Professional Documents
Culture Documents
By
Master of Science
In
ELECTRICAL POWER ENGINEERING
Master of Science
In
ELECTRICAL POWER ENGINEERING
Co-Supervisor
Master of Science
In
ELECTRICAL POWER ENGINEERING
"Professor in Electrical Engineering Department - Faculty of Engineering Fayoum University" (Main Advisor)
ACKNOWLEDGEMENT
ABSTRACT
Electric power systems have rapidly grown for the
past fifty years. As a result, the number of cables in
operation and their total length had largely increased.
Underground cables have been widely implemented due
to reliability and environmental concerns. To improve
the
reliability
of
distribution
system,
accurate
fault
accelerating
location
plays
an
system restoration,
important
reducing
role
outage
in
time,
this
study
presents
fault
based
on
Discrete
Wavelet
Transform
and
tested
using
current
samples
only.
The
results
iii
TABLE OF CONTENTS
Title
Page
Acknowledgment
Abstract
ii
Table of Contents
iv
List of Tables
List of Figures
xii
xvii
1.0
2.0
Introduction
1.1
Problem of Statement
1.2
1.3
1.4
Outline of thesis
1.5
Summary of Work
Literature Survey
2.1
1.1
Introduction
2.2
iv
2.3
11
11
13
3.0
4.0
17
Introduction
22
3.2
Definition of Fault
22
3.3
Distribution faults
23
3.4
25
3.5
Fault types
25
Introduction
28
4.2
Fuzzy Logic
28
4.2.1
Fuzzy Sets
29
4.2.2
4.3
4.2.3
33
4.2.4
37
4.4
31
38
39
4.5
5.0
4.4.1
Fuzzification
42
4.4.2
44
4.4.3
Normalization
44
4.4.4
Defuzzification
45
4.4.5
Neuron addition
45
46
47
5.2
48
5.2.1
5.2.2
48
49
5.3
Software Implementation
50
5.4
50
5.5
53
54
5.5.1
5.6
52
vi
54
57
57
62
5.6.2.1
62
5.6.2.2
63
5.6.2.3
65
5.6.2.4
5.6.2.5
5.7
5.8
67
69
74
74
75
81
5.8.3.1
81
5.8.3.2
85
5.8.3.3
88
5.8.3.4
5.9
66
vii
91
94
95
6.0
Simulation Results
6.1
Introduction
101
6.2
101
6.3
6.3.1
location model
Double Line to Ground ANFIS fault
6.3.2
location model
6.3.3
6.3.4
6.4
104
105
109
113
117
121
122
6.4.11
122
6.4.1.2
122
6.4.1.3
122
6.4.1.4
122
viii
7.0
Conclusions
133
7.2
Recommendation
134
135
REFERENCES
Appendix A
A.1
A.2
146
158
Appendix B
B.1
ix
159
LIST OF TABLES
TABLE
TITLE
PAGE
3-1
24
5-1
System Data
58
5-2
61
5-3
Number of simulations
61
5-4
62
6-1
103
6-2
104
6-3
107
6-4
111
6-5
115
6-6
119
6-7
123
6-8
125
6-9
127
6-10
129
xi
LIST OF FIGURES
FIGURE
TITLE
PAGE
1-1
Summary of work
4-1
32
4-2
32
4-3
33
4-4
The concept of A B
34
4-5
36
4-6
43
5-1
56
5-2
57
5-3
Cable Configuration
58
5-4
60
5-5
63
5-6
63
5-7
64
xii
5-8
64
5-9
65
5-10
66
5-11
67
5-12
67
5-13
68
5-14
68
5-15
70
5-16
71
xiii
5-17
72
5-18
72
5-19
73
5-20
78
5-21
78
5-22
79
5-23
79
5-24
83
5-25
83
5-26
84
5-27
84
5-28
86
5-29
86
5-30
87
xiv
5-31
87
5-32
89
5-33
89
5-34
90
5-35
90
5-36
92
5-37
92
5-38
93
5-39
93
5-40
96
5-41
97
5-42
98
5-43
99
5-44
100
6-1
103
xv
6-2
106
6-3
109
6-4
110
6-5
113
6-6
114
6-7
117
6-8
118
6-9
121
6-10
125
6-11
127
6-12
129
6-13
132
xvi
LIST OF ABBREVIATIONS
Wavelet approximation
ANFIS
ANN
AI
Artificial Intelligence
ATP/EMTP
BPNN
Capacitance
Wavelet detail
db4
Daubechies 4
DWT
EPRI
ES
Expert System
Frequency
FACTS
FIS
FL
Fault location
GUI
HIF
HP
HVDC
Inductance
L-G
LIF
L-L
Line to Line
L-L-G
L-L-L-G
xviii
CHAPTER 1
Introduction
Electrical
power
distribution
systems
are
possible
in
the
system.
Through
fault
location
1 - Introduction
prediction of cable, troubled cable can be detected and
located earlier and accurately.
review
classification
the
and
approach
location
of
for
fault
underground
cable.
2. To simulate the network using ATP/EMTP
and MATLAB software.
3. To classify the type of fault in underground
cable.
4. To estimate fault location of underground
cable.
Estimator
capabilities,
fault
area
can
be
1 - Introduction
1.4 Outline of thesis
This thesis consists of seven chapters. In first
chapter, it discusses about the background of the thesis,
statement of problem, objective of the thesis, scope of
the thesis, outline of thesis as long as summary of work.
While
Chapter
2,
discusses
more
on
theory
and
well
discusses
general
view
of
discusses
on
the
methodology
and
1 - Introduction
theoretical studies to enhance knowledge on the thesis done.
Besides that, softwares that are used for the thesis also are
study.
In the second stage the works are more focusing on
simulation and modification. Results obtained from the
simulation of both software (ATP and MATLAB) were discuss
and then tested.
1 - Introduction
Construct underground cable model in ATP/EMTP
CHAPTER 2
Literature Survey
2.1 Introduction
Various fault classification and location methods
have been proposed and they can be categorized as
analytical
methods,
artificial
intelligence
(AI)
based
2 - Literature Survey
and low pass (LP) filter. The samples of output signal
from the HP filter are known as detail coefficients, and
those from the LP filter are known as approximated
coefficients. This is considered the first level. Next, the
obtained samples from LP filter (first level approximated
coefficients) are further processed through HP and LP
filters forming the second level. In this way, the original
signal can be processed through many levels depending
on the need of resolution of signal. Reference [1]
proposes
wavelet
based
technique
for
fault
new
algorithm
based
on
multiple
scale
can
be
detected
and
classified
by
the
2 - Literature Survey
The applications of the discrete wavelet transform
(DWT) for identifying the phase with fault appearance
along the underground cable distribution system have
been investigated in [6]. Daubechies4 (db4) is employed
as mother wavelet in order to decompose high frequency
components from fault signals. Coefficients of positive
sequence current signals are calculated and employed in
fault detection decision algorithm. The results show that
the proposed algorithm can indicate fault types with the
accuracy higher than 90%.
In
[7],
technique
using
discrete
wavelet
distribution
system
has
been
proposed.
signals
obtained
by
the
discrete
wavelet
2 - Literature Survey
[8]
shows
that
studying
on
respective
2 - Literature Survey
[10] Introduces the application of Adaptive NeuroFuzzy Inference System (ANFIS) for fault classification
in transmission lines. The proposed method uses two
ANFIS units. The fault detection unit is built by using
various training data at fault and no fault conditions. The
fault classification unit is built at different situations of
all fault types (i.e. single line to ground, double lines,
double lines to ground and three lines fault). Then the
unit is tested using testing data different from those of
the training stage.
[11] Proposed an ANFIS based fault classification
scheme in neutral non-effectively grounded distribution
system. The transient currents are obtained by wavelet
transform after faults occur. According to the statistic
characteristic
types,
the
of
fault
transient
currents
identifiers
are
in
different
defined.
The
fault
fault
10
2 - Literature Survey
In
[12],
an
algorithm for
fault
detection
and
faults
(HIF)
has
been
developed
using
acquire
three-phase
currents,
calculate
zero
fault
classification,
the
next
step
is
to
2 - Literature Survey
current values are used to calculate apparent impedance.
Based on the equation of apparent impedance, fault
location is estimated [13].
Reference [14] shows the implementation of the circuit theory based
method for rural distribution feeders. Depending on the fault type,
specific voltage and current values are selected to find out the
apparent impedance. The equation of apparent impedance consists
of unknown terms like fault resistance and distance of fault. By
separating the apparent impedance equation into real and imaginary
parts, the fault location is calculated. In this method, the equations
are shown only for a line-to-ground fault but a similar analysis can
be performed for other types of faults also. Instead of eliminating
the fault resistance, the equation of distance of fault is expressed as
a function of fault resistance in [15]. As the equation is in the form
of a square root of quadratic equation, the minimum value of fault
resistance is calculated and substituted in the equation of distance of
the fault to find the location of the fault. In [16], the fault location is
estimated by comparing the reactance of different sections of line
with apparent reactance. If the modified reactance of a section is not
greater than the apparent reactance, then the fault location is
considered to be present after that section. Initially, fault location is
assumed between two points based on the information of fault type
and sequence voltage and currents. Then, the modified reactance of
section is compared with the apparent reactance to find the accurate
12
2 - Literature Survey
location of fault between those two assumed points. If there are any
subsections on the lines such as laterals, then, an equivalent radial
system is considered and the relationship between voltage and
current is analyzed to find the location of fault.
As this involves more calculations, another approach for fault
location is presented in [17]. For balanced systems, the sequence
components of phase voltages, phase currents and impedances are
used to form a quadratic equation in terms of fault distance and fault
resistance. Unknown fault resistance is calculated by the imaginary
part of this equation and the fault resistance is substituted in the real
part of the equation to obtain the fault location. For unbalanced
systems, direct circuit analysis is used to form a quadratic equation
in terms of fault distance and fault resistance. In between, matrix
inverse lemma technique is used for calculating inverse operations
of matrices. Then, the quadratic equation is solved to find the fault
location.
2 - Literature Survey
propagation velocity and time of waveforms, the location
of fault is known. This technique is mostly used for extra
high voltage transmission lines than distribution lines
because distribution system contains many subsections
such as laterals and feeders. These subsections may
create disturbances for the traveling waves on the lines.
Reference [20] explains fault location technique based
on the traveling wave theory. It includes two methods
the single ended method and the double ended method.
In the single ended method, only one fault locator is
installed at one substation.
Depending
on
the
arrival
time
of
waves
at
that
ended
method,
there
are
two
fault
locators
14
2 - Literature Survey
In
[21],
wavelet
transform is
used
to
analyze
the
problems
in
reflected
waveforms,
separate
system
is
modeled,
and
the
apparent
lines
and
underground
15
cables.
In
[23],
2 - Literature Survey
wavelet analysis based single ended method is used to
locate the fault on the transmission line. In [24], double
ended based method is used to find the fault location in
underground cables.
In [25] a new single ended fault location method is
proposed for underground cable combined with overhead
lines. In this algorithm fault clearing high frequency
transients are used instead of fault-generated transients
and the line parameters are not needed. In the proposed
algorithm, samples just from voltage transients generated
by fault clearing action of circuit breaker are taken from
the sending end of the cable line. Applying wavelet
transform, the first three inceptions of traveling waves to
the fault locator are detected. Using these, the proposed
algorithm at first identifies fault section, overhead or
cable, and then wave speed is calculated and at last
location of fault is determined accurately. Because of
using only voltage samples taken from one terminal, it is
simple and economic and does not need to GPS and data
communication and synchronization.
16
2 - Literature Survey
2.3.3 Artificial Intelligence (AI) Based Methods
The following are the three major approaches in
AI based methods.
Artificial Neural Networks (ANN) based method.
Fuzzy Logic based method.
Expert System based method.
ANN based method detects the location of fault in
power system by recognizing the patterns of voltages
and currents for different fault locations. References [26,
27] suggest separate fault locators for each type of fault
whereas references [28, 29] propose only one single fault
locator
for
all
types
of
faults.
Depending
on
the
[27]
shows
that
the
Marquardt-
and
currents
along
with
the
neutral
17
2 - Literature Survey
Also, in [26], voltages and currents of faulted
phases along with the neutral current are used as inputs
for a three layer feed-forward ANN to detect the fault
location
in
medium
voltage
cables.
The
Marquardt-
2 - Literature Survey
wavelet
transform
and
finally
the
fault
position
is
combined
overhead
transmission
line
with
i.e.,
currents
and
fundamental
zero
component
sequence
current.
of
three
Another
phase
ANFIS
2 - Literature Survey
the knowledge of protective device settings and ANFIS
nets. Wavelet transform is used to obtain the current
patterns
in
this
paper.
The
proposed
methodology
distribution
substation.
The
output
data
(WT)
and
ANFIS
techniques
for
fault
2 - Literature Survey
23,436 fault cases covering a wide range of operating
conditions. Results of this paper show that the algorithm
has an effective structure in fault diagnosis for series
compensated transmission lines.
In [37], The Proposed ANFIS uses only post-fault
three-phase RMS currents as inputs, and outputs were XY points and integers between 1 and 10. The results
show that the approach can accurately locate the faults
and identify the faults types in the DISCO-Net 47-nodess
test practical distribution system. Its accuracy has been
validated by comparing the actual fault locations and
types
of
fault
obtained
from
21
the
PSS-ADEPT
CHAPTER 3
General Fault Characteristics
3.1 Introduction
Underground cables are used in 33kV, 11kV and
415V distribution networks. Faults are major disturbance
to the power system. Hence the need to locate the faulty
point in an underground cable in order to facilitate
quicker
repair,
improve
the
system
reliability
and
system for
accurate
pinpointing
of
the
fault
the
system
availability
and
performance.
3. Reduce operating cost and save the time
required
by
the
crew
searching
in
bad
impedance
in
the
fault
connection.
