A Novel Fault Diagnosis System On Polymer Insulation of Power Transformers Based On 3-Stage GA-SA-SVM OFC Selection and ABC-SVM Classifier
A Novel Fault Diagnosis System On Polymer Insulation of Power Transformers Based On 3-Stage GA-SA-SVM OFC Selection and ABC-SVM Classifier
A Novel Fault Diagnosis System On Polymer Insulation of Power Transformers Based On 3-Stage GA-SA-SVM OFC Selection and ABC-SVM Classifier
wangke1@epri.sgcc.com.cn
* Correspondence: liujiefeng9999@163.com (J.L.); hanbozheng@163.com (H.Z.);
Tel.: +86-199-6812-0257 (J.L.); +86-199-6801-1211 (H.Z.)
† These authors contributed equally to this work.
Abstract: Dissolved gas analysis (DGA) has been widely used in various scenarios of power
transformers’ online monitoring and diagnoses. However, the diagnostic accuracy of traditional
DGA methods still leaves much room for improvement. In this context, numerous new DGA
diagnostic models that combine artificial intelligence with traditional methods have emerged. In
this paper, a new DGA artificial intelligent diagnostic system is proposed. There are two modules
that make up the diagnosis system. The two modules are the optimal feature combination (OFC)
selection module based on 3-stage GA–SA–SVM and the ABC–SVM fault diagnosis module. The
diagnosis system has been completely realized and embodied in its outstanding performances in
diagnostic accuracy, reliability, and efficiency. Comparing the result with other artificial intelligence
diagnostic methods, the new diagnostic system proposed in this paper performed superiorly.
Keywords: artificial bee colony (ABC); dissolved gas analysis (DGA); fault diagnosis; genetic
algorithm (GA); power transformers; simulated annealing (SA) algorithm; support vector machine
(SVM)
1. Introduction
1.1. Motivation
Transformers are distributed in almost all domains of the entire electrical network, changing the
values of AC voltage (current) at given points to another or several values without altering the
frequency. They not only guarantee the normal operation of the power grid, but also affect people’s
living environment [1]. However, the operating conditions of the transformer (including temperature
and electromagnetic conditions) are harsh and not conducive to its long-term health [2–4]. In the
meantime, the failure of power transformers is often attended by disastrous consequences, which
include equipment burning and large-scale blackouts. Undoubtedly, the operational safety of power
transformers deserves serious concern.
Fault diagnosis is regarded as one of the most important considerations in maintaining the safe
operation of the power transformer. Diagnostic and fault prognosis techniques have been widely and
successfully applied in numerous engineering dynamic systems [5], and are also of extreme
importance to researchers of electrical energy systems [6]. At present, fault diagnosis systems play a
critical role in maintaining the operational safety of power transformers, and their principles and
designs are constantly updated and strengthened [7,8].
We argue that a sound transformer diagnostic method should be strengthened in the following
aspects: (1) Economic efficiency and (2) solving the allowable-time problem. Economic efficiency is
related to diagnostic costs. The average annual failure rate of transformers is not very high; it usually
does not exceed 5% [9]. In other words, the expense of the transformer fault diagnosis is of minor
significance in most cases. When the transformer is operating properly, the diagnosis provides less
valuable guidance to maintenance staff. However, the traditional diagnosis costs for the transformer
are relatively prominent, due to the lack of online diagnostic methods. Transformers need to be shut
down periodically for maintenance and, during such shutdowns, the outage cost of the transformer
is huge. Therefore, it is necessary to control the diagnostic costs and improve the economic efficiency
of the transformer’s diagnostic method. On the other hand, the allowable-time problem is a noticeable
challenge to traditional transformer diagnosis. The fault of the transformer has no obvious
abnormality at the beginning, making the allowable-time for maintenance actions relatively short.
Under these circumstances, this paper aims to propose an online-diagnosis method that is economical
and capable of overcoming the allowable-time problem.
