2018 13th IEEE International Conference on Industry Applications Mo9TrackB.

Optimization using NSGA II for Passive Filters in

Industrial Networks
Carlos Alberto Oliveira de Freitas
Jandecy Cabral Leite Jorge de Almeida Brito Júnior,
Research department
Research department Research department
Institute of Technology and Education
Institute of Technology and Education Institute of Technology and Education
Galileo of Amazon - ITEGAM
Galileo of Amazon - ITEGAM Galileo of Amazon - ITEGAM
Manaus, Brazil
Manaus, Brazil Manaus, Brazil
carlos.freitas@itegam.org.br
jandecy.cabral@itegam.org.br jorgebritojr@gmail.com
Tirso Lorenzo Reyes Carvajal
Manoel Henrique Reis Nascimento Milton Fonseca Junior
Research department
Research department Generation
Institute of Technology and Education
Institute of Technology and Education Eletrobras Amazonas GT
Galileo of Amazon - ITEGAM
Galileo of Amazon - ITEGAM Manaus, Brazil
Manaus, Brazil
Manaus, Brazil milton.fonseca@eletrobrasamazonasgt.com
tirsolrca@gmail.com
hreys@itegam.org.br

Abstract The optimization of passive filters in distribution filters in distribution systems has been approached through
systems has been presented by different computational different approaches. In general, these can be classified as
methods. The objective of this paper is to develop a single goal formulations [1-3] and multiobjective
computational tool with NSGA II for selecting the optimization [4].
configuration and design parameters of a set of passive filters Medium and high-power contaminant sources
for industrial electrical systems. As a methodology, the generally focus on industrial electrical systems. These include
static power converters and electric arc furnaces. For this
optimization problem was addressed using three independent
purpose single goal formulations usually attempt to determine
objective functions of innovative character for compensation
the least costly filters that ensure compliance with relevant
of harmonics through passive filters as a multiobjective standards of power quality standards. In multiobjective
problem. The results were the computational solution to this approaches, other objectives are added to achieve the
problem that determines a set of Pareto optimal solutions following: minimum total current distortion [5], minimum
(Frontier). In addition, the computational tool has several new total demand ratio [6], minimum total voltage distortion [7],
features such as: calculates the parameters that characterize minimum investment cost of filters [8], minimum cost losses
the filters, but also selects the type of configuration and the [9], etc.
number of branches of the filter in each candidate bar In commercial and residential installations, a large
according to a set of pre-established configurations according number of nonlinear loads of small power are employed,
to PRODIST-M8 and IEEE 519-2014. The results obtained which due to their large numbers can not be neglected as a
determine solutions with good performance indicators for source of distortion. This is the case of home and office
several characteristic and non-characteristic scenarios of the equipment, discharge lamps as shown by the standards [10,
11], among others.
system that allow to represent: daily variations of the load,
The harmonics injected into the electrical system by
and variations of system parameters and filters. It evaluates the non-linear loads produce effects: in the electric power
the cost of energy bills in an industrial power grid that has systems themselves and in the electric charges connected to
different operating conditions (characteristic scenarios) and them, as well as in the communications systems [12].
evaluates the economic effect of harmonic filters as reactive
power compensators. II. PASSIVE FILTERS

KeywordsQuality Power, NSGA II, Passive Filters, A. Introdution of Passive Filters

multiobjective optimization. When designing an industrial installation containing
large non-linear loads, the limits recommended by harmonic
I. INTRODUCTION distortion standard are generally violated. Given this premise,
Modern electrical systems contain the quantities of measures must be taken to ensure compliance with these
sources capable of contaminating or producing various limits and, in this way, reduce the undesirable effects of the
harmonic impacts in the distribution network where the non- harmonics in the industrial electrical system, thus avoiding the
linear loads found in industrial sectors, commercial and extension of power quality problems to the external supply
residential installations stand out. The optimization of passive system

978-1-5386-7995-1/18/$31.00 ©2018 IEEE 339 ISBN 978-1-5386-7995-1

 The means of compensation of the harmonic
d) Third-order filter
distortion by excellence are the filters of harmonics that aim
essentially to restrict the circulation of the harmonic currents The third-order filter incorporates a new capacitor C2
by the network, in order to avoid the distortion of the tension in the circuit (Figure 4).
in the bars [13].
For the operating principle, the harmonic filters can
be: passive, active or hybrid (when using a mixture of the first
two). Although active filters have shown advantages in low
voltage systems, passive filters are still the most attractive in
medium and high voltage systems [14]. There are several
types of parallel passive filters that can be classified into tuned
filters and damped filters [15].

