Skip to main content
Top

2021 | Book

Condition Monitoring Algorithms in MATLAB®

insite
SEARCH

About this book

This book offers the first comprehensive and practice-oriented guide to condition monitoring algorithms in MATLAB®. After a concise introduction to vibration theory and signal processing techniques, the attention is moved to the algorithms. Each signal processing algorithm is presented in depth, from the theory to the application, and including extensive explanations on how to use the corresponding toolbox in MATLAB®. In turn, the book introduces various techniques for synthetic signals generation, as well as vibration-based analysis techniques for large data sets. A practical guide on how to directly access data from industrial condition monitoring systems (CMS) using MATLAB® .NET Libraries is also included. Bridging between research and practice, this book offers an extensive guide on condition monitoring algorithms to both scholars and professionals.

“Condition Monitoring Algorithms in MATLAB® is a great resource for anyone in the field of condition monitoring. It is a unique as it presents the theory, and a number of examples in Matlab®, which greatly improve the learning experience. It offers numerous examples of coding styles in Matlab, thus supporting graduate students and professionals writing their own codes."

Dr. Eric Bechhoefer

Founder and CEO of GPMS

Developer of the Foresight MX Health and Usage Monitoring System

Table of Contents

Frontmatter
Chapter 1. Introduction
Abstract
This Chapter illustrates the role of machine vibration measurement. For this purpose, concepts of system response and system frequency response function from practical perspective are given. In this way, this chapter compares vibration-based condition monitoring with modal analysis of the same object. Next, this Chapter describes how a new concept within condition monitoring becomes an industrial standard. From this Chapter, the reader also learns where MATLAB® could be used as a part of a commercial condition monitoring system.
Adam Jablonski
Chapter 2. Principles of Condition Monitoring Systems
Abstract
This Chapter organizes numerous aspects of condition monitoring system, including classification of systems, terminology, configuration, and operation. It explains step-by-step the relation between detection, identification, diagnostics, monitoring, and others. Finally, the Chapter illustrates the how condition monitoring systems are selected and implemented.
Adam Jablonski
Chapter 3. Vibration Components Generated by Rotary Machinery
Abstract
This Chapter describes what signal components are generated by individual machine elements, such as shafts, blades, gearboxes, and rolling-element bearings (REBs). The Chapter also illustrates the concept of characteristics “orders” in parallel to characteristic frequencies. Finally, it gives practical information concerning various harmonic components.
Adam Jablonski
Chapter 4. Signal Processing Algorithms
Abstract
This chapter is self-contained repository on signal processing algorithms, which are used in condition monitoring of rotary machinery. All codes are written in MATLAB®. Each technique is accompanied by description, exemplary figure, full code, and comments. Codes cover windowing, compensation, zero-padding, filtering, resampling, integration, envelope analysis, Time-Synchronous Averaging (TSA), and phase marker processing.
Adam Jablonski
Chapter 5. Vibration-Based Condition Assessment Methods
Abstract
This Chapter turns previously introduced signal processing techniques into particular machine diagnostic methods. It proposed general division into scalar-based diagnostic analysis and figure-based diagnostic analysis. The Chapter gives many practical recipes concerning calcualtion of both types of assessment methods. Scalar indicators covered in this Chapter range from basic statistical indicators to conceptual advanced markers. Presented figures range from basic time or frequency-domain plots to recent, advanced 3D maps.
Adam Jablonski
Chapter 6. Synthetic Signals Generation Methods
Abstract
This Chapter starts the subject of synthetic data generation. It covers generation of sinusoidal components typical for shafts, blades, and gearboxes, and decaying pulses characteristic for rolling-element bearings (REBs). Each component undergoes multiple monotonic and non-monotonic modulations. Additionally, the Chapter shows how to model jitter characteristic for REBs working under significant load.
Adam Jablonski
Chapter 7. Simulating Operational Signals
Abstract
This Chapter continues the subject of synthetic data generation. It illustrates how to use signal components described in the previous Chapter and turn them into a single synthetic vibration time waveform. The model presented in this Chapter uses separate synthetic speed profile data, selected aspects of frequency response function, and synthetic structural noise. As a result, in this model, each pulse response of modeled REB (for each ball passing under the sensor) is slightly different.
Adam Jablonski
Chapter 8. Simulating Long-Term Machine Fault Development
Abstract
This Chapter uses previously generated signal to build an array of signals, which model slow machine fault development. The fault is defined by a Failure Development Function (FDF). This Chapter shows how to generate data corresponding to eight different modes (i.e. clearly separated technical condition state) of a rotary machinery. Introduced faults include shaft imbalance, gearbox degradation, and rolling-element bearing failure.
Adam Jablonski
Chapter 9. Analysis of Long-Term Fault Development
Abstract
This Chapter shows how long-term raw vibration data could be analyzed. For this purpose, synthetic data generated in previous Chapter is used. Analysis methods include statistical techniques, narrowband spectral analysis, spectral comparison, and 3-dimensional data visualization. Each analysis is used to detect shaft imbalance, gearbox degradation, and REB inner race failure.
Adam Jablonski
Chapter 10. Connecting MATLAB® to CMS
Abstract
This Chapter describes different possible architecture scenarios, which connect MATLAB®, local PC computer, and commercial condition monitoring system (CMS) into one platform. It illustrates how MATLAB® data could flow in and out in connection with a commercial system. For this purpose, this Chapter covers data import and data export subjects in MATLAB®. Finally, this Chapter introduces the concept of data direct access, described next.
Adam Jablonski
Chapter 11. Development of Interface for Direct Data Access (DDA)
Abstract
This Chapter describes the concept of Direct Data Access (DDA), which enables direct reading of data recorded by a professional, commercial condition monitoring system in MATLAB®. The procedure uses .NET libraries, and it is described step-by-step. A special attention is given to internal MATLAB® wrapping of such functions, which significantly simplify codes. Finally, this Chapter illustrates selected possible scenarios of DDA implementation.
Adam Jablonski
Chapter 12. Prototype Tools
Abstract
This Chapter illustrates development of two exemplary prototype tools in MATLAB®, the goal of which is to test new, conceptual signal processing algorithms on real data recorded by a commercial condition monitoring system (CMS). The first tool is developed as a script, which tests some machine health indicators from latest MATLAB® Predictive Maintenance Toolbox™ within imbalance detection. The second tool uses MATLAB® GUI application for generation of user-averaged 3-dimensional spectral analysis. This technique is used to detect shaft imbalance and cavitation from industrial data.
Adam Jablonski
Backmatter
Metadata
Title
Condition Monitoring Algorithms in MATLAB®
Author
Adam Jablonski
Copyright Year
2021
Electronic ISBN
978-3-030-62749-2
Print ISBN
978-3-030-62748-5
DOI
https://doi.org/10.1007/978-3-030-62749-2

Premium Partners