a (12)
C = 2.r2 (13)
X (t + 1) = X p (t ) − A.D (14)

Where :
t : the iteration number;
D, A, C : coefficient vectors;
Xp : the position of the prey;
X : the position of grey wolf;
a : variable which linearly decreased from 2 to 0;
r1, r2 : random variable, the value is [0 - 1]
Fig. 3. Shading pattern of solar PV modules
c) Modified Particle Swarm Optimization
a) Flower Pollination Algorithm Modified Particle Swarm Optimization (MPSO) is an
Flower Pollination Algorithm (FPA) is an optimization method under the category of Nature Inspired
optimization method that inspired from pollination process Algorithm. This method is adapted from bird folk’s
in flowers. The flower can produce a new species from two behavior. The swarm of a particle suggests a solution for the
processes there are cross-pollination or self-pollination [21]. optimization process. The position of a particle is affected by
The formula of FPA method [22] can be defined the agent’s position [7].
The formula of MPSO method [8] can be defined
using the equations (8) until (10).
using the equations (15) and (16).

xi t +1 = xi t + L( xi t − g ) (8)
vi (k + 1) = wvi (k ) + c1r1( Pbesti − xi (k ) + c2r2 (Gbest − xi (k )) (15)
πλ
λΓ(λ ) sin( ) xi (k + 1) = xi ( k ) + vi (k + 1) (16)
2 1
L≈ ( s >> s 0 > 0) (9) i = 1,2,3,..., n
π s1+λ
xi t +1 = xi t + ∈ ( x j t − x k t ) (10) Where:
xi : the position of particle i;
Where: vi : the velocity of particle i;
xi : the pollen of flower i; k : the iteration number;
xk : the pollen of another flower in the same species; w : the inertia weight, the value is [0 - 0.5];
t : the iteration number; r1, r2 : random variables, the value is [0 - 1];
g : the optimal solution of all pollens; c1, c2 : the cognitive and social coefficient, respectively, the
L : levy flights; value is [0.8];
s : steps; Pbest i : used to store the best position that the ith particle
has found so far;
Γ(λ ) : standard gamma function;
Gbest : used to store the best position of all the particles.
∈ : random variables, the value is [0 - 1];

317
 The algorithms will be applied to the real condition
by using simulation. There are three solar PV modules that
series connected. The output of PV connected to the Sepic
converter. The output power of Sepic converter is distributed
to DC load. Block diagram of a system is shown in Fig. 4.

PV 1
(50 WP)

SENSOR SEPIC SENSOR

CONVERTER
PV 2
(50 WP)
I
. .
V
DC – DC
CONVERTER
.
V I
. DC LOAD

PWM (b)
PV 3
MICROCONTROLER
(50 WP)

Fig. 4. Block diagram system

The algorithms are applied to realize the MPPT

Partial Shading where the P–V characteristic curve has more
maximum points.

V. SIMULATION RESULT
To observe the AI method to find GMPP condition in
partial shading condition can be using software PSIM by
changing the irradiation of each PV modules in array
connection. From the simulation, P-V characteristic curve is (c)
shown in Fig. 5.
Fig. 5. P-V characteristic curve under partial shading
(a) Pattern 1 (b) Pattern 2 (c) Pattern 3

Fig. 5. are shown a P-V characteristic curve formed

when the PV is under partial shading. The maximum peak
power amount is equal to a number of the different
irradiance levels in the PV array. More irradiation
differences occurred, more maximum peak power created.
The maximum power point on the P-V characteristic
curve can be reached with build three PV modules that series
connection whose as same as specifications as shown in
Table I. PV modules are connected to Sepic converter which
has parameters in Table II. Furthermore, the output of the
Sepic converter is connected in DC Microgrid system to
supply the DC load. The simulation result of MPPT Partial
(a) Shading is shown in Fig. 6.