B. Filter types
Figure 4: Third order filter. a) Topology, b) Impedance x frequency.
a) Filters tuned Source: Adapted from [19]
In the tuned filters or bandpass filters, the passive
e) C type filter
circuit consists of a capacitor and a series inductance to a low
value resistor [8]. Figure 1 shows the tuned filter The design of this filter shown in Figure 5 is based
on the fact that Xl is equal to Xc2 and therefore produces a
series resonance therebetween at the fundamental frequency
so that the resistance is short-circuited at this frequency and
the filter operate as a capacitor [19]

Figure 1: Filter tuned. a) Topology, b) Impedance versus frequency.

Source: Adapted from [16]
b) Damped Filter (High pass)
The damped filters shown in Figure 2 are
characterized by having an impedance characteristic versus Figure 5: Filter type C. a) Topology, b) Impedance versus
smoother frequency, which allows the passage of high frequency. Source: Adapted from [19]
frequencies and therefore their elimination.
The impedance of the type C filter for the frequency
n in (1) is:
jR
Xl  n  Xc2 / n  (1)
Z  jXc1 / n
R  j
Xl  n  Xc2  n 

Figure 2: Damped filters. (a) 1st order, (b) 2nd order, (c) 3rd
order, (d) type C. Source: [17] Following the described procedure, the resistance is
obtained from [20] according to (2):
c) Second order filter




 
The most used damping filter in practice is the Xc1 (2)
R n2 Q2  2  n4 1 

second order filter as shown in Figure 3 whose behavior

n3 n 4  1 Q
depends on the quality factor used in its design. A high quality





 n12 Q 4  2Q 2  2  2n10 3Q 2  4  n8 4Q 2  11  2n 6 Q 2  2  2n 4 Q 2  2  4n 2  1 
factor implies a more selective bandwidth, while a low quality
factor reduces the impedance of the filter for high frequencies C. Saturation of components

The standard [21] states that the power capacitors of

the harmonic filters must be able to operate on a continuous
basis under any condition of the system provided that the
following conditions are met:
1) The rms voltage applied to the capacitor does not
exceed 110% of its rated voltage rms (Vcnom).
2) The peak voltage applied to the capacitor
(including harmonics but not transients) does not exceed
120% of its rated peak voltage.
Figure 3: Filter of second order: a) Topology, b) Impedance 3) The rms current flowing through the capacitor
versus frequency. Source: Adapted from [18] does not exceed 135% of its nominal current rms (I cnom).

340
 4) The reactive power generated by the capacitor coding in real numbers, facilitating the interpretation of the
does not exceed 135% of its nominal reactive power (Q cnom). data stored in this chromosome. To represent a set of data on
a chromosome that can be of variable size (the type of filter
Thus, if h is the set of harmonics to which the and the number of branches chosen may differ from one
capacitor is subjected, one can establish the following solution to another), this chromosome must be able to
relations (3), (4), (5) e (6): represent the maximum number of data it defines a filter.
For the location of the passive filters in the industrial
V
hH
h
2

1.1Vcnom (3) installation, a set of K-bars should be determined where such
filters are to be installed. These bars are usually those in which
where Vh represents the harmonic voltage h applied to the there are significant nonlinear loads or distribution centers
capacitor. which have a set of such loads [15]
The chromosome representing an individual's data
 #   
 
















%& consists of an arrangement of the K elements, where each Sk
element as shown in Table 1 is an arrangement of integer and
real data representing the various parameters of the harmonic
where Vpeak is the peak voltage applied to the capacitor.
filter to be located on bar k.
I
hH
h
2

1.35  Icnom (5)
Table 1: Variables that describe a filter represented on the
where Ih represents the current of the harmonic h that chromosome [22][25].
circulates through the capacitor Variable Description

