Download Gas Turbine Diagnostics Signal Processing and Fault Isolation Edit By RANJAN GANGULI
Contents Mechanical Engineering:
1. Introduction
2. Idempotent Median Filters
3. Median-Rational Hybrid Filters
4. FIR-Median Hybrid Filters
5. Transient Data and the Myriad Filter
6. Trend Shift Detection
7. Optimally Weighted Recursive Median Filters
8. Kalman Filter
9. Neural Network Architecture
10. Fuzzy Logic System
11. Soft Computing Approach
12. Vibration-Based Diagnostics
Preface Gas Turbine Diagnostics Signal Processing and Fault Isolation:
Gas turbines are very important components of modern infrastructure and are widely used in power generation. In particular, gas turbines are used for propulsion in jet engines that power most commercial and military aircraft. Faults in gas turbine engines can result in major problems, such as delays and cancellations of flights. Engine in-flight shutdowns (IFSDs) are particularly problematic and can have an impact on flight safety. Unscheduled engine removals add to the cost of air transport.
A systematic analysis of engine data has shown that most engine malfunction is preceded by a so-called single fault, which is a fault in one engine module or component. These single faults occur as sharp changes in measurement deviations in the jet engine, when compared to a baseline good engine. In this book, we present and illustrate a number of algorithms for fault diagnosis in gas turbine engines. These methods focus on the aspects of filtering or cleaning the measurement data and on fault isolation algorithms that use simple engine models for finding the type of fault in the engine.
Novel methods for detecting the damage by finding the time location of a sudden change in the signal are also given. These methods include those based on Kalman filters, neural networks, and fuzzy logic and a hybrid soft computing approach.
The book provides a discussion of the different methods in data filtering, trend shift detection, and fault isolation developed over the past decade. Each method is demonstrated through numerical simulations that can be easily done by the reader using worksheets such as MS Excel or through MATLAB®.
The book provides a variety of new research tools for use in the condition monitoring of jet engines. Though the measurements and models are specific to a turbofan engine, the algorithms given in this book will be useful to all engineers and scientists working on fault diagnosis of gas turbine engines.
The data cleaning algorithms based on nonlinear signal processing shown in this book are also applicable to condition and health monitoring problems in general, and as in all such problems, sharp changes in measurement data herald the onset of a fault.
This book will be useful for engineers and scientists interested in gas turbine diagnostics. It will also be of interest to researchers in signal processing and those working on the fault isolation of systems. The algorithms presented in this book have broad appeal and can be used for condition and health monitoring of a variety of systems.
Post a Comment