318
 (a) (b) (c)
Fig. 6(a) Tracking process of GMPP on Pattern 1 (a) FPA (b) GWO (c) MPSO

(a) (b) (c)

Fig. 6(b) Tracking process of GMPP on Pattern 2 (a) FPA (b) GWO (c) MPSO

(a) (b) (c)

Fig. 6(c) Tracking process of GMPP on Pattern 3 (a) FPA (b) GWO (c) MPSO

Table III. COMPARISON RESULT OF FPA, GWO, and MPSO METHOD

Pmax (W) Accuracy (%) Convergence (second)
Shading Pattern
Target FPA GWO MPSO FPA GWO MPSO FPA GWO MPSO
1 63.61 62.21 58.02 62.33 97.80 91.21 97.99 0.2 sec 0.6 sec 0.3 sec
2 86.99 85.17 77.69 84.42 97.91 89.31 97.05 0.2 sec 0.7 sec 0.3 sec
3 105.69 100.64 95.78 102.4 95.22 90.62 96.89 0.2 sec 0.6 sec 0.6 sec

According to data in Table III, the accuracy of MPPT simulation results also show that the MPSO method has
partial shading using proposed methods overall is greater more advantages compared to other proposed methods in
than 89% and can achieve the convergence situation lower tracking accuracy and the FPA method has more faster to
than 0.8 second. The proposed method can reach GMPP reach convergence situation by other proposed method.
without trapped in LMPP so the power losses from PV is
small. REFERENCES
[1] Kashif Ishaque, Zainal Salam, Muhammad Amjad, and Saad Mekhilef,
“An Improved Particle Swarm Optimization (PSO)–Based MPPT for PV
VI. CONCLUSION With Reduced Steady-State Oscillation,” IEEE TRANSACTIONS ON
The purpose of this paper is to compare the proposed POWER ELECTRONICS, VOL. 27, NO. 8, AUGUST 2012.
method to track the global peak from P-V characteristic [2] Yi-Hwa Liu, Shyh-Ching Huang, Jia-Wei Huang, and Wen-Cheng Liang,
curve when the solar PV modules under partial shading. “A Particle Swarm Optimization-Based Maximum Power Point Tracking
Algorithm for PV Systems Operating Under Partially Shaded
Overall, the proposed method has tracking accuracy greater
Conditions,” IEEE TRANSACTIONS ON ENERGY CONVERSION,
than 89% and can reach the GMPP under 0.8 second. The VOL. 27, NO. 4, DECEMBER 2012.