Cfg Configuration Type (1, 2, 3, 4)
Qch
1.35  Qcnom (6)
m Number of branches tuned (if it is type 1 filter)
hH
Qc Total reactive power in capacitors
where Qch is the reactive power of the harmonic h generated
Fd1 Factors for the distribution of reactive power among
by the capacitor.
Fdw+1 all branches
In the saturation of reactors and resistors, although
Fq1 Tuning frequencies of all branches
there is no specific norm [12], it is considered that the values Fqw+3
of rms voltage, rms current and the nominal power of these
Q1 Quality factors of all branches
elements, cannot surpass any condition of stable operation of Qw+3
the filter.
IV. RESULTS AND DISCUSSIONS
D. Filter costs
a) NPV of filters design
The investment cost of a filter is the sum of the costs
of its component elements [12]: The installation of the harmonic filters in the system
1) Capacitors, reactors and resistors; has two fundamental effects: the reduction of the harmonic
2) Protection (fuses, switches, etc.), and; distortion of the voltages and currents; and the compensation
3) Housing (Chassis, etc.) of the reactive power of the load. The reduction of the
harmonic distortion of the voltage in the system bars improves
With respect to the cost of the other elements, this the quality of energy supplied to the loads supplied from the
can be considered as 1% of the total cost of the filter [12]. electrical system of the industrial installation as well as from
other consumers that are fed from the PCC or the bar under
III. MATERIAL AND METHODS consideration. Although it is known that increasing the quality
A. Problem formulation of energy means reducing the operating costs of electrical
Given the issues raised, this thesis formulates the equipment, it is difficult to economically evaluate this result.
problem of optimizing the design of passive filters in
industrial electrical systems as a multiobjective problem that In addition, when the filters are installed, the currents
seeks the selection and design of passive filters necessary to circulating in the network are reduced at the fundamental
meet the following objectives: Maximize Net Present Value frequency, due to the large increase in the power factor, and
(NPV) installation filters design; Minimize total distortion of the harmonic frequencies due to the filtering effect of the
current in the CCP, and; Minimize the total distortion of the filters. Reducing harmonic currents through the network
voltage in the bars of the industrial electrical system. Subject reduces the loading of generators, transformers, cables, and
to the restrictions of: 1) Meeting the current energy quality other elements of the system, reducing system losses. The
standards; 2) Compliance with technical specifications. impact of reducing filter losses is easier to assess
economically when the required network and load data are
A.1. Variables of the problem known. In an industrial company, the cost components of the
The independent variables of the optimization electric energy bill is a convenient way to measure the annual
problem, represented by the X arrangement, are the types of cost of the electrical energy consumption of the facilities
passive filters to install and their respective design (Costs). To determine the economic effect of filter
parameters. In a genetic algorithm the problem variables are installation, we choose (L) scenarios typical of the daily load
somehow encoded on a chromosome representing the data variation to calculate the power consumption and the power
corresponding to a solution or individual. The computational factor of the installation. These characteristic scenarios
implementation of the NSGA-II used in this work uses a direct

341
correspond to the different load levels that are repeated daily As an objective of the filters installation project, the
for a given time. NPV(x) must be maximized. However, genetic algorithms
For each daily load scenario, the total active power usually work by minimizing the objective functions. In this
(PT) and the total reactive power (QT) supplied by the network, way, the first objective function to be minimized is defined as
as well as the active and reactive power losses in each element (13):
of the installation (including the filters) can be calculated by  %& " ! 
