The
term
intentionally,
between
two
points
of
different
23
balanced.
In
practice,
most
faults
in
power
are
either
temporary
or
permanent.
or
failed
equipment.
Virtually
all
faults
on
damage
any
25
system
equipment.
If
the
operates
normally.
Temporary
faults
cause
27
CHAPTER 4
Adaptive Neuro Fuzzy Inference System (ANFIS)
4.1 Introduction
Adaptive neuro fuzzy inference system (ANFIS) is
based on fuzzy logic modeling and uses artificial neural
network as the learning algorithm. The system can teach,
change the data environment or respond to the remote
stimulus for adapting to the change of data environment
[41]. ANFIS produces constant and linear target by using
respective zero and first-order polynomial equations and
is also known as a Sugeno-type of fuzzy inference
system (FIS).
logic
calculus
28
but
its
application
was
critics
continually
state
that
all-
with
special
continuous
function
kind
of
fuzzy
controllers,
on
compact
set
can
each
be
obtain
fuzzy
sets
by
extending
the
degree
of
membership
can
be
considered
in
( ) |
(4.1)
where
to
membership
grade
(or
membership
value)
between 0 and 1.
The definition of a fuzzy set is a simple extension
of
the
definition
of
classical
set
in
which
the
30
Usually X
is
referred
to
as
the
universe
of
4.2.2
Membership
function
formulation
and
parameterization
Since most fuzzy sets in use have a universe of
discourse X consisting of the real line R, it would be
impractical to list all the pairs defining a membership
function. A more convenient way to define a MF is to
express it as a mathematical formula. Here are some of
the MFs used in the fuzzy sets literature
Definition 4.2 Triangular function [46]
A triangular MF is specified by three parameters
{a,b,c} as follows:
( , , , )=
,
,
,
,
(4.2)
31
1+
(4.3)
32
, )=
(4.4)
using
complement
intersection
(NOT)
(AND),
operations.
union
The
(OR)
and
mathematical
33
A B A (x) B (x) x X
(4.5)
34
(4.6)
(x) = 1 - A (x) x X
(4.8)
35
(4.7)
Figure 4.5 (a) two fuzzy sets A and B; (b) A ; (c) AB ; (d) A B .
only
possible
generalizations
of
the
crisp
set
implication,
or
fuzzy
conditional
statement)
sets
on
universes
of
discourse
and
Y,
Examples
of
fuzzy
if-then
rules
are
the
road
is
slippery,
then
driving
is
dangerous.
If a tomato is red, then it is ripe.
If the speed is high, then apply the brake a
little.
37
performance
according
parameters
to
are
of
the
membership
particular
identified
data
based
function
input.
on
the
The
curve
curve
respective
38
ANFIS
model
exhibits
predicted
target
where,
(4.9)
can
be
noticed
even
though
the
epoch
was
process
which
is
combination
of
two
algorithms
hybrid
method
improves
the
bad
features
are popular in
of
ANN
implementation.
In hybrid learning algorithm, MF parameters are
adjusted to identify the best prediction value. There are
MF curve in input and output parts of ANFIS model.
Back-propagation algorithm takes responsibility to vary
MF parameter in input side of the model, whereas least
square estimate (LSE) takes into consideration on the
output side as a linear line. In MF parameters, the input
side varies, whereas for output, they are static and vice
versa.
The prediction values are performed after the MF
parameters in both sides of the ANFIS model converge
the values according to the given training epoch [49].
ANFISs
learning
process
can
be
implemented
to
form a
fault
diagnosis
system in
power
following equations:
( )=
( )=
1+
1+
(4.10)
(4.11)
where, Xi(x) and Yi(y) are fuzzied values for each input
data, whereas ai, bi and ci are MF parameters for
respective representative of middle, width and slope of
the curve as shown in Figure 4.6. These parameters are
varied accordingly to get a suitable curve in order to get
fuzzy signal.
FIGURE 4-6 A basic ANFIS model with two inputs data and two
MFs.
43
output
signal
from
the
fuzzification
stage
( )
( )
(4.12)
( )
( )
(4.14)
=
=
( )
( )
( )
(4.13)
( )
(4.15)
= 1,2,3,4
where, RT = R1 + R2 + R3 + R4
44
(4.16)
4.4.4 Defuzzification
The next process is signal defuzzification in which
the output signal from the normalization stage becomes
an input signal to this defuzzification stage. In this stage,
a normalized signal is gained again through a linear
equation that is formed from the MF of the output signal
as shown in the following equation,
= 1,2,3,4
(4.17)
= 1,2,3,4
OT is a predicted value.
45
(4.18)
intelligent
techniques
such
as
an
artificial
ANFIS
gives
classification
and
high
prediction
precision
models.
in
This
problem
that
involves
new
power
training
data
while
considering
46
CHAPTER 5
Modeling and Simulation of the Underground
Cable
5.1 Cable types and their characteristics [50]
Cable
types
are
basically
defined
as
low-,
211
kV.
Primarily used
to
to
the
cable
type,
different
maintenance
strategy
are
defined.
Three-conductor
improvement
of
the
XLPE
insulation
material
types
of
power
cable
mainly
used
for
distribution purposes in medium voltage range are paperinsulated lead covered (PILC) cables and Cross-linked
Polyethylene (XLPE) cables.
5.2.1 Paper-insulated lead covered (PILC) cables
PILC cables have provided reliable underground
power
delivery
for
decades.
The
PILC
cables
are
applications.
The
designation
6.35/11
kV
for
provides
partial
insulation
as
well
as
Besides
that,
the
XLPE
cables
are
different
MATLAB
is
types
of faults. On
used
for
wavelet
the other
decomposition
hand,
and
variables
of
interest
within
electric
power
was
developed
at
the
Bonneville
Power
as
approximately
5,000
lines
of
code
used
preprocessor
Electromagnetic
to
the
Transients
ATP
version
Program
of
the
(EMTP).
In
components
from
an
extensive
palette,
51
initiated
by
some
disturbances.
Basically,
or
unsymmetrical
disturbances
are
resistances
and
inductances,
hysteretic
with
nonlinearities:
transformers
switches,
time-dependent
and
voltage-
sources:
step,
ramp,
sinusoidal,
surge
analysis,
protective
relay
modeling,
53
scales
certain
selected
on
the
wavelet
features
of
the
transforms,
wavelet
and
then
transformed
54
sine
waves
of
various
frequencies.
Similarly,
techniques,
or
de-noise
wavelet
a
signal
analysis
without
can
often
appreciable
degradation.
In DWT, a time-scale representation of a digital
signal is obtained using digital filtering techniques. In
this case, filters of different cut-off frequencies are used
to analyze the signal at different scales [56]. The signal
is passed through a series of high pass filters to analyze
55
signal
(S)
components
approximation
is
decomposed
approximation
is
(A)
the
high
into
and
two
detail
scale,
types
(D).
of
The
low-frequency
wavelet
decomposition
tree
which
is
shown
in
lower
frequency
components
are
filtered
progressively.
out
is
used
to
simulate
medium
voltage
components
used
are
three
phase
voltage
57
Core conductor
Insulation
i = 1.0 , i = 2.7
Sheath
58
Where:
c : Resistivity of the conductor material.
sh : Resistivity of the sheath material.
c: Relative permeability of the conductor material.
sh: Relative permeability of the sheath material.
i : Relative permeability of the insulating material
outside the conductor.
i : Relative permittivity of the insulating material
outside the conductor.
Three phase voltage source:
V= 11KV with f = 50 Hz
Load:
Three-Phase Grounded-Wye load with parallel R, L
elements:
2 MVA
R = 71.157
L = 365.475mH
59
Fault type
Inception angle
0 , 45 , 90 , 135 ,180
Training
(21)
Fault distance
Testing
(20)
Training
Testing
630
630
630
630
2520
123
180
121
120
544
200 KHz
5e-6 sec
0.1 sec
power frequency
50 Hz
62
64
the
current
for
the
faulted
phase
is
higher
65
FIGURE
5.11
shows
the
single
line
diagram
66
phases
is
higher
68
Discrete
wavelet
transform
for
fault
parameterization
In this study, the line current signals are used as the
input signals of the wavelet analysis. The DWT, using a
Daubechies-4 wavelet (db4), performed better at identifying
the start and end of a disturbance [57- 60]. The Daubechies
wavelet is very well suited for identifying short-time, high
frequency transients, as well as low frequency behavior over
longer periods of time. In both cases the signals are nonperiodic or non-stationary [61].
The fault transients of the study cases are analyzed
through discrete wavelet transform at levels one to five. Both
approximation and details information related fault current are
extracted from the original signal with the multi-resolution
analysis.
When a fault occurs in the cable, it can be seen that
variations within the decomposition coefficient of the current
signals contain useful fault signatures. FIGURE 5-15 shows the
DWT detailed coefficients at level 1 to level 5 for a particular
type of fault studied in the work. Nature of the plot of detailed
coefficients at level 1 shows a sharp spike corresponding to the
fault initiation. According to DWT theory, this spike represents
69
70
71
72
74
1,
claim
fault.
Where
is
75
MATLAB code:
[Ea0,Ed0] = wenergy (C0,L0);
Ed0_max = max (Ed0);
[Eaa,Eda] = wenergy(Ca,La);
Eda_max = max (Eda);
[Eab,Edb] = wenergy(Cb,Lb);
Edb_max = max (Edb);
[Eac,Edc] = wenergy(Cc,Lc);
Edc_max = max (Edc);
Where:
Ea: the percentage of energy corresponding to the
approximation.
Ed: the vector containing
energy corresponding to the details.
2. Construct
the
fault
the
percentages
classification
of
system with
ANFIS:
Data preparation
The training and the testing data sets consisted of
four
for
inputs
(the
maximum
detail
energy
Name = class
Type = sugeno
Number of Inputs=4
Number of Outputs=1
Number of membership functions = 4
Membership function type = generalized bell
Number of Rules=256
77
78
79
MATLAB code:
if Fault==1
FUZ = readfis('class');
clas = evalfis([Ed0_max Eda_max Edb_max Edc_max ], FUZ)
if (clas>0.5)&& (clas<1.5)
disp('Three Phase Fault ')
th_ph_f= 1;
elseif (clas>1.5)&& (clas<2.5)
disp(' Double line to ground Fault between phase A and B')
dlg_ab=1;
elseif (clas>2.5)&& (clas<3.5)
disp(' line to line Fault between phase A and B')
l2l_ab=1;
elseif (clas>3.5)&& (clas<4.5)
disp('Single line to ground Fault at phase A')
slf_A=1;
end
80
Data preparation
The training and the testing data sets consisted of
four
inputs
for
(the
)
maximum
and
the
desired
detail
output
location in Km).
energy
(fault
82
83
84
MATLAB code:
if th_ph_f==1
FUZ = readfis('anfis_3ph');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max ], FUZ)
end
5.8.3.2 Single Line to Ground Fault
Data preparation
The training and the testing data sets consisted of four
inputs (the maximum detail energy for
the desired output (fault location in Km).
85
) and
86
87
MATLAB code:
if slf_A == 1
FUZ = readfis('anfis_slg');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max ],
FUZ)
end
5.8.3.3 Line to Line Fault
Data preparation
The training and the testing data sets consisted of four
inputs (the maximum detail energy for
the desired output (fault location in Km).
) and
89
90
MATLAB code:
if l2l_ab == 1
FUZ = readfis('anfis_l2l');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max ],
FUZ)
end
5.8.3.4 Double Line to Ground Fault
Data preparation
The training and the testing data sets consisted of four
inputs (the maximum detail energy for
) and
the desired output (fault location in Km).
92
93
95
96
FIGURE 5.40: No Fault (normal conditions)
97
FIGURE 5.41: Three phase to ground fault
CHAPTER 6
Simulation Results
6.1 Introduction
For data analysis; location of fault and type of fault is
considered. The input data, sending end current obtained from
the ATP are chosen randomly and are arranged in Microsoft
Office Excel before entered into program design in MATLAB.
Results are analyzed by comparing the real fault location and
the estimation fault location from the network based on the
right type fault type classification.
By using MATLAB ANFIS models are trained for fault
classification and fault location. Once trained, the network
performance was tested using testing data that were different
from the training data.