DGA Samples
OFC Binary
Feature Selection
Sequence
System
State of
Transformer
Randomly Initialize
Generation = 1
The Population
Generation=Gener
Decoding
ation+1
YES
Sort The Generation>= NO Inverse Simulated
Chromosome MaxIteration? Annealing Operation
Genetic
Operations
YES
YES
NO
SVM Fitness Evaluation Generation>= 40?
T ( xi ) + b −1, if yi = −1
(1)
T ( xi ) + b +1, if yi = +1
Polymers 2018, 10, 1096 5 of 20
where the φ(xi) is a nonlinear mapping. Both the φ(xi) and ω contain infinite dimensions. They form
the optimal hyperplane together.
When the data are linearly inseparable, a non-negative slack variable ξi is introduced to
transform the SVM into (2):
l
1
min ( , ) = + C i
2
2 i =1
yi ( T xi + b ) 1 − i (2)
s.t.
i 0, i = 1, 2,..., l
where parameter C is the penalty factor. C was determined through optimization, which depends on
the GA and SA algorithms.
Build a Lagrangian function to solve the QP problem of (2):
L( , b, , , ) = (, )
l l
(3)
− i { yi [ T ( xi ) + b] − 1 + i } − ii
i =1 i =1
Among them, αi > 0 and βi > 0 are Lagrange multipliers; then, the original problem has been
transformed into a quadratic programming problem:
1 l l
max ( ) = −
2 i =1
i j yi y j K ( xi , x j ) + i
i =1
(4)
y
i =1
i i = 0, i [0, C ], i = 1,..., l (5)
where K(xi, xj) is called the kernel function that satisfies (6). σ was a given parameter, which was
determined through the GA and SA optimization.
xi − x j
K ( xi , x j ) = exp − (6)
2
Using the One-Against-One (OAO) method to extend the two-class SVM to a multiclass SVM,
the optimization problem translated into (7):
l
1 jk T jk
min ( , ) = ( ) + C i jk
2 i =1
( jk )T ( xi ) + b jk 1 − i jk , yi = j (7)
s.t. ( jk )T ( xi ) + b jk i jk − 1, yi = k
jk
i 0, i − 1, 2,..., l
Therefore, the expression of the decision function is written as (8):
Polymers 2018, 10, 1096 6 of 20
f jk ( x) = sign[( jk )T ( x) + b jb ] (8)
c σ DGA
L1 L2 L3
1 k lTi
f ( L1 , L2 , L3 ) = − i 100%
k i =1 l
(9)
i
where, li is the number of samples in the ith verification set; lT is the correct classified number in the
verification set; and k is the number of cross validation. The concept of K-fold cross-classification will
be illustrated in the following description.
• Genetic operations
The old solution generates new solutions through genetic operations.
Genetic operation refers to the fact that in each generation, individual chromosomes are chosen
according to their selection probability. After that, chromosomes still need to experience crossover
and mutation in order to generate a new population. This process ensures that the new population is
more adaptable to the environment than the previous generation. The selection probability of each
individual is calculated as follows in (10):
fi
Pi = N
(10)
f
i =1
i
The mutation operation (12) refers to randomly selecting a mutation bit j in the mutated
chromosome and setting it as a normalized random number U(ai, bi). ai, bi are the upper and lower
constraints of the corresponding mutation.
U (ai , bi ) if i = j
xj = (12)
xi otherwise
YES NO YES NO
(Ei-Ej)<=0? (Ei-Ej)>=0?
Generation>= Generation>=
NO MaxIteration? MaxIteration?
NO
YES YES
(a) (b)
Figure 5. The flowcharts of the simulated annealing operation and the inverse annealing operation.
(a) The simulated annealing operation; (b) The inverse simulated annealing operation.
