...

International Electrical Engineering Journal (IEEJ) Vol. 6 (2015) No.4, pp. 1850-1855

by user

on
3

views

Report

Comments

Transcript

International Electrical Engineering Journal (IEEJ) Vol. 6 (2015) No.4, pp. 1850-1855
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
Evaluation of Fault Diagnostic System
for a Multistring Multilevel Inverter
Using an Artificial Neural Network
C.Dinakaran, Assistant Professor, Dept. of EEE, Sri Venkateswara College of Engg. & Tech., Chittoor
[email protected]
Abstract — In recent years, the demand for higher power ratings
for industries has increased and multilevel inverter system has
become a solution for high power applications. A multi level
inverter system not only achieves high power ratings but also
implement the use of renewable energy sources. As multilevel
inverter systems are utilized in high power applications and
reliability of the power electronics equipment. For example,
industrial applications such as manufacturing industries are
dependent upon induction motors and their inverter systems for
manufacturing process control. In general, the conventional
protection systems are passive devices such as fuses, overload
relays and circuit breakers to protect the inverter system.
Whenever a fault occurs, the protective devices will disconnect
power sources from multilevel inverter system. Downtime of
manufacturing equipment results in huge loss. Many engineers
and researchers have focused on incipient fault detection and
preventive maintenance to avoid inverter and motor faults. The
most important factor is to operate the system continuously
during abnormal conditions. In order to maintain continuous
operation for a multilevel inverter system, knowledge of fault
behavior, fault prediction and fault diagnosis are necessary.
Faults should be detected as soon as possible after they occur,
because if a motor drive runs continuously under abnormal
conditions, the drive or motor may quickly fail. In this paper a
fault diagnostic system in a multilevel- inverter using neural
network is developed. The performance of proposed system is
verified by MATLAB/SIMULINK, PROTEUS software and
Hardware setup.
Key Words — Diagnostic system, Fault Diagnosis, Multi
string multilevel inverter, Artificial Neural Network.
I. INTRODUCTION
Now a day’s day to day, the demand for higher
power ratings in industries. Medium voltage level in drives
and utility applications require medium voltage and megawatt
power level. High rating level is normally interfaced with a
medium voltage network. For inverter drive a medium
voltage grid, it is troublesome to connect only one power
semiconductor switch directly. As a result, a multilevel power
converter structure has been introduced as an alternative in
high power and medium voltage situations and also multilevel
inverter drive (MLID) systems have become a solution for
high power drive applications [1] – [2].
A cascaded Multi Level Induction Drive is a general fit
for large automotive all-electric drives because of the high
VA rating possible and because it uses several DC voltage
sources which would be available from batteries or fuel cells.
A multilevel inverter not only achieves high power ratings,
but also enables the use of renewable energy sources such as
photovoltaic, wind and fuel cells that can be easily interfaced
to a multilevel converter system for a high power application
[3]. Multilevel converter topologies have been developed
during the last two decades, the elementary concept of a
multilevel converter to achieve higher power is to use a series
of power semiconductor switches with several low voltage
DC sources to perform the power conversion by synthesizing
a staircase voltage waveform [4].
Capacitors, batteries and renewable energy voltage
sources can be used as the multiple DC voltage sources. The
commutations of the power switches aggregate these multiple
DC sources in order to achieve high voltage at the output [5].
However, the rated voltage of the power semiconductor
switches depends only upon the rating of the DC voltage
sources to which they are connected [6].
The main disadvantage of multilevel inverters is that
they use a high number of power semiconductors, switches for
this reason, multilevel inverters may be considered less