6 Results
(a) Training
102
6 Results
(b) Testing
FIGURE 6.1: The performance graph for the ANFIS Fault
Type classification model.
classification
Fault Type
Target
L-L-L-G
L-L-G
L-L
L-G
4
103
6 Results
Average accuracy
training
testing
630
630
630
180
120
121
100 %
100 %
100 %
630
123
100 %
=6
104
6 Results
105
6 Results
(a) Training
(b) Testing
FIGURE 6.2: The performance graph for the three phase
ANFIS fault location model
106
6 Results
Table 6.3: Three phase ANFIS fault location training output results
inception angle = 0
Fault
resistance
*X
(KM)
*D
(KM)
*E%
inception angle = 45
inception angle = 90
*X
(KM)
*X
(KM)
*X
(KM)
*X
(KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
0.5
0.502 0.033
1.5
1.504 0.069
0.5
0.500 0.002
1.5
1.501 0.010
0.5
0.499 0.024
0.75
0.748 0.028
1.995 0.088
0.75
0.750 0.007
1.999 0.016
2.75
2.750 0.000
3.002 0.026
3.5
3.497 0.054
3.000 0.001
3.5
3.499 0.013
4.75
4.748 0.025
3.25
3.242 0.129
3.75
3.751 0.011
3.25
3.250 0.001
3.75
3.750 0.004
5.25
5.251 0.025
1.75
1.750 0.001
0.5
0.500 0.001
1.75
1.752 0.038
0.5
0.500 0.007
0.75
0.751 0.011
1.999 0.012
0.75
0.750 0.001
2.002 0.028
0.75
0.748 0.027
1.000 0.005
4.25
4.250 0.002
3.000 0.004
4.25
4.254 0.066
2.998 0.025
3.5
3.500 0.002
4.5
4.500 0.003
3.25
3.250 0.005
4.5
4.504 0.070
3.25
3.250 0.005
5.002 0.028
0.999 0.019
1.5
1.500 0.005
1.003 0.045
1.5
1.495 0.081
1.5
1.497 0.057
1.25
1.251 0.016
2.25
2.252 0.030
1.25
1.248 0.038
2.25
2.249 0.023
2.25
2.250 0.001
5.250 0.007
3.999 0.016
4.997 0.057
3.997 0.051
4.001 0.022
5.25
5.500 0.006
5.5
5.494 0.105
5.25
5.247 0.057
5.5
5.497 0.054
5.25
5.253 0.047
107
10
30
6 Results
inception angle = 0
Fault
resistance
*X
(KM)
*D
(KM)
*E%
inception angle = 45
*X
(KM)
inception angle = 90
Fault
*D
*X
resistan
(KM)
(KM)
*D
(KM)
ce
*E%
Fault
*D
*X
resistan
(KM)
(KM)
*D
(KM)
ce
*E%
1.5
1.498 0.027
2.5
2.502 0.028
1.5
1.477 0.391
2.5
2.500 0.002
1.75
1.748 0.033
2.25
2.250 0.008
2.75
2.718 0.534
2.25
2.269 0.320
2.75
2.746 0.063
1.993 0.116
3.998 0.032
4.75
4.773 0.377
4.026 0.429
4.75
4.751 0.013
4.25
4.256 0.102
5.5
5.510 0.169
5.25
5.212 0.627
5.5
5.468 0.534
5.25
5.246 0.060
5.5
5.487 0.218
2.5
2.494 0.104
1.032 0.536
2.5
2.468 0.536
1.006 0.104
2.5
2.497 0.053
2.75
2.741 0.158
1.25
1.239 0.190
2.75
2.767 0.286
1.25
1.246 0.072
2.994 0.098
4.75
4.755 0.075
5.000 0.001
4.75
4.783 0.554
5.011 0.182
3.25
3.253 0.049
5.25
5.239 0.186
5.25
5.207 0.722
5.25
5.263 0.224
5.25
5.239 0.189
4.5
4.499 0.023
1.5
1.504 0.064
1.75
1.748 0.035
1.5
1.486 0.240
1.75
1.746 0.060
1.5
1.506 0.101
1.997 0.049
1.976 0.399
2.004 0.060
1.998 0.030
2.000 0.001
3.5
3.499 0.020
4.25
4.240 0.159
3.5
3.498 0.031
4.25
4.249 0.019
3.5
3.499 0.022
3.75
3.748 0.032
4.5
4.494 0.098
3.75
3.768 0.299
4.5
4.501 0.011
3.75
3.745 0.084
50
108
100
200
6 Results
1.8
1.6
1.4
Percentage error
1.2
1
0.8
0.6
0.4
0.2
1
28
55
82
109
136
163
190
217
244
271
298
325
352
379
406
433
460
487
514
541
568
595
622
0
Number of Smulation
109
6 Results
(a) Training
(b) Testing
FIGURE 6.4: The training performance graph for the ANFIS
Double Line to Ground Fault Location model.
110
6 Results
Table 6.4: Double Line to Ground (ABG)ANFIS fault location training output results
inception angle = 0
Fault
resistance
inception angle = 45
inception angle = 90
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
0.5
0.500
0.000
1.5
1.497
0.042
0.5
0.500
0.000
1.5
1.499
0.009
0.5
0.500
0.003
0.75
0.735
0.251
2.000
0.000
0.75
0.750
0.003
2.000
0.000
2.75
2.749
0.016
2.965
0.583
3.5
3.502
0.033
3.000
0.002
3.5
3.499
0.012
4.75
4.749
0.012
3.25
3.200
0.837
3.75
3.750
0.008
3.25
3.249
0.015
3.75
3.750
0.004
5.25
5.248
0.036
1.75
1.751
0.012
0.5
0.500
0.000
1.75
1.750
0.001
0.5
0.500
0.000
0.75
0.750
0.003
2.000
0.000
0.75
0.750
0.002
2.000
0.000
0.75
0.750
0.000
1.000
0.000
4.25
4.249
0.010
3.000
0.004
4.25
4.250
0.001
3.249
0.018
3.5
3.499
0.011
4.5
4.500
0.006
3.25
3.250
0.000
4.5
4.511
0.175
3.25
3.498
0.041
5.000
0.000
1.000
0.005
1.5
1.500
0.005
1.000
0.003
1.5
1.504
0.065
1.5
1.500
0.000
1.25
1.253
0.053
2.25
2.249
0.021
1.25
1.242
0.131
2.25
2.250
0.007
2.25
2.250
0.004
5.000
0.005
4.000
0.001
5.000
0.001
4.000
0.000
4.000
0.000
5.25
5.249
0.022
5.5
5.500
0.000
5.25
5.242
0.131
5.5
5.500
0.000
5.25
5.251
0.010
111
10
30
6 Results
inception angle = 0
Fault
resistance
inception angle = 45
inception angle = 90
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
1.5
1.506
0.101
2.5
2.498
0.025
1.5
1.507
0.124
2.5
2.496
0.059
1.75
1.755
0.090
2.25
2.254
0.066
2.75
2.750
0.001
2.25
2.257
0.120
2.75
2.746
0.066
2.000
0.001
4.001
0.012
4.75
4.751
0.020
4.006
0.107
4.75
4.752
0.038
4.25
4.256
0.096
5.5
5.500
0.000
5.25
5.247
0.047
5.5
5.500
0.000
5.25
5.231
0.309
5.5
5.500
0.000
2.5
2.493
0.109
1.000
0.001
2.5
2.506
0.103
1.000
0.000
2.5
2.498
0.036
2.75
2.742
0.131
1.25
1.252
0.041
2.75
2.753
0.052
1.25
1.250
0.001
3.000
0.001
4.75
4.760
0.169
5.000
0.002
4.75
4.745
0.077
5.000
0.006
3.25
3.250
0.003
5.25
5.267
0.284
5.25
5.255
0.085
5.25
5.265
0.251
5.25
5.234
0.268
4.5
4.502
0.028
1.5
1.501
0.015
1.75
1.749
0.021
1.5
1.505
0.082
1.75
1.750
0.008
1.5
1.498
0.027
2.000
0.001
2.000
0.004
2.001
0.020
2.000
0.004
2.000
0.001
3.5
3.495
0.091
4.25
4.249
0.019
3.5
3.488
0.206
4.25
4.251
0.012
3.5
3.505
0.086
3.75
3.745
0.082
4.5
4.501
0.023
3.75
3.741
0.148
4.5
4.502
0.035
3.75
3.753
0.056
50
112
100
200
6 Results
1
0.9
0.8
Percentage error
0.7
0.6
0.5
0.4
0.3
0.2
0
1
31
61
91
121
151
181
211
241
271
301
331
361
391
421
451
481
511
541
571
601
0.1
Number of Smulation
113
6 Results
(a) Training
(b)Testing
FIGURE 6.6: The performance graph for the ANFIS Line to Line Fault
Location model.
114
6 Results
Table 6.5: Line to Line (AB) ANFIS fault location training output results
inception angle = 0
Fault
*X
*D
*E%
resistance (KM) (KM)
inception angle = 45 inception angle = 90 inception angle = 135 inception angle = 180
*X
*D
*E%
(KM) (KM)
*X
*D
*E%
(KM) (KM)
*X
*D
*E%
(KM) (KM)
*X
*D
*E%
(KM) (KM)
0.5
0.502
0.028
1.5
1.498
0.026
0.5
0.498
0.034
1.5
1.500
0.001
0.5
0.499
0.018
0.75
0.747
0.047
2.000
0.005
0.75
0.750
0.006
2.000
0.002
2.75
2.750
0.004
3.003
0.046
3.5
3.500
0.005
2.998
0.030
3.5
3.500
0.005
4.75
4.750
0.005
3.25
3.246
0.065
3.75
3.750
0.000
3.25
3.248
0.040
3.75
3.750
0.000
5.25
5.250
0.002
1.75
1.749
0.015
0.5
0.503
0.047
1.75
1.749
0.018
0.5
0.511
0.185
0.75
0.750
0.003
2.000
0.000
0.75
0.745
0.088
2.000
0.000
0.75
0.734
0.263
1.000
0.000
4.25
4.250
0.002
2.999
0.009
4.25
4.249
0.014
2.992
0.130
3.5
3.509
0.144
4.5
4.501
0.009
3.25
3.236
0.228
4.5
4.506
0.102
3.25
3.254
0.072
5.005
0.082
1.000
0.000
1.5
1.508
0.128
1.000
0.000
1.5
1.500
0.002
1.5
1.501
0.020
1.25
1.228
0.360
2.25
2.250
0.001
1.25
1.236
0.239
2.25
2.250
0.001
1.75
1.728
0.361
4.987
0.223
3.975
0.409
5.012
0.204
4.007
0.124
2.984
0.259
5.25
5.249
0.020
5.5
5.500
0.006
5.25
5.248
0.032
5.5
5.500
0.000
5.25
5.244
0.102
115
10
30
6 Results
inception angle = 0
Fault
*X
*D
*E%
resistance (KM) (KM)
50
116
100
200
inception angle = 45 inception angle = 90 inception angle = 135 inception angle = 180
Fault
*X
*D
*X
*D
resista
*E%
(KM) (KM)
(KM) (KM)
nce
2.500 0.000
1.492 0.129
2.5
1.5
Fault
*X
*D
*X
*D
resista
*E%
(KM) (KM)
(KM) (KM)
nce
2.500 0.000 1.75 1.765 0.245
2.5
1.5
1.480
0.334
2.25
2.250
0.001
2.75
2.743
0.117
2.25
2.252
0.026
2.75
2.743
0.111
2.000
0.000
3.997
0.046
4.75
4.729
0.357
3.995
0.076
4.75
4.755
0.089
4.25
4.251
0.019
5.5
5.500
0.000
5.25
5.252
0.033
5.5
5.500
0.000
5.25
5.251
0.017
5.5
5.500
0.000
2.5
1.000
0.000
1.000
0.000
2.5
2.497
0.044
1.000
0.000
2.5
2.500
0.005
2.75
1.248
0.038
1.25
1.254
0.072
2.75
2.711
0.648
1.25
1.244
0.099
3.021
0.347
4.75
4.727
0.385
4.982
0.308
4.75
4.814
1.063
5.022
0.362
3.25
3.258
0.138
5.25
5.262
0.201
5.25
5.263
0.222
5.25
5.262
0.208
5.25
5.273
0.387
3.993
0.117
1.5
1.477
0.377
1.75
1.759
0.143
1.5
1.498
0.028
1.75
1.747
0.055
1.5
1.522
0.372
1.999
0.024
2.000
0.004
2.000
0.003
1.999
0.013
2.005
0.088
3.5
3.525
0.410
4.25
4.248
0.037
3.5
3.492
0.136
4.25
4.246
0.066
3.5
3.493
0.113
3.75
3.749
0.014
4.5
4.506
0.101
3.75
3.750
0.007
4.5
4.498
0.035
3.75
3.731
0.323
6 Results
2.0
1.8
Percentage error
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
1
35
69
103
137
171
205
239
273
307
341
375
409
443
477
511
545
579
613
0.0
Number of Smulation
117
6 Results
(a)Training
(b) Testing
FIGURE 6.8: The performance graph for the ANFIS Single Line to
Ground (AG) Fault Location model.
118
6 Results
Table 6.6: Single Line to Ground (AG) ANFIS fault location training output results
inception angle = 0
Fault
resistance
119
10
30
*X
(KM)
*D
(KM)
*E%
0.5
0.500
0.002
0.75
0.750
inception angle = 45
inception angle = 90
*X
*D
(KM) (KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
*X
(KM)
*D
(KM)
*E%
1.5
1.494
0.107
0.5
0.520
0.340
1.5
1.498
0.038
0.5
0.497
0.050
0.001
1.994
0.102
0.75
0.835
1.424
2.005
0.087
3.75
3.773
0.384
3.000
0.000
3.5
3.500
0.002
2.946
0.892
3.5
3.493
0.113
4.75
4.712
0.626
3.25
3.265
0.242
3.75
3.807
0.952
3.25
3.241
0.142
3.75
3.817
1.117
5.25
5.265
0.253
1.75
1.751
0.011
0.5
0.500
0.003
1.75
1.750
0.002
0.5
0.500
0.008
0.75
0.750
0.001
2.000
0.006
0.75
0.750
0.003
1.996
0.066
0.75
0.747
0.043
0.999
0.012
4.25
4.250
0.005
3.000
0.003
4.25
4.250
0.002
2.991
0.147
3.5
3.500
0.001
4.5
4.501
0.011
3.25
3.250
0.008
4.5
4.500
0.001
3.25
3.248
0.036
5.000
0.000
0.984
0.262
1.5
1.547
0.787
0.999
0.013
1.5
1.524
0.393
1.5
1.565
1.081
1.25
1.271
0.353
2.25
2.233
0.277
1.25
1.253
0.044
2.75
2.744
0.105
2.25
2.255
0.084
5.000
0.001
4.000
0.002
5.002
0.037
3.992
0.126
3.990
0.174
5.25
5.258
0.141
5.5
5.500
0.001
5.25
5.251
0.012
5.5
5.491
0.155
5.25
5.248
0.026
6 Results
inception angle = 0
Fault
resistance
50
120
100
200
inception angle = 45
inception angle = 90
Fault
*X
*D
*X
resistanc
(KM) (KM)
(KM)
e
2.499 0.021
2.5
1.5
*D
(KM)
*E%
*X
(KM)
1.506
0.096
2.5
Fault
*D
*X
resistan
(KM)
(KM)
ce
2.522 0.374
1.75
2.25
2.245
0.086
2.75
2.756
0.098
0.052
4.009
0.148
4.75
4.737
5.263
0.217
5.5
5.508
0.129
5.25
1.010
0.164
2.5
2.495
0.085
0.267
1.25
1.240
0.173
2.75
2.745
4.763
0.213
5.025
0.417
4.75
5.25
5.298
0.802
5.25
5.277
0.443
1.5
1.464
0.604
1.75
1.762
2.033
0.550
3.5
3.441
0.984
3.75
3.814
1.061
*X
(KM)
*D
(KM)
*E%
*D
(KM)
*E%
1.5
1.482
0.296
1.734
0.264
2.25
2.271
0.348
2.75
2.754
0.066
1.998
0.027
4.017
0.287
4.75
4.747
0.217
4.25
4.251
0.021
5.5
5.485
0.255
5.25
5.278
0.470
5.5
5.488
0.204
2.5
2.530
0.501
1.007
0.109
2.5
2.484
0.260
2.75
2.734
0.077
1.25
1.270
0.326
3.033
0.554
4.75
4.745
0.078
5.033
0.548
3.25
3.267
0.286
5.25
5.253
0.057
5.25
5.206
0.737
4.5
4.471
0.478
0.194
1.5
1.470
0.507
1.75
1.798
0.796
1.5
1.593
1.549
1.995
0.087
2.004
0.066
1.989
0.192
1.952
0.803
4.25
4.274
0.394
3.5
3.486
0.238
4.25
4.254
0.064
3.5
3.594
1.558
4.5
4.497
0.051
3.75
3.760
0.164
4.5
4.382
1.968
3.75
3.769
0.308
6 Results
3.0
Percentage error
2.5
2.0
1.5
1.0
0.5
1
33
65
97
129
161
193
225
257
289
321
353
385
417
449
481
513
545
577
609
0.0
Number of Smulation
121
6 Results
results. From the table, the error produce in estimating the fault
distance is less than 0.7 %, which is a quit good accuracy for the
fault location estimation.