E j − Ei
P = exp − (13)
KT
Bees: Possible
Solutions
All of The Possible Solution
Initial Solution
upper constraints of the search space, respectively, and d is a random integer in [1, D]. The initial
position of the honey source i is randomly generated in the search space according to (14).
xid = Ld + rand(0,1) (U d − Ld ) (14)
To start the search, the lead bee searches around the honey source i according to (15) to generate
a new honey source:
vid = xid + ( xid − x jd ) (15)
where j {1, 2,..., NP}, j i , this indicates randomly selecting a honey source that is not equal to i
among the NP honey sources; φ is a random number of [−1, 1], which is uniformly distributed and
determines the magnitude of perturbation.
Then, the follow bee calculates the fitness of the new honey source Vi = [vi1vi 2 ...vid ] according to
(16) and decides whether or not to replace Xi or keep Xi by using the greedy choice method.
1/ (1 + fi ), f 0
fiti = (16)
1 + abs( fi ), otherwise
fi represents the objective function whose functional value is numerically equal to the mean
square error (MSE) of the accuracy of the SVM prediction model.
After that, follow bees use the Roulette Wheel Selection to determine the lead bees they follow.
The probability in the Roulette Wheel Selection was calculated through (17):
fi
Pi = N
(17)
f
i =1
i
During the search process, if a source Xi reaches the limit through trial iterations without finding
a better source, the source Xi will be abandoned. The lead bee turns into the role of reconnaissance
bee and generates a new source of honey in the search space followed randomly (18). ABC algorithm
flowcharts were depicted in Figure 7.
Ld + rand(0,1) (U d − ld ), t limit
X it +1 = t (18)
X i , t limit
Number of sources
n . found in one
generation N=1
n=1
N+1
YES
The Follow Bee Marks The
NO
New Source, Calculating n+1, n>100?
The Quality of Honey
YES
Obtained The Contemporary
Tagged Source
NO YES
Condition Satisfied? Output
Input Data
Set
LpO CV
ABC
Testing Set Training Set
Parameter
Optimization
NO
Evalaute
YES
SVM
Label Quantity
1 23
2 45
3 10
4 14
5 26
The data need to be preprocessed: Normalizing the data follows (19) to eliminate differences
caused by ratio magnitude differences:
xi − xi.min
xi.result = (19)
xi.max − xi.min
where xi.result is the result of normalization, xi is the ratio which needs to be normalized, xi.max and xi.min
are the maximum and the minimum members among entire samples.
In 3-stage GA–SA–SVM optimization, several parameters were specified in Tables 3 and 4. The
maximum iteration number was set at 200. The population scale was determined at 20. The number
of chromosome segments was 3. Both L1 and L2 took 18, which guarantees that the upper bound of
both C and σ is 255 and they can be accurate to 10−4. L3 was 28. In the optimization process, the first
40 generations were fully optimized using the GA-SVM method, 40–180 generations utilized the GA–
SA–SVM algorithm, and the Inverse SA–GA–SVM algorithm was applied in the last 20 generations.
(a)
Polymers 2018, 10, 1096 13 of 20
(b)
(c)
(d)
Figure 9. Results of four optimal feature selection methods: (a) The result of the GA–SVM method;
(b) The result of the GA–SA–SVM method; (c) The result of the 2-stage GA–SA–SVM method; (d) The
result of the 3-stage GA–SA–SVM method.