%&
a power flow program at fundamental frequency and with a
harmonic penetration program. Using the calculated values of b) Harmonic control objectives
PT and QT, the maximum demand for active and reactive Passive harmonic filters are primarily harmonic
power as well as the active and reactive power consumption control devices whose function is to avoid the circulation of
of the installation can be estimated for a typical working day. distorted currents through the elements of the system,
Therefore, the monthly and annual electricity bill can be reducing the harmonic distortions of voltage in the bars. To
estimated when considering a number of typical working days evaluate the effect of filters on distortion rates, all possible
per year. This method of aggregation can be more or less operating scenarios of the system should be evaluated,
exact, insofar as the load of the installation is better including the L characteristic scenarios considered and
characterized. The Electric Energy Billing Manual defines another set of special system and load conditions. These
various types of tariffs to be used in electric energy billing and special conditions may include variations in network
defines the concepts of DREX Excess Reactive Demand and impedance, different modes of operation of harmonic
EREX Excess Reactive Energy, magnitudes that are producing loads, tuning of filters, etc. They are non-
calculated as the demand and the reactive energy that exceeds characteristic operating states for which a daily operating time
the reactive energy and demand values corresponding to a is not allocated, with impacts on energy calculations, power
power factor of 0.92. As the circuit load varies in different L factor, etc., but with influence on the determination of
characteristic load states, the billed energy is the sum of the harmonic distortion rates.
active power and the reactive power consumed in each state k For each scenario k considered, the total distortion of
of the annual duration tk: the current in the PCC (TDDk) and the total distortion of the
L L voltage in each bar i (THDk, i) can be calculated by a harmonic
E 
P  t  EREX  
P  DREX   t
(8)
F
k 1
Tk k k
k 1
Tk k k
flow program. Both rates, the total distortion of current in the
In the same way, the demand billed DF is composed PCC and the total distortions of voltage in the bars should be
of the sum of the active demand and the surplus reactive minimized by the optimization process [10, 23].
demand of the scenario k of maximum load To minimize TDD in all possible scenarios it would
be necessary to define an objective function for each scenario.
DF  max kL PTk  DREX k (9) However, considering that all harmonic control standards
Thus, the annual costs of electricity billing are limit only the maximum value (95% or 99% probability) of
harmonic distortion, it is only easier to minimize only the
calculated by:
maximum TDD value of all the operating scenarios of the
Cost ( x)  cD  DF ( x)  cE  EF ( x) (10) system as shows (14) [10, 23]:
Where cD($/kW) and cE($/kWh) are coefficients of the
corresponding electric charge cost. f 2 ( x)  max kW TDDk ( x) (14)
Thus, the benefits of installing the filters for the
characteristic L scenarios are determined as the difference Following the same reasoning, the maximum THD
between the annual cost of electricity bill before Cost(0) and value between all operating scenarios and all system buses
after the installation of the Cost(x) harmonic filters. The using the function f3 according to (15) is minimized.
investment cost of I(x) filters is composed of the costs of
capacitor, reactor, resistor and other elements. The cost of the f 3 ( x)  max kW THDk ,i ( x) (15)
capacitor, reactor and resistor depends linearly on its power iU
for each voltage level, while the other components of the cost c) Restrictions
can be assumed proportional to the reactive power of the filter
The body of constraints of the filter optimization
[12]. Thus, the investment cost of the filter is:
problem consider:
I ( x)  K C i QC i  K L i QL i  K R i PR i (11)
1. The stress quality constraints on the system bars;
iCC iC L iC R
2. The quality constraints of the currents in the PCC,
where Kc($/kvar), KL($/kvar) and KR($/kW) are the power cost and;
coefficients of capacitors QC, inductors QL and resistors PR 3. Saturation constraints on the filter components.
respectively, and CC, CL and CR represent the sets of each one The way of evaluating the quality restrictions of the
of these types of elements. voltage and current depends on the standard adopted to
Considering a period of evaluation of N years with formulate the problem. If the standards of [23] or [12] are
an interest rate i, the NPV of the installation project of the used, there are no limits for current distortion in the CCP, so
filters is calculated as shown by the relation (12): this set of constraints is not taken into account. Therefore, the
 formulation used here considers all possible restrictions

%& !
%&
 %& " !%& $











%& according to the adopted norms.
% & In addition, to ensure that the optimization program

obtains feasible solutions to the problem, a fourth objective

342
function to be minimized, which represents the quadratic sum c) Limit the reactive power generated by the capacitor
of all constraint violations of the problem is defined as (16): Qck,i. according (25).
f 4 ( x) 
vi ( x)  li 2
(16)
if Qc k , j  1.35Qcnom j , f 4  f 4 
Qc k , j  1.35Qcnom j 
2
vi ( x )li (25)
Where vi and li represent the calculated value and the limit Having evaluated all constraints of the problem, f4 is
value of parameter i bounded by the corresponding the quadratic sum of all violations of such constraints. If f4 is
constraints. zero, solution x will be feasible, otherwise (f4 