6.4.1.2 Double Line to Ground fault
120 current samples were used to test the double line to ground
ANFIS fault location model, Table 6.8
results. From the table, the error produce in estimating the fault
distance is less than 1.7%, which is a quit good accuracy for the
fault location estimation.
6.4.1.3 Line to Line fault
120 current samples were used to test the line to line ANFIS
fault location model, Table 6.9 represent the testing results. From
the table, the error produce in estimating the fault distance is less
than 1.8%, which is a quit good accuracy for the fault location
estimation.
6.4.1.4 Single Line to Ground fault
6 Results
testing results. From the table, the error produce in estimating the
fault distance is less than 3%, which is a quit good accuracy for
the fault location estimation.
Table 6.7: Three phase ANFIS fault location testing output
results
inception angle = 0
Fault
*X
resistance (KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
0.625
0.625
0.001
0.627
0.041
0.601
0.394
0.875
0.875
0.004
0.875
0.007
0.873
0.026
1.125
1.125
0.005
1.126
0.009
1.126
0.008
1.375
1.375
0.008
1.375
0.002
1.376
0.017
1.625
1.625
0.007
1.625
0.000
1.624
0.019
1.875
1.875
0.008
1.875
0.000
1.877
0.026
2.125
2.125
0.006
2.124
0.009
2.132
0.122
2.375
2.375
0.002
2.375
0.004
2.372
0.043
2.625
2.626
0.010
2.625
0.001
2.626
0.009
2.875
2.875
0.005
2.875
0.002
2.878
0.048
3.125
3.125
0.001
3.125
0.000
3.129
0.064
3.375
3.375
0.001
3.375
0.004
3.374
0.016
3.625
3.624
0.016
3.625
0.004
3.622
0.044
3.875
3.875
0.002
3.875
0.002
3.878
0.047
4.125
4.125
0.000
4.125
0.006
4.127
0.030
4.375
4.375
0.006
4.375
0.002
4.372
0.045
4.625
4.625
0.004
4.625
0.001
4.629
0.063
4.875
4.875
0.002
4.875
0.000
4.882
0.110
10
123
6 Results
inception angle = 0
Fault
*X
resistance (KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
5.125
5.125
0.001
5.125
0.006
5.122
0.051
5.375
5.375
0.004
5.375
0.002
5.380
0.089
0.625
0.631
0.102
0.641
0.261
0.634
0.146
0.875
0.873
0.041
0.851
0.392
0.854
0.350
1.125
1.125
0.008
1.141
0.274
1.140
0.253
1.375
1.376
0.017
1.364
0.182
1.365
0.167
1.625
1.617
0.131
1.619
0.105
1.623
0.031
1.875
1.889
0.237
1.876
0.010
1.875
0.004
2.125
2.123
0.025
2.094
0.518
2.103
0.366
2.375
2.380
0.092
2.388
0.218
2.382
0.122
2.625
2.620
0.080
2.614
0.189
2.616
0.148
2.875
2.871
0.074
2.905
0.503
2.899
0.401
3.125
3.119
0.097
3.130
0.077
3.129
0.066
3.375
3.372
0.057
3.386
0.175
3.404
0.491
3.625
3.608
0.276
3.647
0.359
3.639
0.232
3.875
3.884
0.145
3.874
0.012
3.870
0.076
4.125
4.113
0.196
4.165
0.669
4.138
0.215
4.375
4.364
0.186
4.380
0.087
4.383
0.128
4.625
4.619
0.093
4.626
0.021
4.624
0.012
4.875
4.872
0.057
4.875
0.003
4.873
0.037
5.125
5.127
0.040
5.112
0.219
5.115
0.159
5.375
5.382
0.125
5.357
0.301
5.349
0.427
10
50
0.8
0.7
0.6
0.5
0.4
0.3
10
0.2
0.1
0.0
1
6
11
16
21
26
31
36
41
46
51
56
61
66
71
76
81
86
91
96
101
106
111
116
Percentage error
6 Results
Number of Smulation
Fault
resistance
10
*X
(KM)
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
3.375
3.625
3.875
*D
(KM)
0.630
0.878
1.125
1.377
1.623
1.874
2.124
2.375
2.626
2.873
3.125
3.376
3.623
3.876
*E%
0.086
0.042
0.002
0.035
0.037
0.018
0.012
0.001
0.010
0.040
0.008
0.018
0.029
0.017
125
*D
(KM)
0.626
0.875
1.125
1.376
1.624
1.875
2.125
2.374
2.625
2.875
3.124
3.375
3.627
3.876
*E%
0.016
0.001
0.004
0.014
0.022
0.000
0.007
0.013
0.000
0.003
0.014
0.007
0.025
0.017
*D
(KM)
0.637
0.862
1.124
1.373
1.627
1.875
2.122
2.368
2.618
2.883
3.133
3.373
3.628
3.851
*E%
0.200
0.224
0.013
0.031
0.030
0.005
0.056
0.109
0.110
0.135
0.136
0.027
0.044
0.392
6 Results
inception angle = 0
Fault
resistance
10
200
*X
(KM)
4.125
4.375
4.625
4.875
5.125
5.375
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
3.375
3.625
3.875
4.125
4.375
4.625
4.875
5.125
5.375
*D
(KM)
4.125
4.375
4.623
4.876
5.122
5.363
0.643
0.888
1.132
1.378
1.625
1.872
2.119
2.368
2.617
2.866
3.118
3.368
3.618
3.872
4.123
4.377
4.630
4.882
5.138
5.391
*E%
0.005
0.007
0.033
0.015
0.055
0.201
0.306
0.215
0.113
0.046
0.004
0.056
0.094
0.113
0.130
0.143
0.121
0.117
0.111
0.042
0.025
0.033
0.091
0.119
0.212
0.268
*D
(KM)
4.125
4.377
4.624
4.874
5.126
5.377
0.638
0.885
1.131
1.377
1.626
1.874
2.121
2.371
2.620
2.868
3.119
3.370
3.619
3.872
4.123
4.375
4.627
4.880
5.133
5.386
126
*E%
0.005
0.028
0.011
0.024
0.012
0.037
0.219
0.163
0.101
0.034
0.024
0.017
0.064
0.067
0.078
0.109
0.092
0.086
0.107
0.057
0.028
0.008
0.037
0.091
0.139
0.188
*D
(KM)
4.122
4.366
4.627
4.898
5.107
5.384
0.672
0.909
1.146
1.384
1.627
1.870
2.114
2.361
2.610
2.858
3.109
3.361
3.613
3.868
4.123
4.379
4.634
4.890
5.148
5.404
*E%
0.048
0.142
0.040
0.375
0.299
0.156
0.791
0.560
0.346
0.152
0.026
0.089
0.183
0.230
0.254
0.279
0.260
0.225
0.196
0.112
0.034
0.063
0.147
0.255
0.388
0.486
6 Results
1
6
11
16
21
26
31
36
41
46
51
56
61
66
71
76
81
86
91
96
101
106
111
116
Percentage error
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Number of Smulation
Fault
resistance
10
*X
(KM)
0.625
*D
(KM)
0.609
0.875
inception angle =
45
0.263
*D
(KM)
0.723
0.866
0.156
1.125
1.226
1.375
1.638
*D
(KM)
0.590
0.591
0.936
1.021
0.861
0.241
1.678
1.157
0.530
1.114
0.191
1.341
0.567
1.371
0.070
1.373
0.026
1.625
1.571
0.893
1.611
0.233
1.631
0.092
1.875
1.890
0.250
1.846
0.483
1.902
0.458
2.125
2.152
0.444
2.088
0.612
2.128
0.046
2.375
2.331
0.742
2.340
0.578
2.375
0.006
2.625
2.660
0.577
2.647
0.372
2.622
0.047
2.875
2.892
0.285
2.853
0.361
2.874
0.024
3.125
3.136
0.180
3.115
0.166
3.126
0.011
3.375
3.394
0.310
3.403
0.470
3.367
0.138
*E%
127
*E%
*E%
6 Results
inception angle = 0
Fault
resistance
10
200
*X
(KM)
3.625
3.875
4.125
4.375
4.625
4.875
5.125
5.375
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
3.375
3.625
3.875
4.125
4.375
4.625
4.875
5.125
5.375
*D
(KM)
3.565
3.885
4.128
4.357
4.627
4.863
5.120
5.310
0.577
0.858
1.127
1.347
1.599
1.864
2.117
2.376
2.638
2.886
3.144
3.398
3.641
3.903
4.151
4.397
4.653
4.893
5.118
5.350
*E%
1.008
0.165
0.055
0.293
0.036
0.200
0.076
1.080
0.800
0.284
0.030
0.474
0.439
0.182
0.131
0.021
0.215
0.180
0.313
0.381
0.260
0.459
0.432
0.371
0.470
0.295
0.118
0.416
inception angle =
45
*D
(KM)
3.583
3.912
4.160
4.340
4.642
4.862
5.101
5.372
0.659
0.643
0.886
1.127
1.381
1.627
1.888
2.135
2.380
2.633
2.867
3.125
3.383
3.612
3.875
4.123
4.354
4.617
4.879
5.097
128
*E%
0.701
0.613
0.576
0.583
0.288
0.222
0.404
0.049
0.571
0.294
0.191
0.030
0.093
0.040
0.221
0.159
0.078
0.138
0.134
0.005
0.138
0.214
0.004
0.038
0.357
0.127
0.059
0.466
*D
(KM)
3.617
3.867
4.120
4.381
4.624
4.889
5.117
5.308
0.642
0.876
1.115
1.376
1.622
1.872
2.115
2.361
2.615
2.856
3.111
3.366
3.607
3.873
4.127
4.375
4.636
4.888
5.141
5.400
*E%
0.139
0.127
0.076
0.102
0.019
0.226
0.128
1.119
0.279
0.025
0.173
0.010
0.044
0.048
0.169
0.240
0.170
0.321
0.233
0.143
0.299
0.032
0.040
0.006
0.188
0.209
0.266
0.409
6 Results
1.8
1.6
Percentage error
1.4
1.2
1.0
0.8
0.6
0.4
0.2
1
7
13
19
25
31
37
43
49
55
61
67
73
79
85
91
97
103
109
115
0.0
Number of Smulation
Fault
resistance
*X
(KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
10
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
0.623
0.874
1.126
1.375
1.627
1.874
2.123
2.375
2.625
2.875
3.125
0.025
0.018
0.024
0.002
0.027
0.012
0.040
0.002
0.003
0.005
0.006
0.625
0.875
1.124
1.375
1.624
1.875
2.125
2.375
2.625
2.875
3.125
0.006
0.006
0.010
0.003
0.011
0.006
0.005
0.003
0.004
0.002
0.004
0.625
0.872
1.133
1.373
1.624
1.879
2.116
2.371
2.631
2.875
3.128
0.005
0.044
0.133
0.036
0.021
0.062
0.153
0.059
0.107
0.001
0.049
129
6 Results
inception angle = 0
Fault
resistance
10
50
inception angle =
45
inception angle =
90
*X
(KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
3.375
3.625
3.875
4.125
4.375
4.625
4.875
5.125
5.375
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
3.375
3.625
3.875
4.125
4.375
4.625
4.875
5.125
3.376
3.624
3.875
4.128
4.372
4.628
4.871
5.123
5.376
0.559
0.826
1.091
1.368
1.620
1.885
2.143
2.390
2.655
2.902
3.161
3.412
3.646
3.914
4.144
4.377
4.632
4.848
5.087
0.020
0.023
0.002
0.050
0.042
0.053
0.072
0.031
0.019
1.105
0.815
0.570
0.118
0.086
0.170
0.307
0.249
0.500
0.457
0.604
0.625
0.355
0.651
0.322
0.037
0.122
0.442
0.641
3.375
3.625
3.874
4.127
4.375
4.625
4.876
5.122
5.376
0.629
0.875
1.122
1.379
1.630
1.880
2.130
2.381
2.631
2.868
3.161
3.379
3.611
3.889
4.132
4.367
4.634
4.871
5.125
0.005
0.006
0.011
0.032
0.001
0.002
0.015
0.047
0.024
0.075
0.006
0.058
0.070
0.086
0.089
0.080
0.097
0.101
0.121
0.604
0.067
0.226
0.232
0.123
0.137
0.142
0.073
0.002
3.379
3.621
3.890
4.104
4.358
4.624
4.798
5.099
5.374
0.620
0.892
1.123
1.389
1.621
1.868
2.128
2.371
2.622
2.871
3.138
3.382
3.629
3.909
4.127
4.376
4.637
4.833
5.091
0.060
0.059
0.245
0.345
0.285
0.010
1.276
0.433
0.019
0.081
0.288
0.040
0.227
0.064
0.124
0.050
0.065
0.049
0.066
0.215
0.109
0.063
0.566
0.036
0.015
0.192
0.702
0.564
130
6 Results
inception angle = 0
Fault
resistance
200
inception angle =
45
inception angle =
90
*X
(KM)
*D
(KM)
*E%
*D
(KM)
*E%
*D
(KM)
*E%
5.375
0.625
0.875
1.125
1.375
1.625
1.875
2.125
2.375
2.625
2.875
3.125
3.375
3.625
3.875
4.125
4.375
4.625
4.875
5.125
5.375
5.320
0.538
0.739
1.033
1.356
1.599
1.885
2.150
2.371
2.681
2.918
3.174
3.448
3.650
3.933
4.167
4.375
4.639
4.833
5.068
5.290
0.911
1.445
2.264
1.531
0.318
0.439
0.163
0.418
0.068
0.928
0.713
0.813
1.222
0.424
0.973
0.694
0.007
0.226
0.701
0.944
1.411
5.356
0.670
0.882
1.115
1.386
1.627
1.868
2.129
2.365
2.620
2.872
3.130
3.379
3.624
3.900
4.130
4.378
4.637
4.843
5.101
5.336
0.320
0.753
0.116
0.174
0.188
0.041
0.122
0.061
0.166
0.091
0.050
0.078
0.065
0.016
0.418
0.083
0.047
0.195
0.541
0.402
0.648
5.350
0.463
0.895
1.128
1.401
1.628
1.871
2.135
2.369
2.620
2.880
3.138
3.381
3.639
3.910
4.129
4.379
4.634
4.826
5.072
5.275
0.422
2.704
0.341
0.056
0.434
0.057
0.070
0.162
0.094
0.082
0.088
0.220
0.096
0.233
0.576
0.072
0.065
0.146
0.823
0.875
1.668
131
6 Results
Percentage error
2.5
2
1.5
1
1
9
17
25
33
41
49
57
65
73
81
89
97
105
113
121
129
137
145
153
161
169
177
0.5
Number of Smulation
132
CHAPTER 7
Conclusion And Recommendation
7.1 Conclusion
Nowadays,
electrical
energy
is
being
increasingly
begin,
this
approach
is
based
on
computer
Models
were
trained
and
then
tested
with
7.2 Recommendation
Since this study had only focused on distribution
level of voltage, it is recommended that further studies be
carried out on variety level of voltage in power system. The
network should be able to perform in any case of power
system. More studies can be done on the effect of cable
aging on fault location and for locating faults in a combined
overhead transmission line with underground power cable.