CV
Method Selected Combinations
Accuracy
H2/CO, H2/CO2, H2/TH, CO/CO2, CO/C2H2, CO2/CH4, CO2/C2H4,
GA–SVM 88.17%
CH4/TH, C2H4/TH, C2H4/C2H6
Polymers 2018, 10, 1096 14 of 20
In Table 5, 3-stage GA–SA–SVM has the highest CV accuracy among all algorithms. In Figure
9a, GA–SVM’s fitness reaches the highest value within 20 generations and takes the shortest time,
which is only about 200 s, to end the iteration. However, due to the fitness curve no longer climbing
after reaching a platform, the GA algorithm is more likely to be trapped in the local optimal solution,
making the result unstable. Also, the accuracy of the GA–SVM is slightly lower than that of other
algorithms. The GA–SA–SVM algorithm made some improvements based on the GA algorithm. In
Figure 9b, the fitness of the GA–SA–SVM changed after arriving at a local optimization platform,
which means that it is easier for the GA–SA–SVM to jump out of the local optimal solution. Therefore,
the result of the GA–SA–SVM looks more stable and accurate. However, the weakness of the GA–
SA–SVM is marked. The GA–SA–SVM’s fitness reaches the platform period at around the 40th
generation and requires a long running time of more than 600 s. The adoption of the 2-stage GA–SA–
SVM has already made some improvements to this problem. In Figure 9c, the 2-stage GA–SA–SVM
was able to jump out of the local optimal solution and merely required 474 s to complete the
optimization, and 20 generations to reach the local optimal platform. The shortcoming that remains
in the 2-stage GA–SA–SVM is that the fitness may jump out of the global optimal solution at the end
of the optimization. This situation is due to the temperature T in the SA algorithm is already very
low within the last ten generations, and the probability of accepting a positive-direction-mutation
(which makes the fitness grow) is quite low. In contrast, the probability of receiving a negative-
direction-mutation is 100%. For example, maximum fitness in the 194th generation of Figure 9c
decreased during accepting a negative-direction-mutation. This generation was very close to the
maximum number of iterations. At this moment, it was a risk that the result may return to the local
optimal solution and never grow again until the end of the iteration. This is a typical defect of the SA
algorithm when the maximum number of iterations is set in the first place. A similar situation also
occurs in the GA–SA–SVM hybrid algorithm: At the 190th generation in Figure 9b, fitness returns to
the local optimal solution until the end of the iteration. The 3-stage GA–SA–SVM was intended to
overcome this shortage. In Figure 9d, the 3-stage GA–SA–SVM retains all of the benefits of the 2-stage
GA–SA–SVM, except that the solution takes a little longer—up to 506 s. After the 180th generation,
the inverse SA algorithm not only eliminated the decrease in fitness that might occur in the SA
algorithm, but also provided two opportunities for the fitness to raise. The 3-stage GA–SA–SVM is
therefore more stable and accurate than the 2-stage GA–SA–SVM method.
Based on the high accuracy and stability of the 3-stage GA–SA–SVM, the selection results of the
3-stage GA–SA–SVM are considered to be the most reasonable OFC. DGA ratio components of the
OFC are set out in Table 6.
Ratios
H2/CO2, H2/C2H2, H2/C2H4, H2/TH, CO2/CH4,
CO2/C2H2, CO2/C2H4, CH4/C2H6, C2H6/TH,
CH4/TH, C2H2/C2H4, C2H2/C2H6
Based on the LpO CV, 118 sets of IEC TC 10 samples were divided into two groups. Among
these, 93 sets were for training and 25 sets for testing. The states arrangement of testing samples is
listed in Table 7.
Label Quantity
1 5
2 10
3 3
4 3
5 4
In the ABC algorithm, we set the scale of the bee colony to 20. The number of honey sources
(solutions) is set to half of the scale, that is, 10. In each generation, the maximum number of extra
honey sources that can be found are 100. That is, if the reconnaissance bees discover more than 100
fresh honey sources and the quality of the honey sources does not increase, reinitialize the honey
sources. This setting is to prevent ABC from being trapped in the local optimal solution. The
maximum number of loops is 10 and the dimension of the vector to be optimized is 2. The parameters
are arranged as shown in Table 8.
Number of Honey Scale of The Bee Max Number of New Sources in Max Number of
Source Colony One Generation Loops
10 20 100 10
(a) (b)
(c)
Figure 11. Testing accuracies for all the points (c, σ) (a) Cross-sections of the points at (100, 90.19); (b)
Cross-sections of the points at (100, 90.72); (c) is a 3D visualization of all the points.
selected by the author after 50 times of repeated experiments, and are the closest to the average
results.
The advantages of the ABC–SVM in terms of accuracy can be clearly seen from the comparison
between Figures 10 and 12. The ABC–SVM is the only one of all algorithms with a precision of over
90%. The diagnostic accuracy of ABC–SVM is obviously improved compared to that of standard
SVM, and it is better than that of BPNN. The accuracy of ABC–SVM is also higher than that of other
wrapper algorithms.
