Ga based selective harmonic minimization of a 11 level

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Special Issue: 07 | May-2014, Available @ http://www.ijret.org 237
GA BASED SELECTIVE HARMONIC MINIMIZATION OF A 11 LEVEL
CASCADED INVERTER USING PV PANELS
Kokila R1
, F. T. Josh2
1
PG scholar, EEE department, Karunya University, Tamil Nadu, India
2
Assistant Professor (SG), EEE department, Karunya University, Tamil Nadu, India
Abstract
A method for eliminating the lower order harmonics in a 11 level cascaded inverter is presented in this paper. The input for the
multilevel inverter is taken from solar panels so that the input voltage will be varying with respect to time. Genetic algorithm is a
selective harmonic minimization technique where the switching angles are calculated real time for a set of DC inputs from solar
panels. These values of switching angles are applied to a 11 level cascaded H-bridge inverter. This paper will give details regarding
the method along with the simulation results and calculation of Total Harmonic Distortion(THD). The target harmonics considered
for minimization were 5th
,7th
,11th
and 13th
harmonics.
Keywords: GA, PV, Cascaded MLI, Selective Harmonic Minimization, THD.
----------------------------------------------------------------------***------------------------------------------------------------------------
1. INTRODUCTION
One of the most prominent issues that is encountered in case
of high power converters is the presence of harmonics. For
elimination of harmonics, several modulation methods are
used. One of the most recent modulation method is Selective
Harmonic Minimization Pulse Width Modulation
(SHMPWM) method.
Selective harmonic elimination techniques for separate DC
sources were proposed in [1],[5] according to which
SHEPWM is an efficient modulation method for obtaining
output waveforms with acceptable harmonic content. But there
are certain drawbacks to SHEPWM method [2] such as the
non-cancelled harmonics are not considered in the algorithm
and could reach high amplitudes and the value of non-zeroed
harmonics cannot be managed to get any optimization
objective. [3] elaborates the elimination of harmonics in
multilevel converters by solving the harmonic elimination
equations by the method of resultant from elimination theory
for the polynomial form of equations. In general case, if the
voltage sources are unequal and varying with time, the
solution set increases. As a result, the degree of polynomials
increase and the time required to solve will also become more.
As a result, the look-up tables will increase exponentially as
the number of DC sources increase. So it needs to deal with
extrapolation leading to time consuming algorithms. A method
was proposed in [2] where the optimal switching angles were
calculated using genetic algorithm. But it did not consider the
variation of DC sources with respect to time. It was limited to
equal DC sources. In [7] , a method in which the harmonics
are eliminated using the theory of symmetric polynomials and
resultants was developed . The problem with this method is
that as the number of DC sources increases, the order becomes
very high. This leads to computational complexity. The
calculation of switching angles was calculated online using
Newton Raphson method in [4]. The limitations of this paper
is that it considers that the voltage sources will be varying ,
but they have the same value at the same time. They vary
keeping the same relative value. All the above papers
considered only the case of equal DC sources. It did not
consider the variation of DC sources with respect to time. In
practical cases, the variation of DC sources with respect to
time has to be considered. So in this work, the DC sources
feeding the multilevel inverters are taken from solar panels,
the value of input voltages of which are considered to be
varying with respect to time. The switching angles are adapted
to this variation of DC sources. Genetic Algorithm(GA) is
used to obtain the switching angles that corresponds to real
values of DC sources. Here 11 level multilevel inverter is
taken into consideration. The target harmonics to be reduced
are 5th
, 7th
, 11th
and 13th.
This paper is organized as follows. In section II, the set of
equations for harmonic minimization is described and GA is
introduced. In section III, GA based control is explained.
Section IV gives the simulation results of 11 level multilevel
inverter using GA. The value of THD and lower order
harmonics are calculated.
2. PROBLEM DEFINITION
In general case, the variation of DC sources with respect to
time also has to be considered. For example, if we consider the
output of solar panels, fuel cells, batteries etc. to be given as
the input of multilevel inverter, their variations should be
taken into consideration. The corresponding switching angles
given by the algorithm has to respond to this variation of DC