Moreover,
to
increase
accuracy
in
results,
it
is
134
REFERENCES
REFERENCES
[1]
135
REFERENCES
[5]
136
REFERENCES
[10] Tamer S. Kamel M. A. Moustafa Hassan, "Adaptive
Neuro Fuzzy Inference System (ANFIS) For Fault
Classification in the Transmission Lines", the
Online Journal on Electronics and Electrical
Engineering (OJEEE), Vol. (2), No. (1), January,
2010.
[11] J. Zhang, Z.Y. He, S. Lin, Y.B. Zhang, Q.Q. Qian,
"An ANFIS-based fault classification approach in
power distribution system ", Electrical Power and
Energy Systems, pp. 243252, vol. 49 (2013).
[12] S.M. Yeo, C.H. Kim, K.S. Hong, Y.B. Lim, R.K.
Aggarwal, A.T. Johns, M.S. Choi, " A novel
algorithm for fault classification in transmission
lines using a combined adaptive network and fuzzy
inference system ", Electrical Power and Energy
Systems, pp. 747758, vol. 25 (2003).
[13]M.Mirzaei, M.Z. A Ab Kadir, E.Moazami, H.Hizam,
Review of fault location methods for distribution
power system, Australian Journal of Basic and
Applied Sciences, 3(3), page(s): 2670-2676, 2009.
[14]Adly
A.Girgis,
Christopher
M.Fallon,
David
L.Lubkeman, A fault location technique for rural
distribution feeders, IEEE Transactions on Industry
Applicatuions, Vol.29, No.6, Page(s):1170-1175,
1993
[15]Santoso, S., Dugan, R.C., Lamoree, J., Sundaram,
A., Distance estimation technique for single line137
REFERENCES
to-ground faults in a radial distribution system,
IEEE Power Engineering Society Winter Meeting,
Vol.4, Page(s): 2551-2555, 2000
[16] Das, R., Sachdev, M.S., Sidhu, T.S., A fault
locator for radial subtransmission and distribution
lines, IEEE Power Engineering Society Summer
Meeting, Vol.1, Page(s): 443-448, 2000
[17] Myeon-Song Choi, Seung-Jae Lee, Duck-Su Lee,
Bo-Gun Jin, A new fault location algorithm using
direct circuit analysis for distribution systems,
IEEE Transactions on Power Delivery, Vol.19,
No.1, Page(s):35-41, 2004.
[18] A.H.A. Bakar, M.S. Ali, ChiaKwang Tan, H.
Mokhlis, H. Arof, H.A. Illias, " High impedance
fault location in 11 kV underground distribution
systems using wavelet transforms", International
Journal of Electrical Power & Energy Systems,
Volume 55, Pages 723-730, February 2014.
[42] Wanjing Xiu, Yuan Liao, "Novel fault location
methods for ungrounded radial distribution systems
using measurements at substation ", Electric Power
Systems Research, Volume 106, Pages 95-100,
January 2014.
[43]
REFERENCES
Developments in Power System Protection, Vol.1,
Page(s): 200-203, 2004
[21] Magnago, F.H., Abur, A., Fault Location using
wavelets, IEEE Transactions on Power Deleivery,
Vol.13, No.4, Page(s): 1475-1480, 1998
[22] Magnago, F.H., Abur, A., A new fault location
technique for radial distribution systems based on
high frequency signals, IEEE Power Engineering
Society Summer Meeting, Vol.1, Page(s): 426-431,
1999
[23] Thomas, D.W.P., Christopoulos, C., Tang, Y., Gale,
P., Stokoe, J., Single ended travelling wave fault
location scheme based on wavelet analysis, Eighth
IEE International Conference on Developments in
Power System Protection, Vol.1, Page(s): 196-199,
2004
[24] Chul-Hwan Kim, Woo-Gon Chung, Myong-Chul
Shin, Sung-Wan Park, Myong-Hee Lee, Yong-Pil
Seo, Hae-Sul Choi, A study on the expert system
for fault location estimation in underground cable,
International Conference on Energy Management
and Power Delivery, Proceedings of EMPD, Vol.1,
Page(s): 67-72, 1995
[25] Ismail Niazy, Javad Sadeh, " A new single ended
fault location algorithm for combined transmission
line considering fault clearing transients without
using line parameters ", International Journal of
139
REFERENCES
Electrical Power & Energy Systems, Volume 44,
Issue 1, Pages 816-823, January 2013.
[26] H.Khorashadi-Zadeh, M.R.Aghaebrahimi, A novel
approach to fault classification and fault location for
medium voltage cables based on artificial neural
network, International Journal of Computational
Intelligence, Vol.2, No.2, pp. 90-93, 2005
[27]
[28]
REFERENCES
[31] H.H. Sherwali, P.A.Crossley, Expert system for
fault location and protection analysis, Fourth
International Conference on Power System Control
and Management, Conference Publication No.421,
Page(s): 181-186, 1996
[32] K.K. Kuan, K.Warwick, Real-time expert system
for fault location on high voltage underground
distribution cables, IEE Proceedings C, Generation,
Transmission and Distribution, Vol.139, No.3,
Page(s): 235- 240, 1992
[33] J. Moshtagh, R. K. Aggarwal, A new approach to
fault location in a single core underground cable
system using combined fuzzy logic & wavelet
analysis, The Eight IEE International Conference
on Developments In Power System Protection, pp.
228-231, April 2004.
[34] Javad Sadeh, Hamid Afradi, "A new and accurate
fault location algorithm for combined transmission
lines
using
Adaptive
Network-Based
Fuzzy
Inference
System",
Electric
Power
Systems
Research, pp. 15381545, vol. 79 (2009).
[57] J. J. Mora, G. Carrillo, "
Distribution Systems using
Patterns ", 2006 IEEE
Distribution Conference
America, Venezuela, 2006.
141
REFERENCES
[36] Huseyin Eristi, "Fault diagnosis system for series
compensated transmission line based on wavelet
transform
and
adaptive
neuro-fuzzy
inference
system ", pp. 393401, Measurement vol.46 (2013).
[37] Rasli A Ghani, MohdFauzi Othman, A.M. Hashim, "
ANFIS approach for locating precise fault points in
distribution system ", International Journal of
Emerging Technology and Advanced Engineering ,
Volume 2, Issue 6, June 2012.
[38] T.A Short, Electric power distribution handbook,
CRC PRESS, 2004.
[39] Johnston, L., Tweed, N. B., Ward, D. J., and Burke,
J. J., An Analysis of Vepcos 34.5 kV Distribution
Feeder Faults as Related to Through Fault Failures
of Substation Transformers, IEEE Transactions on
Power Apparatus and Systems, vol. PAS-97, no. 5,
pp. 187684, 1978.
[40] Mohd Wazir Mustafa, Power System Analysis,
Second edition, 2009.
[41] Michael Negnevitsky. Artificil Intelligent: A Guide
to Intelligent System, Addison-Wesley, ISBN 0321-20466-2, Harlow, England, (2005).
[42] L. A. Zadeh. Fuzzy sets, Information and Control 8,
338-353, (1965).
[43] J. Lukasiewicz. O logice trjwartociowej, Ruch
Filozoficzny 5, 170f, (1920).
142
REFERENCES
[44] L. A. Zadeh. A rationale for Fuzzy Control, Journal
of Dynamic Systems, Measurement, Control 94 (6),
3-4, (1972).
[45]
J. J. Buckley. Universal
Automatica, 28:1245-1248, (1992).
fuzzy
[46]
controllers.
Toolbox
REFERENCES
[52] Canadian/American EMTP User Group, Alternative
Transients Program (ATP) Rule Book, CoChairmen W. S. Meyer and T. Liu, Portland,
Oregon, 1995.
[53] Gunther, E., Grebe, T., Rambabu, A., and Mader, D.,
Running EMTP on PCs, IEEE Computer Applications in
Power, pp. 33-38, Jan. 1993.
[54] Lszl Prikler, Hans Kristian Hidalen, "ATPDRAW
Users' Manual", Preliminary Release No. 1.0 November
2009.
[55]
REFERENCES
Transmission & Distribution, Vol. 2, No. 1, pp. 110 - 118,
Jan. 2008.
[60] Zhaoa W, Songa YH, Minb Y. Wavelet analysis based
scheme for fault detection and classification in
underground power cable systems. Electric Power System
Research;53(1):2330, 2000.
[61] K. Chul Hwan and A. Raj, "Wavelet transforms in power
systems. I. General introduction to the wavelet
transforms," Power Engineering Journal [see also Power
Engineer], vol. 14, pp. 81-87, 2000.
[62] "IEEE Guide for Determining Fault Location on AC
Transmission and Distribution Lines", Jun. 2005, IEEE
Standard C37.114, (2004).
145
Appendix A
A.1 Development of the fault location system tool
(FL)
function varargout = fl(varargin)
gui_Singleton = 1;
gui_State = struct('gui_Name', mfilename,
...
'gui_Singleton',
gui_Singleton, ...
'gui_OpeningFcn',
@fl_OpeningFcn, ...
'gui_OutputFcn',
@fl_OutputFcn, ...
'gui_LayoutFcn', [] , ...