In addition, it can be seen that the ABC algorithm has better convergence characteristics when
compared to other optimization algorithms, which guarantees that ABC–SVM performs better than
other wrapper algorithms. It runs steadily, has fewer iterations, and has a rapid convergence time
and a high termination fitness. Unlike GA or PSO, which require hundreds of generations of
calculations, ABC reached the optimal platform within five iterations. Besides, as seen in Figure 13,
termination fitness of ABC exceeded 95%, which is almost 10% higher than that of the GA and the
PSO. This shows the ABC’s outstanding preferment in convergence.
(a) (b)
(c) (d)
Figure 12. Fault diagnosis results and the spatial distribution of the optimal solution using different
methods (a) PSO–SVM; (b) GA–SVM; (c) SVM; (d) BPNN.
Polymers 2018, 10, 1096 18 of 20
(a)
(b)
(c)
Figure 13. Average fitness and best fitness of SVM based methods (a) ABC method; (b) PSO method;
(c) GA–SVM.
Author Contributions: In this research activity, all the authors were involved in the data collection and
preprocessing phase, model constructing, empirical research, results analysis and discussion, and manuscript
preparation. All authors have approved the submitted manuscript.
Acknowledgments: The authors acknowledge the National Natural Science Foundation of China (Grant No.
51867003), the National Basic Research Program of China (973 Program, 2013CB228205), the National High-tech
R & D Program of China (863 Program, 2015AA050204), the Natural Science Foundation of Guangxi
(2015GXNSFBA139235), the Foundation of Guangxi Science and Technology Department (AE020069), the
Foundation of Guangxi Education Department (T3020097903), and the National Key Research and Development
Program of China (2016YFB0900101) in support of this work.
References
1. Li, J.S.; Zhou, H.W.; Meng, J.; Yang, Q.; Chen, B. Carbon emissions and their drivers for a typical urban
economy from multiple perspectives: A case analysis for Beijing city. Appl. Energy 2018, 226, 1076–1086.
2. Liu, J.; Zheng, H.; Zhang, Y.; Zhou, T.; Zhao, J.; Li, J.; Liu, J.; Li, J. Comparative Investigation on the
Performance of Modified System Poles and Traditional System Poles Obtained from PDC Data for
Diagnosing the Ageing Condition of Transformer Polymer Insulation Materials. Polymers 2018, 10, 191.
3. Zhang, Y.; Liu, J.; Zheng, H.; Wei, H.; Liao, R.; Sciubba, E. Study on Quantitative Correlations between the
Ageing Condition of Transformer Cellulose Insulation and the Large Time Constant Obtained from the
Extended Debye Model. Energies 2017, 10, 1842.
Polymers 2018, 10, 1096 19 of 20
4. Liu, J.; Zheng, H.; Zhang, Y.; Wei, H.; and Liao, R. Grey Relational Analysis for Insulation Condition
Assessment of Power Transformers Based Upon Conventional Dielectric Response Measurement. Energies
2017, 10, 1526.
5. Borutzky, W. Bond Graph Modelling of Engineering Systems. Springer: New York, NY, USA, 2011; pp. 105–
135, 978-1-4419-9367-0.
6. Djeziri, M.A.; Ananou B.; Ouladsine M. Data driven and model based fault prognosis applied to a
mechatronic system. In Proceedings of Fourth International Conference on Power Engineering, Energy and
Electrical Drives. IEEE, Istanbul, Turkey, 13–17 May 2013.
7. Sun, H.-C.; Huang, Y.-C.; Huang, C.-M. A Review of Dissolved Gas Analysis in Power Transformers.
Energy Procedia 2012, 14, 1220–1225.
8. Sica, F.C.; Guimarães, F.G.; Duarte, R.d.O.; Reis, A.J.R. A cognitive system for fault prognosis in power
transformers. Electr. Power Syst. Res. 2015, 127, 109–117.
