Download Gas Turbine Diagnostics - Signal Processing and Fault Isolation - Ranjan Ganguli (CRC, 2013) PDF

TitleGas Turbine Diagnostics - Signal Processing and Fault Isolation - Ranjan Ganguli (CRC, 2013)
File Size7.6 MB
Total Pages235
Table of Contents
                            Front Cover
Contents
Preface
About the Author
Chapter 1 - Introduction
Chapter 2 - Idempotent Median Filters
Chapter 3 - Median-Rational Hybrid Filters
Chapter 4 - FIR-Median Hybrid Filters
Chapter 5 - Transient Data and the Myriad Filter
Chapter 6 - Trend Shift Detection
Chapter 7 - Optimally Weighted Recursive Median Filters
Chapter 8 - Kalman Filter
Chapter 9 - Neural Network Architecture
Chapter 10 - Fuzzy Logic System
Chapter 11 - Soft Computing Approach
Chapter 12 - Vibration-Based Diagnostics
References
Back Cover
                        
Document Text Contents
Page 1

R a n j a n G a n G u l i

Gas Turbine
Diagnostics
Signal Processing and Fault Isolation

ISBN: 978-1-4665-0272-7

9 781466 502727

9 0 0 0 0

K14444

Signal ProceSSing

“Excellent book on the topic with comprehensive description of the
theory and a simple approach for gas turbine engine performance
diagnostics.”
—Ashwani K. Gupta, University of Maryland, College Park, USA

“... a single reference for numerous techniques of fault analysis and
isolation. The book in its 12 chapters provides an organized way
for fault analysis in gas turbines. Simple algorithms using MATLAB®
are developed based on Kalman filters, neural networks and fuzzy
logic, and a hybrid soft computing approach. The book is useful for
both engineers and scientists interested in gas turbine diagnostics.”
—Dr. Ahmed F. El-Sayed, Zagazig University, Egypt

Widely used for power generation, gas turbine engines are susceptible to faults due to
the harsh working environment. Most engine problems are preceded by a sharp change
in measurement deviations compared to a baseline engine, but the trend data of these
deviations over time are contaminated with noise and non-Gaussian outliers. Gas
Turbine Diagnostics: Signal Processing and Fault Isolation presents signal processing
algorithms to improve fault diagnosis in gas turbine engines, particularly jet engines. The
algorithms focus on removing noise and outliers while keeping the key signal features
that may indicate a fault.

The book brings together recent methods in data filtering, trend shift detection, and fault
isolation, including several novel approaches proposed by the author. Each method is
demonstrated through numerical simulations that can be easily performed by the reader.
Coverage includes hybrid filters, nonlinear myriad filters, innovative nonlinear filters for
data cleaning developed using optimization methods, an edge detector based on gradient
and Laplacian calculations, and more.

Using simple examples, the book describes new research tools to more effectively isolate
faults in gas turbine engines. These algorithms may also be useful for condition and health
monitoring in other systems where sharp changes in measurement data indicate the
onset of a fault.

Gas Turbine
Diagnostics

K14444_Cover_mech.indd 1 11/5/12 3:42 PM

Page 117

103Optimally Weighted Recursive Median Filters

These figures clearly illustrate the capability of the weighted RM filters to
preserve sharp edges or trend shifts in a signal and to remove noise from
stationary regions.

7.4 Test Signal with Outliers

The signal in Figure 7.9 considers the combination signal with added noise
(SNR = 1.5) and outliers. Outliers represent the impulsive noise that may be
present in a signal. Here, the outliers are selected at three different levels. The
first is equal to 4.23°C and is added at k = 10, 80, and 140 and subtracted at
k = 40 and 120. The 8.46°C outlier is added at k = 20, 100, and 190 and sub-
tracted at k = 30 and 170. The last outlier has a value of 12.69°C, and this is
added at k = 110 and 160 and subtracted at k = 60 and 130. Similarly, outliers
are added to the step, ramp, and transient signals. The weights obtained after
putting in the outliers are the same as shown in Table 7.3. This is primarily
because all median architectures are good at removing outliers, and the
weights serve to address the ideal signal characteristics.