'gui_Callback',
[]);
if nargin && ischar(varargin{1})
gui_State.gui_Callback =
str2func(varargin{1});
end
if nargout
[varargout{1:nargout}] =
gui_mainfcn(gui_State, varargin{:});
else
gui_mainfcn(gui_State, varargin{:});
end
% --- Executes on button press in delete.
function delete_Callback(hObject, eventdata,
handles)
% hObject
handle to delete (see GCBO)
% eventdata reserved - to be defined in a
future version of MATLAB
146
Appendix A
% handles
structure with handles and user
data (see GUIDATA)
axes(handles.axes1); cla reset
axes(handles.axes2); cla reset
axes(handles.axes3); cla reset
set(handles.t1,'String',' ')
axes(handles.axes7); cla reset
axes(handles.axes8); cla reset
axes(handles.axes9); cla reset
set(handles.t2,'String',' ')
set(handles.t4,'String',' ')
set(handles.et1,'String',' ')
set(handles.t_error,'String',' ')
% --- Executes just before fl is made
visible.
function fl_OpeningFcn(hObject, eventdata,
handles, varargin)
handles.output = hObject;
guidata(hObject, handles);
% --- Outputs from this function are returned
to the command line.
function varargout = fl_OutputFcn(hObject,
eventdata, handles)
% varargout cell array for returning output
args (see
% Get default command line output from
handles structure
varargout{1} = handles.output;
% --- Executes on button press in b.
function b_Callback(hObject, eventdata,
handles)
147
Appendix A
% hObject
handle to b (see GCBO)
global th_ph_f;
global dlg_ab;
global l2l_ab;
global slf_A;
global D;
th_ph_f=0; dlg_ab=0; l2l_ab=0; slf_A=0; D=0;
file = uigetfile('*.csv');
newData1 = importdata(file);
% Create new variables in the base workspace
from those fields.
vars = fieldnames(newData1);
for i = 1:length(vars)
assignin('base', vars{i},
newData1.(vars{i}));
end
data = newData1.(vars{1});
t=data(1:20000,1);
A =data(1:20000,2);
B =data(1:20000,3);
C =data(1:20000,4);
I0 = (1/3)* (A+B+C);
I1 = (1/3)*(2*A - B -C);
I2 = (1/3)*(sqrt(3)*B - sqrt(3)*C);
% Perform a multilevel wavelet decomposition
of I0 at level 5
[C0,L0] = wavedec(I0,5,'db4');
[Ea0,Ed0] = wenergy(C0,L0);
148
Appendix A
% Extract approximation and detail
coefficients.
c0A5
c0A4
c0A3
c0A2
c0A1
=
=
=
=
=
appcoef(C0,L0,'db4',5);
appcoef(C0,L0,'db4',4);
appcoef(C0,L0,'db4',3);
appcoef(C0,L0,'db4',2);
appcoef(C0,L0,'db4',1);
c0D5
c0D4
c0D3
c0D2
c0D1
=
=
=
=
=
detcoef(C0,L0,5);
detcoef(C0,L0,4);
detcoef(C0,L0,3);
detcoef(C0,L0,2);
detcoef(C0,L0,1);
=
=
=
=
=
wrcoef('d',C0,L0,'db4',1);
wrcoef('d',C0,L0,'db4',2);
wrcoef('d',C0,L0,'db4',3);
wrcoef('d',C0,L0,'db4',4);
wrcoef('d',C0,L0,'db4',5);
sum_s_D01=sum(s_D01);
sum_s_D02=sum(s_D02);
sum_s_D03=sum(s_D03);
sum_s_D04=sum(s_D04);
149
Appendix A
s_D05=D05.*D05;
sum_s_D05=sum(s_D05);
=
=
=
=
=
appcoef(Ca,La,'db4',5);
appcoef(Ca,La,'db4',4);
appcoef(Ca,La,'db4',3);
appcoef(Ca,La,'db4',2);
appcoef(Ca,La,'db4',1);
caD5
caD4
caD3
caD2
caD1
=
=
=
=
=
detcoef(Ca,La,5);
detcoef(Ca,La,4);
detcoef(Ca,La,3);
detcoef(Ca,La,2);
detcoef(Ca,La,1);
=
=
=
=
wrcoef('a',Ca,La,'db4',11);
wrcoef('d',Ca,La,'db4',1);
wrcoef('d',Ca,La,'db4',2);
wrcoef('d',Ca,La,'db4',3);
150
Appendix A
Da4 = wrcoef('d',Ca,La,'db4',4);
Da5 = wrcoef('d',Ca,La,'db4',5);
%Energy for 1-D wavelet or wavelet packet
decomposition
[Eaa,Eda] = wenergy(Ca,La);
Eda_max = max(Eda);
% Evaluation of the square of the wavelet
coefficients
s_Da1=Da1.*Da1;
sum_s_Da1=sum(s_Da1);
s_Da2=Da2.*Da2;
sum_s_Da2=sum(s_Da2);
s_Da3=Da3.*Da3;
sum_s_Da3=sum(s_Da3);
s_Da4=Da4.*Da4;
sum_s_Da4=sum(s_Da4);
s_Da5=Da5.*Da5;
sum_s_Da5=sum(s_Da5);
% caculation of the disotrted signal
energy%%%%
da1=sum_s_Da1;
da2=da1+sum_s_Da2;
da3=da2+sum_s_Da3;
da4=da3+sum_s_Da4;
da5=da4+sum_s_Da5;
=
=
=
=
appcoef(Cb,Lb,'db4',5);
appcoef(Cb,Lb,'db4',4);
appcoef(Cb,Lb,'db4',3);
appcoef(Cb,Lb,'db4',2);
151
Appendix A
cbA1 = appcoef(Cb,Lb,'db4',1);
cbD5
cbD4
cbD3
cbD2
cbD1
=
=
=
=
=
detcoef(Cb,Lb,5);
detcoef(Cb,Lb,4);
detcoef(Cb,Lb,3);
detcoef(Cb,Lb,2);
detcoef(Cb,Lb,1);
=
=
=
=
=
wrcoef('d',Cb,Lb,'db4',1);
wrcoef('d',Cb,Lb,'db4',2);
wrcoef('d',Cb,Lb,'db4',3);
wrcoef('d',Cb,Lb,'db4',4);
wrcoef('d',Cb,Lb,'db4',5);
sum_s_Db1=sum(s_Db1);
sum_s_Db2=sum(s_Db2);
sum_s_Db3=sum(s_Db3);
sum_s_Db4=sum(s_Db4);
sum_s_Db5=sum(s_Db5);
152
Appendix A
db2=db1+sum_s_Db2;
db3=db2+sum_s_Db3;
db4=db3+sum_s_Db4;
db5=db4+sum_s_Db5;
% Perform a multilevel wavelet decomposition
of phase C at level 5
[Cc,Lc] = wavedec(C,5,'db4');
% Extract approximation and detail
coefficients.
ccA5
ccA4
ccA3
ccA2
ccA1
=
=
=
=
=
appcoef(Cc,Lc,'db4',5);
appcoef(Cc,Lc,'db4',4);
appcoef(Cc,Lc,'db4',3);
appcoef(Cc,Lc,'db4',2);
appcoef(Cc,Lc,'db4',1);
ccD5
ccD4
ccD3
ccD2
ccD1
=
=
=
=
=
detcoef(Cc,Lc,5);
detcoef(Cc,Lc,4);
detcoef(Cc,Lc,3);
detcoef(Cc,Lc,2);
detcoef(Cc,Lc,1);
=
=
=
=
=
wrcoef('d',Cc,Lc,'db4',1);
wrcoef('d',Cc,Lc,'db4',2);
wrcoef('d',Cc,Lc,'db4',3);
wrcoef('d',Cc,Lc,'db4',4);
wrcoef('d',Cc,Lc,'db4',5);
153
Appendix A
%Energy for 1-D wavelet or wavelet packet
decomposition
[Eac,Edc] = wenergy(Cc,Lc);
Edc_max = max(Edc);
% Evaluation of the square of the wavelet
coefficients
s_Dc1=Dc1.*Dc1;
s_Dc2=Dc2.*Dc2;
s_Dc3=Dc3.*Dc3;
s_Dc4=Dc4.*Dc4;
s_Dc5=Dc5.*Dc5;
sum_s_Dc1=sum(s_Dc1);
sum_s_Dc2=sum(s_Dc2);
sum_s_Dc3=sum(s_Dc3);
sum_s_Dc4=sum(s_Dc4);
sum_s_Dc5=sum(s_Dc5);
axes(handles.axes5);
imshow('panel_b.png')
154
Appendix A
axes(handles.axes6);
imshow('panel_c.png')
set(handles.t1,'String',file)
axes(handles.axes7); plot(t,Da5);
axes(handles.axes8); plot(t,Db5);
axes(handles.axes9); plot(t,Db5);
% no fault
if (da5<15) && (db5<15) && (dc5<15)
set(handles.t2,'String','No Fault')
Fault=0;
else
Fault=1;
end
%%%%%%
Fault Classification
%%%%%
if Fault==1
FUZ = readfis('class');
clas = evalfis([Ed0_max Eda_max Edb_max
Edc_max ], FUZ)
if (clas>0.5)&& (clas<1.5)
set(handles.t2,'String','Three Phase
Fault')
th_ph_f= 1;
elseif (clas>1.5)&& (clas<2.5)
set(handles.t2,'String','Double line to
ground Fault between phase A and B')
155
Appendix A
dlg_ab=1;
elseif (clas>2.5)&& (clas<3.5)
set(handles.t2,'String','line to line
Fault between phase A and B')
l2l_ab=1;
elseif (clas>3.5)&& (clas<4.5)
set(handles.t2,'String','Single line to
ground Fault at phase A')
slf_A=1;
else
set(handles.t2,'String','Wrong')
end
%%%%%%%%
%%%%%%%%%%%%%%%
Fault location
if th_ph_f==1
FUZ = readfis('anfis_3ph');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max
], FUZ)
set(handles.t4,'String',D);
end
if slf_A == 1
FUZ = readfis('anfis_slg');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max
], FUZ)
set(handles.t4,'String',D);
end
if l2l_ab == 1
156
Appendix A
FUZ = readfis('anfis_l2l');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max
], FUZ)
set(handles.t4,'String',D);
end
if dlg_ab ==1
FUZ = readfis('anfis_dlg');
D = evalfis([Ed0_max Eda_max Edb_max Edc_max
], FUZ)
set(handles.t4,'String',D);
end
end
function et1_Callback(hObject, eventdata,
handles)
function et1_CreateFcn(hObject, eventdata,
handles)
if ispc &&
isequal(get(hObject,'BackgroundColor'),
get(0,'defaultUicontrolBackgroundColor'))
set(hObject,'BackgroundColor','white');
end
% --- Executes on button press in calc_error.
function calc_error_Callback(hObject,
eventdata, handles)
d = get(handles.t4,'String');
d_val = str2double(d);
AFL = get(handles.et1,'String');
AFL_val = str2double(AFL);
error = 100 * abs(AFL_val-d_val)/6 ;
set(handles.t_error,'String',error);
157
Appendix A
A.2 Creating and Training the ANFIS MODEL
files_names = dir(fullfile(matlabroot,
'anfis'));
no_files = length (files_names);
for nf=3:no_files
fp = files_names(nf).name;
savefile = 'trnError_3ph.csv';
file_Path = ['C:\Program
Files\MATLAB2\R2011a\anfis\' , fp];
importfile_anfis(file_Path)
trnData = train_3ph;
numMFs = 5;
mfType = 'gbellmf';
in_fis = genfis1(trnData,numMFs,mfType);
[out_fis,trnError_class] = anfis(trnData,
in_fis,300);
plot(trnError_class , '*');grid;
xlabel('Epochs');
ylabel('RMSE (Root Mean Squared Error)');
title('Error Curves');
figure(2)
out_fis_o = evalfis(trnData(:,1:4),out_fis);
plot([out_fis_o ,trnData(:,5)], 'x');
end
158
Appendix B
ANFIS Fault Type classification Training Results
B.1 Three phase ANFIS Fault Type classification Training
Results.