9. Bengtsson, C. Status and trends in transformer monitoring. IEEE Trans. Power Deliv. 1996, 11, 1379–1384.
10. Yang, M.-T.; Hu, L.-S. Intelligent fault types diagnostic system for dissolved gas analysis of oil-immersed
power transformer. IEEE Trans. Dielectr. Electr. Insul. 2013, 20, 2317–2324.
11. Singh, S.; Bandyopadhyay, M.N. Dissolved gas analysis technique for incipient fault diagnosis in power
transformers: A bibliographic survey. IEEE Electr. Insul. Mag. 2010, 26, 41–46.
12. Duval, M.; Pablo, A.D.; Atanasova-Hoehlein, I.; Grisaru, M. Significance and detection of very low degree
of polymerization of paper in transformers. IEEE Electr. Insul. Mag. 2017, 1, 31–38.
13. Peischl, S.; Walker, J.P.; Ryu, D.; Kerr, Y.H. Analysis of Data Acquisition Time on Soil Moisture Retrieval
from Multiangle L-Band Observations. IEEE Trans. Geosci. Remote 2017, 56, 966–971.
14. Unsworth, J.; Mitchell, F. Degradation of electrical insulating paper monitored with high performance
liquid chromatography. IEEE Trans. Electr. Insul. 1990, 25, 737–746.
15. Verma, H.C.; Baral, A.; Pradhan, A.K.; Chakravorti, S. A method to estimate activation energy of power
transformer insulation using time domain spectroscopy data. IEEE Trans. Dielectr. Electr. Insul. 2017, 24,
3245–3253.
16. Saha, T.K.; Purkait, P.; Muller, F. An attempt to correlate time & frequency domain polarisation
measurements for the insulation diagnosis of power transformer. In Proceedings of Power Engineering
Soc. General Meeting, Denver, CO, USA, 6–10 June, 2004.
17. Bakar, N., A.; Abu-Siada, A.; and Islam, S. A review of dissolved gas analysis measurement and
interpretation techniques. IEEE Electr. Insul. Mag. 2014, 30, 39–49.
18. Gómez, N.A.; Wilhelm, H.M.; Santos, C.C.; Stocco, G.B. Dissolved gas analysis (DGA) of natural ester
insulating fluids with different chemical compositions. IEEE Trans. Dielectr. Electr. Insul. 2014, 21, 1071–
1078.
19. Rogers, R.R. IEEE and IEC Codes to Interpret Incipient Faults in Transformers, Using Gas in Oil Analysis.
IEEE Trans. Electr. Insul. 1978, EI-13, 349–354.
20. Duval, M.; Depabla, A. Interpretation of gas-in-oil analysis using new IEC publication 60599 and IEC TC
10 databases. IEEE Electr. Insul. Mag. 2001, 31–41.
21. Irungu, G.K.; Akumu, A.O. Munda, J.L. A new fault diagnostic technique in oil-filled electrical equipment;
the dual of Duval triangle. IEEE Trans. Dielectr. Electr. Insul. 2016, 23, 3405–3410.
22. Irungu, G.K.; Akumu, A.O.; Munda, J.L. Comparison of IEC 60599 gas ratios and an integrated fuzzy-
evidential reasoning approach in fault identification using dissolved gas analysis. In Proceedings of
International Universities Power Engineering Conference (UPEC), Coimbra, Portugal, 6–9 Sept 2016.
23. Barbosa, T.M.; Ferreira, J.G.; Finocchio, M.A.F.; Endo, W. Development of an Application Based on the
Duval Triangle Method. IEEE Latin Am. Trans. 2017, 15, 1439–1446.
24. Benmahamed, Y.; Teguar, M.; Boubakeur, A. Application of SVM and KNN to Duval Pentagon 1 for
transformer oil diagnosis. IEEE Trans. Dielectr. Electr. Insul. 2017, 24, 3443–3451.