To determine the value of f4 for a set of x filters will not be efficient with the characteristics adopted for one
installed, we have the following procedure (the sentences are or more constraints.
written in pseudocode): Then the global optimization problem is defined as
1) Initialize with f4 = 0. shown in (26):
2) For each operating scenario k and each bar i of
the system, the voltage quality constraints of the min  f1 ( x), f 2 ( x), f 3 ( x) sujeito a  f 4 ( x)  0 (26)
type are evaluated:
a) Limit the value of the voltage modulus Vmk, i There are different ways to manipulate constraints in
according to (17). an optimization problem. However, since zero is the smallest
possible value of f4 and there is a multiobjective optimization
if Vmk ,i  Vlimi , f 4  f 4 
Vmk ,i  Vlimi 
2 method, the problem can be formulated as shown in (27).
(17)
min  f1 ( x), f 2 ( x), f 3 ( x), f 4 ( x) (27)
b) Limit the total voltage distortion THDk, i. as shown
(18). By minimizing f4, the algorithm tries to obtain the
zero value of this function, in other words, it looks for the
if THDk ,i  THDlimi , f 4  f 4 
THDk ,i  THDlimi  (18)
2
viable solutions of the problem. In this way, both feasible and
quasi-viable solutions (f4   
     
For each harmonic h, the limiting of the individual advantageous in very difficult solution problems.
distortion limit of the voltage IHDk,i,h, is evaluated. as shown
(19) IV Optimization algorithm
if IHDk ,i ,h  IHDlimi ,h , f 4  f 4 
IHDk ,i ,h  IHDlimi ,h  (19)
2
For the problem formulated for the design of filters
whose nonlinear features with real and integer variables
3) For each system operating scenario k (only for whose solution requires an optimization algorithm using the
standard [23], the current quality restrictions in NSGA II. The types of optimization problems present several
the PCC of the type:: objective functions, which are almost always in conflict, and
a) Limit the total distortion of demand TDDk. according if one wishes to optimize simultaneously in this case, in an
to (20). innovative way, it presents three objective functions (f1, f2 and
f3). In multiobjective optimization, the notion of optimal
if TDDk  TDDlim, f 4  f 4 
TDDk  TDDlim 
2
(20) solution is replaced by the notion of Pareto unpaired or
optimal solution [24].
For each harmonic h, the limit constraint of the
individual distortion of the current demand IDDk,h, is IV Application examples
evaluated as shown (21):
This example corresponds to an industry that contains
if IDDk ,h  IDDlimh , f 4  f 4 
IDDk ,h  IDDlimh 
2
(21)
medium and low voltage loads. The electrical system uses a
4) For each system operating scenario k and each primary distribution network of 4160V that feeds the medium
capacitor j of the filters installed, the saturation voltage loads and four substations that feed the loads of 480V.
restrictions of capacitors of the type are The nonlinear loads are concentrated in the low voltage part
evaluated: and are formed by three-phase six-pulse converters.
In this case it is considered that the voltage of all the
a) Limit the voltage applied to the capacitor Vck, i. nodes of the network must comply with the quality indicators
according to (22). as established in the standard [23]. The industrial plant is
described according to the single-line diagram shown in
if Vck , j  1.1Vcnom j , f 4  f 4 
Vck , j  1.1Vcnom j  (22)
2
Figure 6.
Limit the peak voltage applied to the capacitor Vcpeakk,j.
according to (23).

if Vcpeakk , j  1.2 2Vcnom j , f 4  f 4  Vcpeakk , j  1.2 2Vcnom j 
2
(23)

b) Limit the circulating current through the capacitor

Ick,i. according (24)

if Ick , j  1.35Icnom j , f 4  f 4 
Ick , j  1.35Icnom j  Figure 6: Industrial plant single line diagram.
2
(24)
Source: [20].