reliable. Also multilevel inverter systems are utilized in high
power applications. Power electronics equipment is more
reliability [7]. Generally, the conventional protection systems
are passive devices such as fuses, overload relays and circuit
breakers to protect the inverter systems and the induction
motors [8]-[9]. The protection devices will disconnect the
power sources from the multilevel inverter system whenever a
fault occurs, stopping the operated process. Downtime of
manufacturing equipment can add up to be thousands or
hundreds of thousands of dollars per hour. Therefore, fault
detection and diagnosis is vital to a company’s bottom line
[10]-[11].
Multilevel inverters provide more possibilities in the
power circuit to operate under faulty conditions. However,
faults should be detected as soon as possible after they occur,
because if a motor drive runs continuously under abnormal
conditions, the drive or motor may quickly fail. Thus
1850
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
knowledge of fault behaviors, fault prediction and fault
diagnosis are necessary [12]-[13].
the vehicle was connected to an AC supply as shown in
Fig.2.
II. CASCADED H – BRIDGE MULTILEVEL
INVERTER DRIVES
A single-phase structure of an m-level cascaded
inverter is shown in Fig.1. Each separate DC source (SDCS)
is connected to a single-phase full-bridge or H-bridge
inverter. Each inverter level can generate three different
voltage outputs +Vdc, 0 and –Vdc by connecting the DC
source to the AC output by different combinations of the four
switches S1, S2, S3 and S4. To obtain +Vdc switches S1 and S4
are turned ON, whereas –Vdc can be obtained by turning ON
switches S2 and S3. By turning on S1 and S2 or S3 and S4 the
output voltage is 0.
The AC output of each of the different full-bridge
inverter levels are connected in series such that the
synthesized voltage waveform is the sum of the inverter
outputs. The number of output phase voltage levels ‘m’ in a
cascade inverter is defined by m = 2n+1, where ‘n’ is the
number of separate DC sources.
.
Fig. 2 Three Phase Star connection structure for electric vehicle motor drive
and battery charging
III. GENERAL PROTECTION USED IN A
CONVENTIONAL INVERTER DRIVE SYSTEM
A conventional inverter drive (CID) has become one
of the major applications in industry. Since a CID is used in
various industrial applications, the reliability of the power
electronic system of a CID is of outstanding importance. An
example of an open loop volts/Hz speed control of a CID is
shown in Fig.3 as can be seen a CID usually consists of six
diodes on the input side, dc link voltage and six
semiconductor power switches on the output side as shown
in Fig. 3 Shows that the faults occur in the motor, rectifier or
inverter. The conventional protection system used in a CID
is mostly passive devices such as fuses, circuit breakers
(CBs) and overload relays as shown in Fig.4.
Fig. 1 Multilevel Inverter
Cascaded inverters have also been proposed for use as
the main traction drive in electric vehicles, where several
batteries or ultra capacitors are well suited to serve as
SDCSs. The cascaded inverter could also serve as a
rectifier/charger for the batteries of an electric vehicle while
Fig. 3 Conventional Voltage Fed PWM Inverter Drive.
1851
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
Fig. 4 A Typical Protection of a conventional voltage fed inverter drive
The modern power semiconductor switches for a CID is
mostly designed as modular package known as intelligent
power module (IPM). An IPM usually combines a single
phase or three phase rectifier with three phase inverter, gate
drive circuit and protection circuit as one package as
depicted in Fig.5
Fig. 6 Structure of Fault Diagnostic System
V. NEURAL NETWORK ARCHITECTURE DESIGN
Fig. 5 A Typical Protection System of a Conventional Voltage Fed Inverter
Drive using Intelligent Power Module
IV. FAULT DIAGNOSTIC SYSTEM
The structure for a fault diagnostic system is illustrated
in Fig.6. The system is composed of four major states:
 Feature Extraction
 Neural Network classification
 Fault Diagnosis
 Switching pattern calculation with Gate signal
output.
Focus of this paper is feature extraction, neural
classification and Fault Diagnosis. The feature extraction