__________________________________________________________________________________________
sources. In this proposed method, GA is used to obtain the
angles corresponding to real values of DC sources.
Fig-1: 11 level cascaded inverter
Fig 1 shows the 11 level cascaded inverter. It requires 5 DC
inputs, in this case, it is taken from solar panels. The output
voltage equations corresponding to Fig 2 can be expressed as
the following equation
1 1 2 2
1,3,5,7,...
4
( ) .( cos( ) cos( )an dc dc
n
V t V n V n
n
  



  
3 3 4 4 5 5cos( ) cos( ) cos( ))dc dc dcV n V n V n   
(1)
Where 1 5,.....,dc dcV V - Input DC sources
1 5,.......,  - Output angles
anV - Inverter output voltage
The proposed approach works as follows:-The input voltages
from the solar panels will be varying with respect to time.
Genetic algorithm works to obtain the switching angles
corresponding to the variation of DC sources in such a way
that minimum total harmonic distortion is obtained. It also
emphasizes in reduction of lower order harmonics such as 5th
,
7th
, 11th
, and 13th
.
Fig-2:11 level inverter output waveform
3. GA BASED HARMONIC MINIMIZATION
A genetic algorithm (GA) is a stochastic search method shown
to be well suited for problems where many global minimum
and/or highly dimensional search spaces are possible. Each
individual of a set has an associated cost value, referred to as
“fitness function” that is a measure of how well this individual
performs in the population. In previous works, analytical
solutions were found partially in the range space of input
voltage variation; however, this approach still uses GA first,
because this range is used to calibrate it to perform in the
range where there is no analytical solution. The basic
continuous GA execution flow is shown in Fig 3.
Under the variations of DC sources, it is necessary to maintain
the fundamental output voltage and to cancel out the lower
order harmonics. Here the harmonics 5th
,7th
,11th
, and 13th
harmonics are the required harmonics to be cancelled. The
triplen harmonics such 3rd
and 9th
harmonics are not
considered because in three phase application, it is cancelled
in the line voltage.

__________________________________________________________________________________________
Fig-3: Flowchart for GA execution
The equations for the fundamental voltage, 5th
, 7th
, 11th
and
13th
harmonics are given below
1 1 2 2 5 5
4
[ cos( ) cos( ) .... cos( )]fund dc dc dcV V V V  

    (2)
5 1 1 2 2 5 5
4
[ cos(5 ) cos(5 ) .... cos(5 )]
5
th dc dc dcV V V V  

    (3)
7 1 1 2 2 5 5
4
[ cos(7 ) cos(7 ) ... cos(7 )]th dc dc dcV V V V  

   
(4)
11 1 1 2 2 5 5
4
[ cos(11 ) cos(11 ) ... cos(11 )]
11

   
(5)
13 1 1 2 2 5 5
4
[ cos(13 ) cos(13 ) ... cos(13 )]
13

   
(6)
The equations (2) to (6) are given as the equations to solve the
angles using GA. The fundamental voltage is taken as 230V
rms. The equations (3) to (6) are equated to zero in order to
eliminate them. [6] gives the objective function for the GA as
given in equation(7)
5 7 11 13 1( , , , , ) | 230|fund th th th th fundf V V V V V k V  + 2k
5 3 7| | | |th thV k V  4 11| |thk V  5 13| |thk V (7)
The values for 2k , 3k , 4k , 5k are generally taken greater
than 1k . It can be taken as multiples of 10. Calculations were
made for switching angles that correspond to the randomly
taken values of DC sources. The values of coefficients used in
Equation (7) were found through trial and error runs of the
algorithm. They allow the fundamental to be kept very close to
230 V while keeping harmonics close to zero. The GA
program was written using Matlab R2010a. The termination
criteria was given in such a way that the number of iterations
is 100. So the algorithm stops at 100th
generation and it will
return the best individual as the end product. Table 1 shows
the parameters used for running GA.
Table-1: Parameters for GA
Table-2: Dataset obtained through GA run
Input Voltages(V)
[ 1dcV 2dcV 3dcV 4dcV 5dcV ]
Output Angles(
0
)
[ 1 2 3 4 5 ]
[40 41 42 43 44] [0.182 21.709 44.615
60.274 88.363]
[40 42 43 44 45] [1.191 23.622 45.623
59.565 74.464]
[40 42 43 44 46] [1.983 30.864 46.557
61.677 89.58]
[41 42 43 44 45] [0.446 27.372 37.528
57.359 82.694]
[42 43 44 45 46] [0.317 20.038 48.37 59.782
81.719]
[43 44 46 47 48] [0.853 22.946 42.399
56.473 87.567]
[44 45 46 47 48] [0.878 18.767 33.007
56.152 73.215]
Population size 100
Number of generations 100
Stall generations 50
Crossover fraction 0.8