343
 For the optimization process, five possible operating
scenarios are considered, which are presented in Table 2. In order to select the possible solution to the
Table 2: Scenarios for the analyzes. problem, considering that the PRODIST-Module 8 standard
Scenarios only restricts the voltage distortion, we can order the solutions
Parameter
1 2 3 4 5 in ascending order of maxTHD, maxTDD and -NPV
Daily scenario duration (h/day) 6 10 8 0 0 respectively. Figure 8 shows the ordered solutions, where as
Filter Capacitance Depreciation C(%) 0 0 0 0 10
Filter inductance depreciation L(%) 0 0 0 -5 5
maxTHD increases, maxTDD and -NPV decrease.
Short-circuit MVA in PCC (MVA) 250 250 250 125 125

The first three scenarios are load regimes

characteristic of a normal industrial plant work day,
considered to evaluate the 12-month energy bill with 30 days.
These scenarios do not consider depreciation of the filters
components, since they assume that they exactly maintain

-NPV($)
their design parameters. Scenarios four and five are
pessimistic conditions of network operation with reduced
short-circuit MVA in the PCC. In addition, these scenarios
add a depreciation of capacitance (C) and inductance (L)
for all filters that are installed. The bars (N4, N8 and N10)
were selected for the installation of filters considering that
they are the ones that feed non-linear loads.
To evaluate the economic effectiveness (NPV) of the Figure 8: Possible ordered solutions (case 1).
compensation project, it was considered a duration of five
years, with a rate of return of 10% per year. The following Here, different criteria can be used to choose the
cases were analyzed: solution and to be used. If the least voltage distortion solution
1) Design of filters for the three characteristic is selected as shown in Figure 10, a variant is obtained whose
scenarios; parameters are shown in Table 4.
2) Design of filters for the five possible scenarios. Table 4: Parameters of selected filters (case 1)..
In both cases, the limits of voltage harmonics [23]
Bus Parameter Branch 1 Branch 2 Branch 3 Branch 4
were used as energy quality constraints. In addition, 100 Type tuning in 2a order
generations of the algorithm were performed, with a Capacitors 4x50 kvar 4x50 kvar
population of 500 individuals. Frequency 4.7 7.6
N4
Quality factor 21.6 10
IV.3 Design of the filters for the three characteristic scenarios
The initial results of the problem (base case), Type tuning in tuning in tuning in tuning in
considering only the three characteristic scenarios (1, 2 and 3) Capacitors 4x50 kvar 1x50 kvar 2x50 kvar 1x50 kvar

are presented in Table 3. N8 Frequency 4.7 6.6 10.4 13

Quality factor 37.9 19.8 22.2 8
Table 3: Initial results (case 1).
Parameter Value
Annual energy cost ($ / year) 840124 Type tuning in tuning in
Maximum TDD (%) 7.412 Capacitors 3x50 kvar 2x50 kvar
Maximum IDD (%) 6.498 N10
Frequency 4.7 6.6
Maximum THD (%) 8.349
Maximum IHD (%) 6.267 Quality factor 28.1 34
Power factor 0.797

According to [23], these levels of distortion are This solution is composed almost exclusively of
within the established limits. Finished 100 generations, the tuned filters, since the selected second order branches have a
genetic algorithm produced a population of 500 solutions, for high quality factor. Thus, it is possible to obtain a solution
example. Extracting only viable solutions, the results obtained only with tuned filters, adopting the possible configurations
are shown in 9 for the Pareto frontier of the problem, as shown of the variables are given to a single configuration of type 1.
in figure 7. The developed program admits this possibility. The results
obtained, when installing the selected filters, are shown in
Table 5, where a great reduction of the harmonic distortion




indicators and the annual cost of electric power is proven.




Table 5: Final results (case 1)

Parameters Value %
Annual energy cost ($ / year) 638400 75.989
Maximum TDD (%) 2.596 35.024
Maximum IDD (%) 2.064 31.763
Maximum THD (%) 2.346 28.097
Maximum IHD (%) 1.635 26.090
Power factor 0.992 124.582
Cost of investment of the filters ($) 46687
Figure 7: Pareto front. NPV from the project 718005