performs the voltage input signal transformation with rated
signal values as important features and the output of the
transformed signal is transferred to the Neural Network
classification. The networks are trained with both normal
and abnormal data for the MLID. Thus, the output of this
network is nearly 0 and 1 as binary code. The binary code is
sent to the fault diagnosis to decode the fault type and its
location. Then, the switching pattern is calculated.
The architecture of the proposed fault diagnostic neural
network is shown in Fig.7. The five multilayer feed forward
networks or multilayer perceptron (MLP) are used in this
research because the input data contain continuous features. A
network has one hidden layer with 40 input nodes
corresponding to harmonic order and magnitude, 2 hidden
nodes, and 1 output node. The sigmoid activation function is
used LOGSIG for hidden nodes and PURELM for an output
node.
The implementation of the proposed neural network
system consisting of five Neural Networks as shown in Fig. 7.
The number of nodes for the input and output layers depends
on the specific application. The selection of number and
dimension in the hidden layer is based on neural network
accuracy in preliminary tests. Indeed, optimization of the
network architecture is a significant topic in a study of
artificial intelligence aspects.
Fig. 7 Feed Forward Network
1852
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
VI. RESULTS
A. Simulation Result
The proposed method has been tested and simulation
results are shown in Fig. 8. This model has been implemented
using MATLAB/SIMULINK environment with SIMPOWER
system toolbox.
Fig. 9 PWM Signal for Proposed Five Level Inverter
Fig. 10 Output Voltage of Five Level Inverter
Fig. 8 Simulation Circuit for Five Level Inverter using Artificial Neural
Network
Fig. 11 Sinusoidal and Triangular Waveforms
Fig. 12 Multilevel Sinusoidal Pulse Width Modulation
1853
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
R8
100
U9(VC)
Q12
2N3859A
D6
Q8
R7
2N3859A
A
B
10
C
Q13
2N3859A
11
D
12
Q9
VC
HIN
SD
VB
HO
VS
LIN
COM
2N3859A
100
U9
DIODE
3
LO
2
6
7
5
1
IR2112
V1
12v
R6(1)
R6(2)
R6
5k
Q14
2N3859A
Q10
2N3859A
Q11
Q15
2N3859A
2N3859A
A
B
A
C
B
D
C
D
U6(MCLR/Vpp/THV)
D19(A)
R12
D19
10
11
U6
13
14
2
3
4
5
6
7
8
9
10
1
OSC1/CLKIN
RB0/INT
OSC2/CLKOUT
RB1
RB2
RA0/AN0
RB3/PGM
RA1/AN1
RB4
RA2/AN2/VREF-/CVREF
RA3/AN3/VREF+
RB5
RB6/PGC
RA4/T0CKI/C1OUT
RB7/PGD
RA5/AN4/SS/C2OUT
RC0/T1OSO/T1CKI
RE0/AN5/RD
RC1/T1OSI/CCP2
RE1/AN6/WR
RC2/CCP1
RE2/AN7/CS
RC3/SCK/SCL
RC4/SDI/SDA
MCLR/Vpp/THV
RC5/SDO
RC6/TX/CK
RC7/RX/DT
RD0/PSP0
RD1/PSP1
Fig. 13 FFT Analysis from Power Graphical User Interface
RD2/PSP2
RD3/PSP3
RD4/PSP4
RD5/PSP5
RD6/PSP6
RD7/PSP7
33
12
34
D1(A)
36
37
D18
R1
D1
38
40
3
10
10
17
11
23
25
12
HIN
VC
SD
VB
HO
VS
LIN
COM
LO
6
11
7
5
1
2
19
12
HIN
VC
SD
29
7
5
1
IR2112
VS
LIN
COM
LO
R2
6
7
10
5
1
11
12
IR2112
U2
DIODE
3
HIN
VC
SD
VB
HO
VS
LIN
COM
LO
100
U5(VC)
6
D5
7
D3
U3(VC)
5
IR2112
U3
DIODE
3
2
U4(VC)
R5
R3
1
D4
100
10
11
12
HIN
VC
SD
VB
HO
VS
LIN
COM
LO
6
7
10
5
1
11
12
2
U4DIODE
3
100
10
11
30
IR2112
U5
DIODE
3
100
R4
IR2112
27
28
6
U2(VC)
100
21
22
LO
2
U7
DIODE
VB
HO
2
26
20
COM
D2
3
U1
DIODE
15
24
VB
HO
VS
LIN
R11
100
39
18
VC
SD
D18(A)
35
16
HIN
100
U8
DIODE
3
HIN
VC
SD
VB
HO
VS
LIN
COM
LO
6
12
7
HIN
VC
SD
6
VB
7
HO
5
VS
LIN
COM
1
LO
5
1
2
IR2112
PIC16F877A
2
IR2112
Fig. 15 Development of Five Level Inverter in PROTEUS Software
Fig. 16 Output Pulse of Five Level Inverter
Fig. 14 Neural Network Training
B. PROTEUS Model
Proteus is the more familiar simulators. It can be used
to simulate almost every circuit on electric fields. It is easy to
use because of the Graphical User Interface (GUI) that is very
similar to the real prototype system. Moreover, it can be used
to design printed circuit board (PCB).
Fig. 17 Output Voltage for Five Level Inverter
1854
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
International Electrical Engineering Journal (IEEJ)
Vol. 6 (2015) No.4, pp. 1850-1855
ISSN 2078-2365
http://www.ieejournal.com/
C. Hardware Setup
The hardware is implemented for five level inverter
IRF 640 IGBT Switches, IN4007 Diode are used to design for
high voltage, high current carrying capability. Gate driver IC
IR2112 Circuits are used for boosting the pulses which will