__________________________________________________________________________________________
Calculations were made for switching angles that correspond
to the randomly taken values of DC sources. Table 2
represents the dataset obtained through GA run for different
sets of input voltages.
Fig-4: GA fitness plot for 100 generations
Fig 4. shows the fitness plot of GA after 100 generations. The
best and the mean values of fitness functions were obtained.
Best fitness value:1112.7346
Mean fitness value:1154.5118
4. SIMULATION RESULTS AND ANALYSIS
In this section, the simulation results of 11 level cascaded
inverter is presented. The simulation is carried out using
MATLAB/SIMULINK software by the switching angles
obtained from Table 2. The output waveforms and the FFT
analysis were also carried out. The input is given from five
different solar panels. The solar panels were modeled using
the equations as given below.
The photon current is given by
[ ( 298)] /1000ph scr iI I K T S   (8)
Where scrI is the short circuit current
iK is the short circuit current temperature co-efficient
T is the module operating temperature
S is PV module illumination
The module reverse saturation current is
/[exp( / ) 1]rs sc oc sI I qV N KAT  (9)
Where q is electron charge
ocV is the open circuit voltage
sN is number of cells connected in series
K is Boltzman constant
A is ideality factor
The module saturation current is
3
[ / ] exp[( / )(1/ 1/ )]s rs r go rI I T T qE AK T T  (10)
Where rT is the reference temperature
goE is the bandgap
The output current of PV module is
exp ( )
[ 1]s
p ph p s
s
q V IR
I N I N I
N AKT

   (11)
pN is the number of parallel connected cells
sR is series resistance
Fig-5: Simulation Diagram of 11 level inverter

__________________________________________________________________________________________
Fig-6: Subsystem of multilevel inverter
Fig-7: Subsystem of solar panel
The output voltage and THD analysis for 11 level inverter
using solar panel inputs are shown in Fig 8 and Fig 9
respectively.
Fig-8: Output waveform of multilevel inverter
Fig-9: FFT analysis
From the FFT analysis, it can be seen that the THD is less and
the lower order harmonics are minimized.
The value of THD and lower order harmonics are given in
table 3.
Table-3: Parameters obtained from the FFT analysis of output
of GA controller
Parameter Value
Fundamental Voltage fundV (V) 182.5
THD(%) 13.77
5th
Harmonic 5V (%) 0.57
7th
11th
13th
5. CONCLUSIONS
This paper presents an approach for real time computation of
switching angles using GA. The solutions were found in real
time using genetic algorithms to obtain a data set. GA was
also used so as to explore the advantages of approximate
solutions. The 11 level inverter whose inputs were taken from
PV panels was simulated and the output was obtained. The
THD analysis was also carried out. It is observed that the
value of THD was very low. It also reduces the lower order
harmonics such as 5th
,7th
, 11th
and 13th
. Future work points in
the direction of implementation and performance evaluation
over a greater number of angles. This would allow for
harmonic control while keeping the switching frequency
minimum and complying with THD standards. The literature
4
Conn4
3
Conn3
2
Conn2
1 Conn1
gm
DS
Q14
gm
DS
Q13
gm
DS
Q12
gm
DS
Q11
Pulse
Generator3
Pulse
Generator2
Pulse
Generator1
Pulse
Generator