344
 To verify the effectiveness of the solution for Figure 11: Distortion of the tension in the bar N10 (case 1).
variations of the filter parameters, the harmonic penetration However, a frequency sweep study in bar N10,
program is executed for all scenarios with different shown in Figure 12, comparing the impedance characteristics
depreciation of these parameters and the results are shown in vs base case frequency (without filters) and the response
Figure 11, where a reduction of the maximumTDD between obtained for all scenarios and with capacitance depreciation
31 to 37% and the maximumTHD between 27.2 and 47.1% (0 to + 10%) and the inductance (-5% to + 5%), shows that
compared to the base case values for all scenarios of the the impedance peaks occur in low order harmonics in the
problem and considering the possible depreciation of the nonlinear loads in the problem. Thus, this behavior is repeated
filters. in the figures N8 and N4, and it can be concluded that the
selected filters will perform satisfactorily.







Figure 12: Frequency sweep in bar N10.
Figure 9: Results with depreciation of the filters (case 1)..
V. CONCLUSION
As shown in Figure 9, the worst results are obtained It is concluded that the computational solution to this
when the components of the filters have a positive problem was achieved using the genetic algorithm NSGA-II
depreciation, which reduces the frequency of tuning, that determines a set of optimal solutions of Pareto (Border)
separating them from the harmonics to be eliminated. This that allow the designer to choose the most appropriate
same behavior is repeated for the individual harmonics, which solutions to the problem. In addition, the computational tool
is exemplified for the current distortion in the PCC shown in developed has several novelties such as: The parameters that
Figure 10 and the voltage distortion in the N10 bar shown in characterize the filters are calculated, but also the type of
Figure 11 for scenario 1. configuration and the number of branches of the filter in each
candidate bar according to a set of configurations
preestablished; Two standards have been programmed to
evaluate the energy quality constraints that can be selected by
the user; We determine solutions with good performance
indicators for several characteristic and non-characteristic
scenarios of the system that allow us to represent: the daily
variations of the load, and the variations of the system
  


parameters and the filters; It evaluates the cost of energy bills

in an industrial power grid that has different operating
conditions (characteristic scenarios) and evaluates the
economic effect of harmonic filters as reactive power
compensators. The positive results, from the analysis of
several practical examples, show the advantages of the
developed method.
Figure 10. Current distortion in the PCC (case 1).
ACKNOWLEDGMENTS
The authors thank the Institute of Technology and Education
 
 
 
  !"#& and The Amazonas
Research Foundation (FAPEAM) for their support in
completing this study.

REFERENCES
  


[1] M. Ghiasi, V. Rashtchi, and S. Hoseini, "Optimum

location and sizing of passive filters in distribution
networks using genetic algorithm," in Emerging
Technologies, 2008. ICET 2008. 4th International
Conference on, 2008, pp. 162-166.