Fault
resistance
()
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1.004
1.004
1.002
1.002
1.001
1.000
0.999
0.998
0.998
0.997
0.997
0.997
0.997
0.997
0.999
0.998
1.000
1.001
1.001
1.005
1.005
0.978
0.988
0.992
1.000
1.008
1.014
1.019
1.021
1.022
1.019
1.014
1.008
0.998
0.990
0.983
0.978
0.978
0.981
0.990
1.006
1.025
0.977
1.001
1.008
1.010
1.012
1.009
1.005
1.003
0.999
0.996
0.995
0.995
0.994
0.995
0.997
0.997
0.999
1.001
1.001
1.002
1.003
1.054
1.022
0.991
0.964
0.959
0.958
0.960
0.976
0.989
1.006
1.017
1.028
1.035
1.039
1.044
1.035
1.024
1.012
0.989
0.963
0.935
1.021
1.012
1.003
0.995
0.992
0.990
0.989
0.991
0.992
0.994
0.997
1.000
1.003
1.005
1.007
1.007
1.007
1.005
1.002
0.999
0.991
* IA : Inception Angle
159
Appendix B
Fault
resistance
()
10
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.970
0.994
1.005
1.012
1.013
1.012
1.010
1.003
1.000
0.998
0.996
0.996
0.994
0.995
0.997
0.998
1.000
1.001
1.002
1.003
1.003
0.994
1.004
1.003
1.002
1.000
0.998
1.000
0.999
1.000
1.001
1.001
1.001
0.999
0.997
0.996
0.995
0.997
0.998
1.001
1.004
1.009
1.009
1.005
0.998
0.992
0.990
0.989
0.991
0.995
0.999
1.003
1.008
1.011
1.012
1.012
1.013
1.010
1.006
1.001
0.993
0.986
0.977
0.967
0.975
0.987
0.999
1.009
1.016
1.019
1.018
1.015
1.009
1.002
0.995
0.989
0.985
0.984
0.985
0.989
0.995
1.003
1.012
1.020
1.012
1.009
1.000
0.998
0.994
0.995
1.000
0.991
0.992
0.999
0.998
1.005
0.999
1.002
1.003
1.000
1.004
1.002
0.998
1.002
1.001
* IA : Inception Angle
160
Appendix B
Fault
resistance
()
30
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1.012
1.008
1.005
1.002
0.999
0.997
0.995
0.994
0.993
0.993
0.993
0.993
0.994
0.995
0.997
0.999
1.001
1.003
1.005
1.010
1.013
1.000
1.000
1.000
1.001
1.002
1.002
1.003
1.001
0.998
0.999
1.000
0.997
0.993
0.995
0.994
0.993
1.001
1.003
0.971
1.021
1.026
0.980
0.985
0.993
0.995
1.003
1.008
1.008
1.013
1.011
1.012
1.015
1.011
1.010
1.006
1.006
1.002
1.001
0.996
0.961
0.989
0.984
0.999
0.999
1.000
0.999
1.000
1.000
0.999
1.002
1.001
1.002
1.002
1.003
1.003
1.003
1.004
1.003
1.004
1.003
0.996
0.995
1.001
0.996
0.997
0.997
0.997
0.997
0.998
0.998
0.999
0.999
0.999
0.999
1.000
1.001
1.001
1.002
1.002
1.003
1.003
1.002
1.003
1.005
* IA : Inception Angle
161
Appendix B
Fault
resistance
()
50
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.893
0.966
0.983
0.994
1.007
1.019
1.022
1.030
1.036
1.032
1.036
1.034
1.030
1.025
1.022
1.009
1.002
0.991
0.938
0.962
0.950
0.994
0.996
0.996
0.997
1.002
0.999
1.002
0.999
1.002
1.005
1.002
1.004
1.002
1.004
1.005
1.004
1.006
1.003
0.967
1.005
1.004
1.001
1.003
1.005
1.006
1.008
1.008
1.009
1.008
1.008
1.008
1.006
1.006
1.003
1.002
0.999
0.997
0.995
0.992
0.976
0.986
0.983
1.010
1.006
1.004
1.001
0.998
0.999
0.995
0.994
0.996
0.993
0.995
0.996
0.995
0.998
1.000
1.000
1.004
1.007
0.985
0.988
1.016
0.990
0.993
0.996
0.999
1.001
1.003
1.006
1.007
1.008
1.009
1.008
1.008
1.007
1.006
1.004
1.002
0.999
0.996
0.996
0.992
0.983
* IA : Inception Angle
162
Appendix B
Fault
resistance
()
100
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.970
0.976
0.979
0.981
0.983
0.985
0.987
0.990
0.991
0.993
0.995
0.997
0.999
1.001
1.002
1.005
1.006
1.008
1.002
1.012
1.014
0.952
0.958
0.963
0.969
0.975
0.980
0.985
0.990
0.995
1.000
1.005
1.009
1.014
1.018
1.022
1.026
1.030
1.034
1.020
1.041
1.044
1.043
1.039
1.034
1.031
1.027
1.022
1.018
1.014
1.010
1.004
1.000
0.994
0.989
0.983
0.979
0.973
0.968
0.962
0.928
0.951
0.946
0.973
0.975
0.980
0.986
0.991
0.996
1.001
1.004
1.005
1.010
1.015
1.018
1.022
1.025
1.027
1.028
1.031
1.032
0.961
0.961
1.039
0.982
0.986
0.988
0.989
0.992
0.994
0.996
0.997
1.001
1.003
1.005
1.007
1.009
1.011
1.014
1.016
1.017
1.019
1.041
1.043
1.025
* IA : Inception Angle
163
Appendix B
Fault
resistance
()
200
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Output
*X
Output
Target Output Output Output
IA =
(KM)
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
4
4.25
4.5
4.75
5
5.25
5.5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.955
0.970
0.977
0.984
0.990
0.997
1.004
1.011
1.018
1.024
1.031
1.038
1.044
1.051
1.058
1.064
1.070
1.077
1.081
1.090
1.096
1.003
1.000
0.998
0.995
0.992
0.989
0.969
0.972
0.975
0.979
0.982
0.986
0.989
0.993
0.994
0.998
1.002
1.006
1.023
1.014
1.018
0.980
0.983
0.984
0.986
0.987
0.988
0.989
0.991
0.995
0.996
0.998
1.000
1.002
1.004
1.006
1.008
1.012
1.014
1.160
1.021
1.024
0.956
0.961
0.969
0.976
0.983
0.990
0.999
1.003
1.010
1.017
1.024
1.031
1.037
1.041
1.048
1.054
1.060
1.066
0.852
0.857
1.083
0.953
0.956
0.959
0.961
0.963
0.965
0.968
0.971
0.974
0.977
0.979
0.982
0.985
0.989
0.991
0.994
0.997
1.000
1.058
1.061
1.010
* IA : Inception Angle
164
Appendix B
B.2 Double Line to Ground (ABG) ANFIS Fault Type
classification Training Results.
Fault
resistance
()
ANFIS ANFIS
ANFIS ANFIS ANFIS
Output Output
*X
Target Output Output Output
IA =
IA =
(KM)
IA = 0 IA = 45 IA = 90
135
180
0.5
2.000
1.973
2.048
2.107
2.031
0.75
2.005
2.011
2.009
2.022
1.995
1.975
2.019
1.994
1.992
1.980
1.25
2.005
2.022
1.982
1.974
2.000
1.5
2.004
1.987
2.005
1.978
2.002
1.75
2.004
2.017
1.967
1.951
2.003
2.031
2.023
1.964
1.950
1.979
2.25
2.003
2.012
1.965
1.951
2.005
2.5
2.000
2.010
2.003
2.010
2.005
2.75
2.000
1.999
1.973
1.964
2.005
2.001
1.989
1.982
1.974
2.006
3.25
2.000
1.992
1.991
1.985
2.005
3.5
1.997
1.998
1.968
1.956
2.005
3.75
1.997
1.992
2.017
2.016
2.004
1.965
1.991
2.032
2.029
2.009
4.25
1.993
1.986
2.047
2.049
2.002
4.5
1.992
2.214
2.000
2.003
2.001
4.75
1.994
2.006
1.993
2.013
1.999
2.001
1.985
2.082
2.088
1.991
5.25
1.992
1.998
1.989
1.999
1.996
5.5
2.009
1.913
2.011
1.978
2.028
* IA : Inception Angle
165
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
10
0.5
2.000
1.955
2.015
1.906
2.012
0.75
2.000
1.996
1.984
1.949
1.973
2.000
2.009
1.981
1.968
1.990
1.25
2.037
2.018
1.983
1.985
2.012
1.5
2.053
1.987
2.016
1.991
2.022
1.75
2.066
2.029
1.996
2.012
2.028
2.015
2.029
2.002
2.021
2.012
2.25
2.082
2.028
2.008
2.028
2.029
2.5
2.084
2.010
1.980
1.998
2.026
2.75
2.081
2.020
2.014
2.032
2.021
2.075
2.016
2.014
2.027
2.014
3.25
2.064
2.009
2.013
2.032
2.006
3.5
2.050
2.025
2.011
2.010
1.999
3.75
2.033
1.997
2.006
1.999
1.989
2.034
1.990
2.002
1.985
2.008
4.25
1.991
1.983
1.997
2.012
1.973
4.5
1.969
1.987
2.007
2.033
1.964
4.75
1.944
2.012
1.996
2.007
1.958
1.994
1.972
1.987
1.954
1.970
5.25
1.893
2.036
1.998
2.006
1.947
5.5
1.975
1.956
2.001
1.984
2.020
* IA : Inception Angle
166
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
30
0.5
2.012
1.980
2.033
1.963
2.023
0.75
1.995
2.000
2.016
1.978
1.961
2.014
2.008
2.010
1.984
1.968
1.25
2.000
2.015
2.004
1.990
1.974
1.5
2.002
2.038
1.968
2.005
1.980
1.75
2.004
2.022
1.996
1.998
1.985
2.036
2.023
1.993
2.002
1.962
2.25
2.007
2.024
1.991
2.004
1.996
2.5
2.009
2.026
2.050
1.993
2.001
2.75
2.010
2.018
1.991
2.008
2.006
2.011
2.015
1.991
2.008
2.010
3.25
2.011
2.008
1.993
2.009
2.015
3.5
2.012
2.022
1.991
2.006
2.019
3.75
2.012
1.993
1.999
2.009
2.023
1.993
1.985
2.003
2.009
2.007
4.25
2.011
1.974
2.007
2.009
2.030
4.5
2.011
1.992
2.029
1.986
2.034
4.75
2.010
2.010
2.005
2.007
2.037
1.959
1.950
2.019
2.007
2.010
5.25
2.007
1.973
2.014
2.003
2.043
5.5
2.000
1.973
1.963
2.002
2.017
* IA : Inception Angle.
167
Appendix B
ANFIS ANFIS
ANFIS ANFIS ANFIS
*X
Output Output
Target Output Output Output
(KM)
IA =
IA =
resistance
IA = 0 IA = 45 IA = 90
135
180
Fault
50
0.5
2.041
2.003
1.973
2.048
2.031
0.75
1.958
1.998
1.975
2.042
2.059
2.025
1.996
1.977
2.038
2.017
1.25
1.981
1.995
1.978
2.033
2.052
1.5
1.990
2.051
1.925
1.959
2.050
1.75
1.998
1.993
1.984
2.023
2.047
1.992
1.993
1.987
2.017
1.990
2.25
2.008
1.994
1.992
2.011
2.042
2.5
2.012
1.998
2.231
2.013
2.039
2.75
2.012
1.995
2.000
1.998
2.037
2.004
1.995
2.005
1.991
2.031
3.25
2.013
1.996
2.011
1.985
2.032
3.5
2.011
1.996
2.000
2.007
2.028
3.75
2.008
1.999
2.024
1.970
2.025
1.995
2.001
2.031
1.964
2.020
4.25
1.998
2.004
2.039
1.957
2.018
4.5
1.992
2.000
2.021
1.980
2.016
4.75
1.984
1.999
1.911
1.973
2.013
1.968
2.010
2.056
1.944
1.988
5.25
1.965
2.085
2.065
2.035
2.007
5.5
1.955
2.082
2.097
2.033
1.977
* IA : Inception Angle
168
Appendix B
ANFIS ANFIS
ANFIS ANFIS ANFIS
*X
Output Output
Target Output Output Output
(KM)
IA =
IA =
resistance
IA = 0 IA = 45 IA = 90
135
180
Fault
100
0.5
1.994
2.063
1.995
2.045
1.904
0.75
1.949
2.050
1.994
2.034
2.081
1.943
2.044
1.995
2.029
1.951
1.25
1.956
2.038
1.995
2.023
2.070
1.5
1.959
2.009
2.060
1.959
2.065
1.75
1.962
2.026
1.996
2.013
2.060
1.997
2.023
1.997
2.008
2.059
2.25
1.969
2.017
1.998
2.003
2.050
2.5
1.972
1.991
2.224
2.017
2.045
2.75
1.975
2.006
1.999
1.995
2.041
1.970
2.001
2.001
1.990
2.037
3.25
1.982
1.996
2.002
1.986
2.032
3.5
1.985
2.039
2.003
2.012
2.028
3.75
1.988
1.988
2.005
1.977
2.024
1.982
1.984
2.007
1.974
2.020
4.25
1.994
1.979
2.009
1.970
2.017
4.5
1.997
2.021
2.008
1.994
2.013
4.75
2.000
1.991
2.096
2.092
2.010
1.958
1.971
2.013
1.964
2.051
5.25
2.006
1.989
2.065
1.960
2.003
5.5
2.083
1.996
2.114
1.941
2.074
* IA : Inception Angle
169
Appendix B
ANFIS ANFIS
ANFIS ANFIS ANFIS
*X
Output Output
Target Output Output Output
(KM)
IA =
IA =
resistance
IA = 0 IA = 45 IA = 90
135
180
Fault
200
0.5
1.980
1.975
1.981
1.954
2.006
0.75
1.926
1.975
1.984
1.964
2.022
2.021
1.974
1.985
1.968
1.975
1.25
1.935
1.974
1.986
1.973
2.025
1.5
1.940
2.012
1.970
1.908
2.027
1.75
1.945
1.975
1.990
1.981
2.029
1.902
1.976
1.991
1.986
2.057
2.25
1.954
1.976
1.993
1.990
2.032
2.5
1.959
1.955
2.026
2.034
2.034
2.75
1.963
1.977
1.996
1.998
2.035
1.942
1.978
1.998
2.002
2.015
3.25
1.972
1.979
2.000
2.006
2.039
3.5
1.976
2.122
2.011
1.988
2.040
3.75
1.980
1.980
2.004
2.014
2.042
1.959
1.981
2.006
2.018
2.021
4.25
1.988
1.983
2.008
2.022
2.045
4.5
1.992
2.128
2.019
2.005
2.047
4.75
1.996
1.889
2.001
1.954
2.049
2.119
1.986
2.013
2.030
2.005
5.25
2.003
2.176
2.043
1.948
2.052
5.5
1.979
1.988
1.979
1.946
2.011
* IA : Inception Angle
170
Appendix B
B.3 Line to Line (AB) ANFIS Fault Type classification
Training Results.