25. Faiz, J.; Soleimani, M.; Dissolved gas analysis evaluation in electric power transformers using conventional
methods a review. IEEE Trans. Dielectr. Electr. Insul. 2017, 24, 1239–1248.
26. Islam, S.M.; Wu, T.; Ledwich, G. A novel fuzzy logic approach to transformer fault diagnosis. IEEE Trans.
Dielectr. Electr. Insul. 2000, 7, 177–186.
27. Miranda, V.; Castro, A.R.G. Improving the IEC table for transformer failure diagnosis with knowledge
extraction from neural networks. IEEE Trans. Power Deliv. 2005, 20, 2509–2516.
Polymers 2018, 10, 1096 20 of 20
28. Zheng, H.; Zhang, Y.; Liu, J.; Wei, H.; Zhao, J.; Liao, R. A novel model based on wavelet LS-SVM integrated
improved PSO algorithm for forecasting of dissolved gas contents in power transformers. Electr. Power Syst.
Res. 2018, 155, 196–205.
29. Zhang, Y.; Wei, H.; Liao, R.; Wang, Y.; Yang, L.; Yan, C. A New Support Vector Machine Model Based on
Improved Imperialist Competitive Algorithm for Fault Diagnosis of Oil-immersed Transformers. J. Electr.
Eng. Technol. 2017, 12, 830–839.
30. Dai, J.; Song, H.; Sheng, G.; Jiang, X. Dissolved gas analysis of insulating oil for power transformer fault
diagnosis with deep belief network. IEEE Trans. Dielectr. Electr. Insul. 2017, 24, 2828–2835.
31. Mirowski, P.; LeCun, Y. Statistical Machine Learning and Dissolved Gas Analysis: A Review. IEEE Trans.
Power Deliv. 2012, 27, 1791–1799.
32. Huang, Y.C. A new data mining approach to dissolved gas analysis of oil-insulated power apparatus. IEEE
Trans. Power Deliv. 2003, 18, 1257–1261.
33. Chen, W.; Pan, C.; Yun, Y.; Liu, Y. Wavelet Networks in Power Transformers Diagnosis Using Dissolved
Gas Analysis. IEEE Trans. Power Deliv. 2008, 24, 187–194.
34. Zhou, J.; Yang, Y.; Ding, S., X.; Zi, Y.; Wei, M. A Fault Detection and Health Monitoring Scheme for Ship
Propulsion Systems Using SVM Technique. IEEE Access 2018, 6, 16207–16215.
35. Zhang, Y.; Zheng, H.; Liu, J.; Zhao, J.; Sun, P. An Anomaly Identification Model for Wind Turbine State
Parameters. J. Clean. Prod. 2018, 195, 1214–1227.
36. Qin, L.; Wang, J.; Li, H.; Sun, Y.; Li, S. An Approach to Improve the Performance of Simulated Annealing
Algorithm Utilizing the Variable Universe Adaptive Fuzzy Logic System. IEEE Access 2017, 5, 18155–18165.
37. Xin, F.; Ni, S.; Li, H.; Zhou, X. General Regression Neural Network and Artificial-Bee-Colony Based General
Regression Neural Network Approaches to the Number of End-of-Life Vehicles in China. IEEE Access 2018,
6, 19278–19286.
38. Tang, W.H.; Goulermas, J.Y.; Wu, Q.H.; Richardson, Z.J.; Fitch, J. A Probabilistic Classifier for Transformer
Dissolved Gas Analysis with a Particle Swarm Optimizer. IEEE Trans. Power Deliv. 2008, 23, 751–759.
39. Kim, S., W.; Kim, S., J.; Seo, H.D.; Jung, J.R.; Yang, H.J.; Duval, M. New methods of DGA diagnosis using
IEC TC 10 and related databases Part 1: Application of gas-ratio combinations. IEEE Trans. Dielectr. Electr.
Insul. 2013, 20, 685–690.
40. Karaboga, D.; Akay, B. A survey: Algorithms simulating bee swarm intelligence. Artif. Intell. Rev. 2009, 31,
61.
© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).