319
[3] Ramdan B. A. Koad, Ahmed F. Zobaa, and Adel El-Shahat, “A Novel Journal of Engineering Technology, Vol. 2, No. 1, June 2014.
MPPT Algorithm Based on Particle Swarm Optimisation for Photovoltaic [21] J. Prasanth Ram and N. Rajasekar, “A novel Flower Pollination based
Systems,” IEEE Transactions on Sustainable Energy. Global Maximum Power Point method for Solar Maximum Power Point
[4] Satyajit Mohanty, Bidyadhar Subudhi, Senior Member, IEEE, and Pravat Tracking,” TPEL-Reg-2016-06-1119, 2016.
Kumar Ray, Member, IEEE, “A New MPPT Design Using Grey Wolf [22] Rabah Benkercha, Samir Moulahoum, Ilhami Colak, Bilal Taghezouit,
Optimization Technique for Photovoltaic System Under Partial “PV Module Parameters Extraction with Maximum Power Point
Shading Conditions,” IEEE TRANSACTIONS ON SUSTAINABLE Estimation Based on Flower Pollination Algorithm,” 978-1-5090-1798-
ENERGY, 2015. 0/16/$31.00 ©2016 IEEE.
[5] Kashif Ishaque and Zainal Salam, “A Deterministic Particle Swarm [23] Satyajit Mohanty, Bidyadhar Subudhi, Senior Member, IEEE and Pravat
Optimization Maximum Power Point Tracker for Photovoltaic System Kumar Ray, Member, IEE, “A Grey Wolf Optimization Based MPPT for
Under Partial Shading Condition,” IEEE TRANSACTIONS ON PV System under Changing Insolation level,” IEEE Students Technology
INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013. Symposium, 2016.
[6] Duy C. Huynh, Tuong M.Nguyen, Matthew W. Dunnigan, and Markus A.
Mueller, “Global MPPT of Solar PV Modules using a Dynamic PSO
Algorithm under Partial Shading Conditions,” IEEE Conference on Clean
Energy and Technology, 2013.
[7] S.Mohsen Mirhassani, Mohsen Razzazan, and Amin Ramezani, “An
Improved PSO Based MPPT Approach to Cope with Partially Shaded
Condition,” The 22nd Iranian Conference on Electrical Engineering
(ICEE 2014), May 20-22, 2014.
[8] Chun-Liang Liu, Yi-Feng Luo, Jia-Wei Huang, and Yi-Hua Liu, “A PSO-
based MPPT Algorithm for Photovoltaic Systems Subject to
Inhomogeneous Insolation,” SCIS-ISIS 2012, Kobe, Japan, November
20-24, 2012.
[9] K. L. Lian, J. H. Jhang, and I. S. Tian, “A Maximum Power Point
Tracking Method Based on Perturb and Observe Combined With Particle
Swarm Optimization,” IEEE JOURNAL OF PHOTOVOLTAICS, VOL.
4, NO. 2, MARCH 2014.
[10] S.Z. Mirbagheri, M. Aldeen, S. Saha, “A PSO-based MPPT Re-initialized
by Incremental Conductance Method for a Standalone PV System,” 23rd
Mediterranean Conference on Control and Automation (MED) June
16-19, 2015.
[11] Chakkarapani Manickam, Guru Raghav Raman, Guru Praanesh Raman,
Saravana Ilango Ganesan, and Chilakapati Nagamani, “A Hybrid
Algorithm for Tracking of GMPP Based on P&O and PSO with Reduced
Power Oscillation in String Inverters,” IEEE TRANSACTIONS ON
INDUSTRIAL ELECTRONICS.
[12] D. W . Hart, Power Electronics, New York: McGraw-Hill, 2011.
[13] Farid Dwi Murdianto, Ontoseno Penangsang, Ardyono Priyadi,
“Modeling and Simulation of MPPT-Bidirectional Using Adaptive Neuro
Fuzzy Inference System (ANFIS) in Distributed Energy Generation
System,” International Seminar on Intelligent Technology and Its
Applications, 2015.
[14] Silicon Solar PV Module 50 WP.
[15] Suyanto, Farid Dwi Murdianto, Ontoseno Penangsang, A. Suprijanto, R.
Seto, “Wind-PV hybrid system modeling using bidirectional converter
with MPPT-dual adaptive neuro fuzzy inference system (ANFIS) in
microgrid isolated system,” Journal of Engineering and Applied Sciences,
vol. 11, pp. 2353-2359, December 2016.
[16] Soedibyo, Farid Dwi Murdianto, Suyanto, Mochammad Ashari, Ontoseno
Penangsang, “Modeling and simulation of mppt SEPIC combined
bidirectional control invers KY converter using ANFIS in microgrid
system,” Indonesian Journal of Electrical Engineering and Computer
Science, vol. 1, pp. 264-272, February 2016.
[17] Kashif Ishaque, Zainal Salam, Hamed Taheri and Amir Shamsudin,
“Maximum Power Point Tracking for PV System under Partial Shading
Condition via Particle Swarm Optimization,” IEEE Applied Power
Electronics Colloquium (IAPEC), 2012.
[18] Jose J. Chac6n, Reynaldo A. Ortiz, John E. Archila, Maria A. Mantilla,
M6nica A. Botero, Johann F. Petit, “Prototype for the Characterization of
Photovoltaic Panels based on a SEPIC Converter,” IEEE 978-1-5090-
6678-0/17/$31.00 ©2017.
[19] Moh. Zaenal Efendi, Novie Ayub Windarko, “DC POWER SUPPLY
WITH POWER FACTOR CORRECTION USING INTEGRATION OF
FLYBACK-FORWARD TOPOLOGY IN PARALLEL CONNECTION,”
Journal of Theoretical and Applied Information Technology (JATIT),
Vol. 78, No. 3, 31 August 2015.
[20] Moh. Zaenal Efendi, Abdul Rizal, Aldi Erzanuari, Suryono, Novie Ayub
Windarko, “AC-DC PFC Converter Using Combination of Flyback
Converter and Full-bridge DC-DC Converter,” EMITTER International

320