The weighted recursive median filter is able to efficiently discard these
outliers while preserving signal features that can be easily observed from
Figures 7.9–7.12. Results in Tables 7.4 and 7.5 show that the simple median

1.0

Noisy
Filtered

0.95

0.85∆
EG

T

0.7

0.8

0.9

0.75

0.65

0 7 15
Epoch (k)

21 27

FIGURE 7.8
Effect of weighted RM filters on noisy realistic signal with SNR = 1.5. (From Uday, P., and
Ganguli, R., Journal of Engineering for Gas Turbines and Power 132(4):2010. With permission.)

Page 118

104 Gas Turbine Diagnostics: Signal Processing and Fault Isolation

noisy
filtered

40

30

20

10

0

-10

-20
0 50 100 150 200

Epoch (k)


EG

T
(C

)

FIGURE 7.9
Effect of weighted RM filters on noisy combination signal with outliers. (From Uday, P., and
Ganguli, R., Journal of Engineering for Gas Turbines and Power 132(4):2010. With permission.)

25

20

15

Noisy
Filtered

10


EG

T
(C

)

5

0

–5
0 50 100

Epoch (k)
150 200

FIGURE 7.10
Effect of weighted RM filters on noisy step signal with outliers. (From Uday, P., and Ganguli, R.,
Journal of Engineering for Gas Turbines and Power 132(4):2010. With permission.)

Page 234

220 References

137. Greaves, R.W., and White, E.R. 1987. An Overview of Airborne Vibration Monitoring
(AVM) Systems. SAE Technical Paper Series 871731.

138. Harker, R.W., and Handelin, G.W. 1990. Enhanced On-Line Machinery Condition
Monitoring through Automated Start-up/Shutdown Vibration Data Acquisition.
American Society of Mechanical Engineers 90-GT-272.

139. Hartranft, J.J. 1995. Description of a Vibration Diagnostic Trending Approach for a
Condition Monitoring System for the LM2500 Gas Turbine. American Society of
Mechanical Engineers 95-GT-373.

140. Oberholster, A.J., and Heyns, P.S. 2004. On-Line Fan Blade Damage Detection
Using Neural Networks. Mechanical Systems and Signal Processing 20(1):78–93.

141. Larsen, G.C., Hansen, A.M., and Kristensen, O.J.D. 2002. Identification of
Damage to Wind Turbine Blades by Modal Parameter Estimation. Report
Riso-R-1334 (EN). Roskilde, Denmark: Riso National Laboratory.

Page 235

R a n j a n G a n G u l i

Gas Turbine
Diagnostics
Signal Processing and Fault Isolation

ISBN: 978-1-4665-0272-7

9 781466 502727

9 0 0 0 0

K14444

Signal ProceSSing

“Excellent book on the topic with comprehensive description of the
theory and a simple approach for gas turbine engine performance
diagnostics.”
—Ashwani K. Gupta, University of Maryland, College Park, USA

“... a single reference for numerous techniques of fault analysis and
isolation. The book in its 12 chapters provides an organized way
for fault analysis in gas turbines. Simple algorithms using MATLAB®
are developed based on Kalman filters, neural networks and fuzzy
logic, and a hybrid soft computing approach. The book is useful for
both engineers and scientists interested in gas turbine diagnostics.”
—Dr. Ahmed F. El-Sayed, Zagazig University, Egypt

Widely used for power generation, gas turbine engines are susceptible to faults due to
the harsh working environment. Most engine problems are preceded by a sharp change
in measurement deviations compared to a baseline engine, but the trend data of these
deviations over time are contaminated with noise and non-Gaussian outliers. Gas
Turbine Diagnostics: Signal Processing and Fault Isolation presents signal processing
algorithms to improve fault diagnosis in gas turbine engines, particularly jet engines. The
algorithms focus on removing noise and outliers while keeping the key signal features
that may indicate a fault.

The book brings together recent methods in data filtering, trend shift detection, and fault
isolation, including several novel approaches proposed by the author. Each method is
demonstrated through numerical simulations that can be easily performed by the reader.
Coverage includes hybrid filters, nonlinear myriad filters, innovative nonlinear filters for
data cleaning developed using optimization methods, an edge detector based on gradient
and Laplacian calculations, and more.

Using simple examples, the book describes new research tools to more effectively isolate
faults in gas turbine engines. These algorithms may also be useful for condition and health
monitoring in other systems where sharp changes in measurement data indicate the
onset of a fault.

Gas Turbine
Diagnostics

K14444_Cover_mech.indd 1 11/5/12 3:42 PM

Similer Documents