Fault
*X
resistance (KM)
0.5
3.015
3.013
3.000
3.033
2.989
0.75
3.007
3.008
3.004
3.016
2.974
2.996
3.000
3.016
3.001
3.001
1.25
3.007
3.000
3.006
2.992
3.006
1.5
3.004
2.997
3.005
2.984
2.988
1.75
3.002
2.991
3.004
2.978
3.001
3.008
2.996
2.947
2.989
2.988
2.25
3.003
2.993
2.993
3.003
2.997
2.5
3.004
2.993
3.046
3.008
3.020
2.75
2.998
2.995
2.994
2.976
3.000
3.001
2.995
2.991
2.979
2.988
3.25
3.002
2.987
2.988
2.982
3.037
3.5
3.005
2.993
2.988
2.987
3.002
3.75
3.000
3.000
3.000
3.002
3.000
2.998
2.973
2.993
2.998
2.988
4.25
3.000
3.018
3.007
3.004
3.007
4.5
3.000
3.011
3.003
3.010
3.004
4.75
2.992
2.991
2.996
3.017
2.993
2.991
3.009
3.002
3.023
2.999
5.25
3.003
3.010
3.008
3.030
3.019
5.5
3.000
3.000
3.000
2.997
3.000
* IA : Inception Angle
171
Appendix B
Fault
*X
resistance (KM)
10
0.5
3.006
3.005
2.948
2.967
2.984
0.75
3.060
3.007
3.040
2.995
3.007
3.034
2.997
2.998
3.009
2.974
1.25
2.986
2.998
3.004
3.004
3.014
1.5
3.027
2.977
3.007
3.028
2.973
1.75
2.956
2.960
3.036
3.020
3.006
2.999
2.969
3.000
2.990
3.015
2.25
2.958
3.025
3.015
2.993
3.007
2.5
2.986
3.029
2.986
3.014
3.015
2.75
2.974
3.009
3.109
3.004
2.992
2.986
3.003
2.940
3.014
2.990
3.25
2.994
2.987
2.989
2.992
2.996
3.5
3.064
3.038
2.975
2.990
2.954
3.75
3.000
3.000
3.000
2.995
3.000
4
4.25
4.5
4.75
5
5.25
5.5
3
3
3
3
3
3
3
3.019
3.077
2.967
2.975
2.941
2.933
2.999
2.980
3.005
3.018
3.122
2.980
3.031
3.001
2.916
2.929
2.944
3.020
3.063
3.082
3.000
2.994
2.996
2.995
3.000
3.005
2.998
2.998
3.001
3.002
3.005
3.015
3.012
3.008
3.001
* IA : Inception Angle
172
Appendix B
Fault
*X
resistance (KM)
30
0.5
3.000
2.972
3.016
3.077
3.026
0.75
2.952
2.983
2.979
2.975
3.022
3.020
2.994
3.013
2.992
2.969
1.25
3.016
2.992
3.064
3.041
3.021
1.5
3.039
3.001
3.054
2.982
3.019
1.75
3.031
2.989
3.047
2.987
2.990
2.942
3.003
2.978
2.964
3.053
2.25
3.038
3.007
2.974
3.008
3.000
2.5
2.975
3.000
3.028
3.055
2.989
2.75
2.996
2.995
3.017
3.025
2.935
2.992
3.028
2.983
2.910
3.012
3.25
3.036
3.044
2.933
3.043
2.899
3.5
2.907
3.012
2.881
3.052
2.888
3.75
3.001
3.000
3.006
2.992
2.999
3.054
3.001
3.001
3.065
2.975
4.25
3.042
2.996
3.006
3.054
2.977
4.5
2.989
3.029
2.966
2.906
3.013
4.75
3.051
2.993
3.020
3.008
2.974
2.991
2.990
3.040
3.075
3.000
5.25
2.966
2.983
2.947
2.830
3.019
5.5
3.000
3.000
2.997
3.002
3.003
* IA : Inception Angle
173
Appendix B
Fault
*X
resistance (KM)
50
0.5
3.037
3.041
3.025
3.007
2.915
0.75
3.082
3.004
2.992
2.952
2.939
3.014
2.997
3.015
2.988
3.026
1.25
3.029
3.000
3.021
3.017
2.951
1.5
3.019
3.007
3.015
2.991
2.960
1.75
3.005
2.994
3.048
2.988
2.994
2.959
3.002
2.984
2.982
3.036
2.25
2.995
3.007
3.068
2.970
2.990
2.5
3.009
2.995
2.998
3.029
2.962
2.75
2.955
3.005
2.994
2.992
3.017
3.001
2.983
2.984
2.980
2.986
3.25
2.926
3.007
2.997
3.001
3.024
3.5
2.909
3.001
2.998
3.006
3.026
3.75
3.028
3.000
2.931
3.015
2.998
2.943
3.002
2.984
3.019
3.025
4.25
2.897
2.995
2.991
3.023
3.034
4.5
2.993
2.994
2.988
2.970
2.978
4.75
2.927
2.984
2.989
3.003
3.020
2.983
2.986
2.993
3.012
2.998
5.25
3.017
3.021
2.998
2.954
2.980
5.5
3.001
3.001
2.994
3.012
2.975
* IA : Inception Angle
174
Appendix B
Fault
*X
resistance (KM)
100
0.5
2.774
2.963
2.862
3.016
2.938
0.75
2.937
3.111
2.936
3.012
2.967
3.005
3.000
3.087
2.951
3.002
1.25
2.809
3.044
2.921
3.025
2.977
1.5
2.834
2.932
2.922
3.024
2.981
1.75
2.944
3.053
3.061
3.030
2.960
3.037
2.998
2.826
2.873
3.002
2.25
3.003
2.988
2.964
2.921
2.928
2.5
2.989
3.012
2.954
3.040
3.041
2.75
2.965
3.015
3.039
3.018
3.015
2.917
3.082
3.038
3.014
3.023
3.25
2.949
3.040
3.042
3.010
3.032
3.5
2.954
3.038
3.039
3.006
3.039
3.75
2.989
3.001
3.103
3.076
3.021
2.982
2.946
3.034
2.995
3.062
4.25
3.100
2.918
3.045
2.988
3.075
4.5
2.958
2.977
2.992
2.976
3.061
4.75
3.070
3.004
3.110
2.972
3.110
2.972
2.946
3.138
2.961
3.093
5.25
3.045
2.897
2.974
2.941
3.072
5.5
2.995
2.996
2.963
2.972
2.995
* IA : Inception Angle
175
Appendix B
Fault
*X
resistance (KM)
200
0.5
2.973
3.011
3.039
2.979
2.982
0.75
2.867
3.111
2.934
2.982
2.921
3.005
2.999
3.006
2.679
2.973
1.25
2.966
2.908
2.983
2.987
2.954
1.5
2.973
2.765
2.990
2.990
2.964
1.75
2.991
2.875
3.038
2.992
3.002
3.031
3.049
2.826
3.054
2.909
2.25
2.967
2.999
2.933
2.957
2.997
2.5
3.062
3.075
3.161
2.975
3.014
2.75
2.989
3.002
3.029
3.002
2.996
2.982
3.034
3.018
3.005
2.990
3.25
2.983
3.009
3.007
3.007
2.993
3.5
2.981
3.007
2.997
3.009
2.994
3.75
2.809
2.911
2.870
3.097
2.963
2.984
2.996
2.976
3.014
3.003
4.25
2.984
2.998
2.967
3.016
3.006
4.5
2.979
3.063
2.959
3.019
3.010
4.75
2.989
2.990
2.963
3.021
3.015
2.981
2.986
2.964
3.023
3.018
5.25
2.986
2.883
2.957
3.025
3.019
5.5
2.932
3.020
2.914
3.040
3.088
* IA : Inception Angle
176
Appendix B
B.4 Single Line to Ground (AG)
classification Training Results.
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
0.5
4.024
4.001
4.013
3.998
4.008
0.75
4.013
4.000
4.013
3.992
3.983
4.006
3.997
3.997
3.997
4.001
1.25
4.002
3.994
3.996
3.998
4.000
1.5
4.000
3.991
3.996
4.000
4.000
1.75
4.001
3.989
3.998
4.002
4.000
4.002
3.986
4.007
4.003
4.000
2.25
4.012
3.995
4.009
3.997
3.990
2.5
4.006
3.983
4.007
4.005
4.000
2.75
4.009
3.995
4.009
3.998
3.992
3.937
3.993
4.007
3.999
4.005
3.25
4.005
3.991
4.005
3.999
3.994
3.5
4.035
4.015
3.946
3.950
3.994
3.75
4.002
3.988
4.003
4.000
3.995
3.966
3.987
3.972
3.971
4.001
4.25
4.018
3.998
3.984
3.981
4.004
4.5
4.013
3.994
3.995
3.991
4.006
4.75
4.008
3.992
4.004
3.999
4.008
3.996
3.987
4.001
4.002
4.026
5.25
4.002
3.999
4.012
4.011
4.006
5.5
3.999
4.009
4.009
4.013
4.002
* IA : Inception Angle.
177
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
10
0.5
4.017
4.000
3.997
4.005
3.996
0.75
4.000
3.999
3.998
4.008
3.995
4.006
3.999
3.991
4.009
4.000
1.25
4.002
4.000
3.996
4.012
4.003
1.5
4.000
3.998
3.992
4.014
4.001
1.75
4.000
4.000
4.009
4.015
4.010
4.003
3.997
4.001
4.014
4.013
2.25
3.982
3.991
4.002
4.003
4.002
2.5
4.013
3.985
4.002
4.005
4.014
2.75
3.989
4.000
4.005
3.998
4.000
3.993
4.004
4.005
3.996
4.001
3.25
3.997
4.007
4.006
3.993
3.996
3.5
4.002
3.999
3.976
3.987
3.985
3.75
4.004
4.011
4.007
3.988
3.989
4
4.25
4.5
4.75
5
5.25
5.5
4
4
4
4
4
4
4
4.007
3.991
3.988
3.987
4.017
3.991
3.998
4.013
4.010
4.003
4.004
3.974
4.001
4.008
4.007
4.010
4.003
4.007
3.973
4.001
4.004
3.985
3.985
3.990
3.986
4.013
3.994
4.000
3.991
3.994
3.996
3.996
4.023
3.996
3.994
* IA : Inception Angle.
178
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
30
0.5
4.002
4.010
4.012
4.010
4.009
0.75
4.001
3.982
3.984
3.961
3.998
4.002
3.994
4.009
4.008
4.001
1.25
4.002
4.003
4.006
4.008
4.000
1.5
4.003
4.009
4.004
4.008
4.001
1.75
4.005
4.013
4.004
4.010
4.002
4.005
4.004
4.006
4.012
4.004
2.25
3.994
4.004
3.996
3.988
4.001
2.5
4.006
4.003
4.022
4.017
4.010
2.75
3.994
4.003
4.000
3.997
4.002
4.003
4.008
4.001
4.011
4.010
3.25
3.995
4.003
4.003
4.005
4.003
3.5
3.996
3.996
3.994
3.994
3.996
3.75
3.995
4.002
4.006
4.013
4.005
4.001
3.961
3.986
3.995
3.993
4.25
3.994
3.997
3.985
3.995
3.992
4.5
3.993
3.997
3.986
3.996
3.991
4.75
3.992
3.998
3.989
3.998
3.992
4.007
4.019
4.015
4.030
4.008
5.25
3.991
4.002
3.997
4.001
3.997
5.5
3.990
4.004
3.998
4.003
4.002
* IA : Inception Angle.
179
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
50
0.5
3.994
3.975
4.024
3.998
4.005
0.75
3.998
4.004
4.008
3.983
3.990
4.000
3.999
4.004
3.999
4.007
1.25
4.003
3.994
3.994
3.999
4.008
1.5
4.007
3.991
3.986
3.998
4.008
1.75
4.010
3.990
3.984
3.997
4.009
4.013
3.991
3.990
3.996
4.008
2.25
4.000
3.993
4.001
3.994
3.996
2.5
4.016
4.007
4.029
3.996
4.003
2.75
4.001
3.998
3.998
3.997
3.998
4.002
4.000
3.996
3.999
3.998
3.25
4.003
4.002
3.995
4.000
3.999
3.5
4.012
4.002
4.032
3.993
4.005
3.75
4.005
4.006
3.992
4.002
4.001
4.006
4.008
3.990
4.003
4.001
4.25
4.006
3.992
3.987
3.995
4.010
4.5
4.002
3.994
3.987
3.995
4.009
4.75
3.999
4.014
3.993
3.995
4.007
4.009
4.016
3.983
4.006
4.003
5.25
4.010
4.007
4.008
3.993
4.004
5.5
3.994
4.009
4.013
3.990
4.002
* IA : Inception Angle.
180
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
100
0.5
3.996
4.007
4.015
3.996
3.996
0.75
3.999
3.986
4.015
3.997
3.995
4.000
3.986
4.010
4.003
3.999
1.25
4.001
3.986
4.001
4.003
4.000
1.5
4.002
3.986
3.993
4.002
4.001
1.75
4.001
3.989
3.988
4.000
4.002
4.000
4.011
3.989
3.998
4.002
2.25
4.005
4.011
4.007
3.997
3.995
2.5
3.997
4.030
4.019
3.994
4.002
2.75
4.006
4.010
4.004
3.997
3.995
4.006
4.010
4.003
3.996
3.995
3.25
4.006
4.009
4.002
3.996
3.995
3.5
3.995
4.030
4.021
3.997
3.998
3.75
4.007
4.008
3.999
3.996
3.995
4.008
4.008
3.997
3.995
3.995
4.25
4.000
3.995
3.990
4.003
3.999
4.5
4.001
3.993
3.993
4.004
3.998
4.75
4.000
3.994
3.999
4.004
3.998
4.009
4.006
3.992
3.994
3.996
5.25
3.997
3.994
4.007
4.000
3.996
5.5
4.010
3.990
4.002
3.995
3.994
* IA : Inception Angle.
181
Appendix B
ANFIS
ANFIS ANFIS ANFIS
ANFIS
Fault
*X
Output
Target Output Output Output
Output
resistance (KM)
IA =
IA = 0 IA = 45 IA = 90
IA = 180
135
200
0.5
3.992
4.007
4.006
3.998
3.994
0.75
3.998
3.999
4.010
4.003
4.000
4.002
4.000
4.010
4.006
4.000
1.25
4.004
3.999
4.003
4.007
4.004
1.5
4.004
3.996
3.995
4.006
4.003
1.75
4.003
3.996
3.990
4.004
4.002
4.000
3.988
3.990
4.002
3.999
2.25
4.005
3.988
4.006
4.001
4.000
2.5
3.993
4.025
4.015
3.997
3.992
2.75
4.005
3.987
4.004
4.000
4.000
4.005
3.987
4.004
3.999
4.000
3.25
4.005
3.987
4.003
3.999
4.000
3.5
3.994
3.987
4.012
3.999
3.993
3.75
4.005
3.987
4.002
3.998
4.000
4.005
3.986
4.001
3.998
3.999
4.25
4.005
3.986
3.987
3.997
4.003
4.5
4.005
4.002
3.991
4.007
4.004
4.75
4.005
3.986
3.997
4.007
4.004
4.005
3.986
3.998
3.996
3.999
5.25
4.005
4.004
4.000
4.003
3.998
5.5
3.992
3.996
3.990
3.998
3.993
* IA : Inception Angle.
182
:
.
) (Wavelet Transform
.
).(ANFIS
) (ATP/EMTP
MATLAB
183
ANFIS
.
,
, .
.
:
: .
:
.
:
.
:
).(ANFIS
184
.
: .
.
:
.
: .
185
/
) (
-
2014
/
-
. /
-
2014
:
. / ................................... .
" - - "" "
. / ..................................... .
" - - "
. / ...................................... .
" - - "
-
2014