get from a microcontroller (PIC 16F877A) is used for
generating required pulses.
application. In most cases a reduction of the rated power is
more preferable than a complete shutdown.
It is possible that Artificial Intelligence based technique
can be applied in condition monitoring and diagnosis.
Moreover, the reliability of the system can also be improved
by using diagnosis. Therefore, the fault diagnostic system for
cascaded H-bridge multilevel inverter based on artificial
intelligent technique is implemented in the paper.
REFERENCES
Fig. 18 Experimental Setup for Cascaded H – Bridge Five Level Inverter
Fig. 19 Output Waveform for Cascaded H - Bridge Five Level Inverter
VII. CONCLUSION
The cascaded H-bridge multilevel inverter is one of
optimistic solutions for high power drives application. If a
fault (open or short circuit) occurs in a semiconductor power
switch in a cell, it will causes an unbalanced output voltage
and current, while the traction motor is operating. The
unbalanced voltage and current may result in vital damage to
the traction motor if the traction motor is run in this state for a
long time.
Although a cascaded Multilevel Induction Drive
(MLID) has the ability to tolerate a fault for some cycles, it
would be better if one can detect the fault location. Then,
switching patterns and the modulation index of other active
cells of the MLID can be adjusted to maintain the operation
under balanced load condition, the MLID cannot be operated
at full rated power. The amount of reduction of the rated
power that can be tolerated depends upon the MLID
[1] S.Sunisith, K.S.Mann, Janardhan Rao, “Effective Algorithm for
Reducing DC Link Neutral Point Voltage and Total Harmonic
Distortion for Five Level Inverter”, International Journal of
Engineering Journal (IEEJ), Vol. 5, (2014), No. 11, PP. 1613-1618.
[2] S. kumfoi, Leon M. Tolbrt, ‘‘Fault Diagnostic System for a Multilevel
Inverter using a Neural Network’’, IEEE Trans. Power Electron., vol.
22, no. 3, pp. 1062-1069, May 2007.
[3] D. Kastha and B. K. Bose, “Investigation of fault modes of voltage-fed
inverter system for induction motor drive”, IEEE Trans. Ind. Appl.,
vol.30, no. 4, pp. 1028–1038, Jul. 1994.
[4] L. M. Tolbert, F. Z. Peng and T. G. Habetler, ‘‘Multilevel converters
for large electric drives”, IEEE Trans. Ind. Appl., vol.35, no. 1, pp.
36-44, Jan./Feb. 1999.
[5] Chen, L. Hu, L. Chen, Y. Deng and X. He, ‘‘A multilevel converter
topology with fault-tolerant ability”, IEEE Trans. Power Electron., vol.
20, no. 2, pp. 405-415, Mar. 2005.
[6] S. Zhang, T. Asakura, X. Xu and B. Xu, “Fault diagnosis system for
rotary machines based on fuzzy neural networks”, in Proc.
IEEE/ASME Adv. Intell. Mechatron. (AIM), 2003, pp. 199–204.
[7] H. Demuth and M. Beale, “Neural Network Toolbox User’s Guide”,
Natick, MA: The MathWorks, Inc., 1998.
[8] A. Chen, L. Hu, L. Chen, Y. Deng and X. He, “A multilevel converter
topology with fault-tolerant ability”, IEEE Trans. Power Electron., vol.
20, no. 2, pp. 405–415, Mar. 2005.
[9] D. Diallo, M. H. Benbouzid, D. Hamad and X. Pierre, “Fault detection
and diagnosis in an induction machine drive: A pattern recognition
approach based on concordia stator mean current vector”, IEEE
Trans.Energy Conv., vol. 20, no. 3, pp. 512–519, Sep. 2005.
[10] F. Blaabjerg and J. K. Pedersen, “A new low-cost, fully fault-protected
PWM-VSI inverter with true phase-current information”, IEEE
Trans.Power Electron., vol. 12, no. 1, pp. 187–197, Jan. 1997.
[11] J. Rodriguez, P. W. Hammond, J. Pontt, R. Musalem, P. Lezana and
M. J. Escobar, “Operation of a medium-voltage drive under faulty
conditions”, IEEE Trans. Ind. Electron., vol. 52, no. 4, pp. 1080–1085,
Aug. 2005.
[12] S.K.Meeravali, K.Chandrasekhar, “Power Quality Enhancement using
Multilevel Cascaded H – Bridge based D-STATCOM with IRP
Theory”, International Journal of Engineering Journal (IEEJ), Vol. 6
(2015), No. 2, PP. 1756-1764.
[13] S.Sunisith, M.Saritha, D.Krishna Chaitanya, “Grid Connected Solar
System with PWM Operated Thirteen level Inverter using Digital PI
Controller, International Journal of Engineering Journal (IEEJ), Vol. 5
(2014), No. 12, PP. 1688-1695.
1855
C.Dinakaran
Evaluation of Fault Diagnostic System for a Multistring Multilevel Inverter using an Artificial Neural Network
Fly UP