__________________________________________________________________________________________
review also pointed the possibility of determining what the
minimum number of angles needed is in order to comply with
certain THD level. Future work also emphasises on the real
time hardware implementation of the GA tuned ANN
technique using adSpace or DSP platform.
REFERENCES
[1]. J. Napoles, J. I. Leon, R. Portillo, L. G. Franquelo, and
M. A. Aguirre, “Selective harmonic mitigation technique for
high-power converters,” IEEE Trans. Ind. Electron., vol. 57,
no. 7, pp. 2315–2323, Jul. 2010.
[2]. J. N. Chiasson, L. M. Tolbert, K. J. McKenzie, and Z. Du,
“A unified approach to solving the harmonic elimination
equations in multilevel converters,” IEEE Trans. Power
Electron., vol. 19, no. 2, pp. 478–490, Mar. 2005.
[3]. J. N. Chiasson, L. M. Tolbert, K. J. McKenzie, and Z. Du,
“Elimination of harmonics in a multilevel converter using the
theory of symmetric polynomials and resultants,” IEEE Trans.
Control Syst. Technol., vol. 13, no. 2, pp. 216–223, Mar.
2005.IEEE Trans. Control Syst. Technol., vol. 13, no. 2, pp.
216–223, Mar 2005.
[4]. T. Tang, J. Han, and X. Tan, “Selective harmonic
elimination for a cascade multilevel inverter,” in Proc. IEEE
Int. Symp. Ind. Electron., Jul. 2006, vol. 2, pp. 977–981.
[5]. J. Chavarria, D. Biel, F. Guinjoan, C. Meza, and J. J.
Negroni, “Energybalance control of PV cascaded multilevel
grid-connected inverters under level-shifted and phase-shifted
PWMs,” IEEE Trans. Ind. Electron.,vol. 60, no. 1, pp. 98–
111, Jan. 2013.
[6]. Faete Filho, Helder Zandonadi Maia, Tiago H.A. Mateus,
Burak Ozpineci, Leon M. Tolbert and Joao O. P. Pinto,
“Adaptive Selective Harmonic Minimization Based on ANNs
for Cascade Multilevel Inverters With Varying DC Sources” ,
IEEE Transactions on Industrial Electronics, Vol. 60, No.5,
pp. 1955 – 1962, May 2013.
[7]. Chiasson J. N , Ozpineci B, and Tolbert L . M, “Harmonic
optimization of multilevel converters using genetic
algorithms,” IEEE Power Electron Lett., vol. 3, no. 3, pp. 92–
95,Sep.2005.
[8]. T.Ikegami,T.Maezono,F.Nakanishi, Y. Yamagata and
K.Ebihara, “Estimation of equivalent circuit parameters of PV
module and its application to optimal operation of PV
system,”Elsevier solar energy materials and solar
cells,67(2001),pp.389-395,Jan 2001.
BIOGRAPHIES
Kokila R completed her B.Tech degree in
Electrical and Electronics Engineering at
LBS College of Engineering, Kasaragod.
She is pursuing M.Tech in Power
Electronics and Drives at Karunya
University, Coimbatore. Her area of
interest include Multilevel inverter, soft
computing etc.
Mr. F. T. Josh completed his B. E in
Electrical and Electronics Engineering in
2000. He completed his M.E degree in
Power Electronics and Drives. He has 12
years of experience in teaching. He is
currently working as Assistant
Professor(SG) in Karunya University,
Coimbatore. His areas of interest include Multilevel inverters,
soft computing etc.

Ga based selective harmonic minimization of a 11 level

More Related Content

What's hot (19)

Viewers also liked (20)

Similar to Ga based selective harmonic minimization of a 11 level (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Ga based selective harmonic minimization of a 11 level