345
[2] A. Zobaa, A. Vaccaro, H. Zeineldin, A. Lecci, and [13] J. Arrillaga and N. R. Watson, Power system
A. A. Monem, "Sizing of passive filters in time- harmonics, 2nd ed.: John Wiley & Sons, 2004.
varying nonsinusoidal environments," in Harmonics [14] A. B. Nassif and W. Xu, "Passive harmonic filters
and Quality of Power (ICHQP), 2010 14th for medium-voltage industrial systems: Practical
International Conference on, 2010, pp. 1-8. considerations and topology analysis," in Power
[3] J. C. A. Leite, I.P.; Azevedo, M.S.S., Nascimento, Symposium, 2007. NAPS'07. 39th North American,
M.H.R.; Moraes, N. M., Reis, A.M. , "Multicriteria 2007, pp. 301-307.
design of passive harmonic filters for industrial [15] A. B. Nassif, W. Xu, and W. Freitas, "An
installations using evolutionary computation investigation on the selection of filter topologies for
techniques.," Journal of Engineering and passive filter applications," IEEE transactions on
Technology for Industrial Applications (JETIA), vol. Power Delivery, vol. 24, pp. 1710-1718, 2009.
01, pp. 52-60, 2015. [16] N. H. A. Kahar and A. F. Zobaa, "Application of
[4] C. f. Yang, G. G. Lai, and C. T. Su, "Optimal mixed integer distributed ant colony optimization to
planning of passive harmonic filters using hybrid the design of undamped single-tuned passive filters
differential evolution considering variation of based harmonics mitigation," Swarm and
system parameters," International Transactions on Evolutionary Computation, 2018.
Electrical Energy Systems, vol. 23, pp. 1317-1334, [17] S. H. E. Abdel Aleem, A. F. Zobaa, and M. E. Balci,
2013. "Optimal resonance-free third-order high-pass filters
[5] J. Ji, H. Liu, G. Zeng, and J. Zhang, "The multi- based on minimization of the total cost of the filters
objective optimization design of passive power filter using Crow Search Algorithm," Electric Power
based on PSO," in Power and Energy Engineering Systems Research, vol. 151, pp. 381-394,
Conference (APPEEC), 2012 Asia-Pacific, 2012, pp. 2017/10/01/ 2017.
1-4. [18] B. J. Maundy and A. S. Elwakil, "Second order
[6] J. C. Churio-Barboza, J. M. Maza-Ortega, and M. bandstop and bandpass filters using transformers,"
Burgos-Payán, "Optimal design of passive tuned Microelectronics Journal, vol. 46, pp. 690-697,
filters for time-varying non-linear loads," in Power 2015/08/01/ 2015.
Engineering, Energy and Electrical Drives [19] G. Zhang, Y. Wang, W. Xu, and E. Sitther,
(POWERENG), 2011 International Conference on, "Characteristic Parameter-Based Detuned C-Type
2011, pp. 1-6. Filter Design," IEEE Power and Energy Technology
[7] S. Li, Y. Li, X. Luo, L. Zeng, and Z. He, "Multi- Systems Journal, vol. 5, pp. 65-72, 2018.
objective optimal design for passive power filters in [20] I. P. Abril, "Cálculo de parámetros de filtros pasivos
hybrid power filter system based on multi-island de armónicos; Calculation of the harmonics passive
particle swarm optimization," in Power Electronics filters parameters," Ingeniería Energética, vol. 33,
and Motion Control Conference (IPEMC), 2012 7th pp. 134-143, 2012.
International, 2012, pp. 2859-2863. [21] "IEEE Standard for Shunt Power Capacitors," IEEE
[8] R. Dehini and S. Sefiane, "Power Quality and Cost Std 18-2012 (Revision of IEEE Std 18-2002), pp. 1-
Improvement by Passive Power Filters Synthesis 39, 2013.
using Ant Colony Algorithm," Journal of [22] J. C. LEITE, "Projeto multicritério de filtros
Theoretical & Applied Information Technology, vol. harmônicos passivos para instalações industriais
23, 2011. utilizando técnicas de inteligência computacional,"
[9] H. Yue, G. Li, M. Zhou, K. Wang, and J. Wang, Doctor in Electrical Engineering, PPGEE/ITEC,
"Multi-objective optimal power filter planning in UFPA, Belém, 2013.
distribution network based on fast nondominated [23] P. d. D. d. E. ANEEL, "Elétrica no Sistema Elétrico
sorting genetic algorithms," in Electric Utility Nacional>PRODIST: Módulo 8-Qualidade de
Deregulation and Restructuring and Power Energia Elétrica," Revisão, vol. 10, p. 88, 2018.
Technologies (DRPT), 2011 4th International [24] C. Kawann and A. E. Emanuel, "Passive shunt
Conference on, 2011, pp. 234-240. harmonic filters for low and medium voltage: A cost
[10] I. S. Association, "519-2014-IEEE Recommended comparison study," IEEE Transactions on Power
Practices and Requirements for Harmonic Control in Systems, vol. 11, pp. 1825-1831, 1996.
Electric Power Systems," New York, IEEE, 2014. [25] J. C. Leite, I. P. Abril, W. F. Silva, A. de Oliveira
[11] P. R. Maciel, R. D. Lins, and P. R. Cunha, Castro, R. de Mendonça Nogueira and M. S. S.
Introdução às redes de Petri e aplicações: Azevedo, "Planning passive filters using NSGA II
UNICAMP-Instituto de Computacao, 1996. for industry applications," 2014 11th IEEE/IAS
[12] I. Std.153L, "IEEE Guide for Application and International Conference on Industry Applications,
Specification for Harmonic Filters," 2003. Juiz de Fora, 2014, pp. 1-8 .

346