MATLAB Predictive Maintenance Toolbox User s Guide

The Mathworks get pdf

Available on textbookfull.com
https://textbookfull.com/product/matlab-predictive-maintenance-
toolbox-user-s-guide-the-mathworks/

★★★★★
4.7 out of 5.0 (93 reviews )

Instant PDF Download

 MATLAB Predictive Maintenance Toolbox User s Guide The
Mathworks

TEXTBOOK

Available Formats

■ PDF eBook Study Guide Ebook

EXCLUSIVE 2025 ACADEMIC EDITION – LIMITED RELEASE

Available Instantly Access Library

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

MATLAB Econometrics Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-econometrics-toolbox-
user-s-guide-the-mathworks/

MATLAB Bioinformatics Toolbox User s Guide The

Mathworks

https://textbookfull.com/product/matlab-bioinformatics-toolbox-
user-s-guide-the-mathworks/

MATLAB Mapping Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-mapping-toolbox-user-s-
guide-the-mathworks/

MATLAB Optimization Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-optimization-toolbox-
user-s-guide-the-mathworks/
MATLAB Trading Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-trading-toolbox-user-s-
guide-the-mathworks/

MATLAB Computer Vision Toolbox User s Guide The

Mathworks

https://textbookfull.com/product/matlab-computer-vision-toolbox-
user-s-guide-the-mathworks/

MATLAB Curve Fitting Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-curve-fitting-toolbox-
user-s-guide-the-mathworks/

MATLAB Fuzzy Logic Toolbox User s Guide The Mathworks

https://textbookfull.com/product/matlab-fuzzy-logic-toolbox-user-
s-guide-the-mathworks/

MATLAB Global Optimization Toolbox User s Guide The

Mathworks

https://textbookfull.com/product/matlab-global-optimization-
toolbox-user-s-guide-the-mathworks/
Predictive Maintenance Toolbox™
User's Guide

R2020a
How to Contact MathWorks

Latest news: www.mathworks.com

Sales and services: www.mathworks.com/sales_and_services

User community: www.mathworks.com/matlabcentral

Technical support: www.mathworks.com/support/contact_us

Phone: 508-647-7000

The MathWorks, Inc.

1 Apple Hill Drive
Natick, MA 01760-2098
Predictive Maintenance Toolbox™ User's Guide
© COPYRIGHT 2018–2020 by The MathWorks, Inc.
The software described in this document is furnished under a license agreement. The software may be used or copied
only under the terms of the license agreement. No part of this manual may be photocopied or reproduced in any form
without prior written consent from The MathWorks, Inc.
FEDERAL ACQUISITION: This provision applies to all acquisitions of the Program and Documentation by, for, or through
the federal government of the United States. By accepting delivery of the Program or Documentation, the government
hereby agrees that this software or documentation qualifies as commercial computer software or commercial computer
software documentation as such terms are used or defined in FAR 12.212, DFARS Part 227.72, and DFARS 252.227-7014.
Accordingly, the terms and conditions of this Agreement and only those rights specified in this Agreement, shall pertain
to and govern the use, modification, reproduction, release, performance, display, and disclosure of the Program and
Documentation by the federal government (or other entity acquiring for or through the federal government) and shall
supersede any conflicting contractual terms or conditions. If this License fails to meet the government's needs or is
inconsistent in any respect with federal procurement law, the government agrees to return the Program and
Documentation, unused, to The MathWorks, Inc.
Trademarks
MATLAB and Simulink are registered trademarks of The MathWorks, Inc. See
www.mathworks.com/trademarks for a list of additional trademarks. Other product or brand names may be
trademarks or registered trademarks of their respective holders.
Patents
MathWorks products are protected by one or more U.S. patents. Please see www.mathworks.com/patents for
more information.
Revision History
March 2018 Online only New for Version 1.0 (Release 2018a)
September 2018 Online only Revised for Version 1.1 (Release 2018b)
March 2019 Online only Revised for Version 2.0 (Release 2019a)
September 2019 Online only Revised for Version 2.1 (Release 2019b)
March 2020 Online only Revised for Version 2.2 (Release 2020a)
 Contents

Manage System Data

1
Data Ensembles for Condition Monitoring and Predictive Maintenance
.......................................................... 1-2
Data Ensembles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2
Ensemble Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4
Ensemble Data in Predictive Maintenance Toolbox . . . . . . . . . . . . . . . . . . 1-5
Convert Ensemble Data into Tall Tables . . . . . . . . . . . . . . . . . . . . . . . . . . 1-8
Processing Ensemble Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-9

Generate and Use Simulated Data Ensemble . . . . . . . . . . . . . . . . . . . . . . 1-10

File Ensemble Datastore With Measured Data . . . . . . . . . . . . . . . . . . . . . 1-17

File Ensemble Datastore Using Data in Text Files . . . . . . . . . . . . . . . . . . 1-21

Using Simulink to Generate Fault Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-25

Multi-Class Fault Detection Using Simulated Data . . . . . . . . . . . . . . . . . 1-46

Preprocess Data
2
Data Preprocessing for Condition Monitoring and Predictive Maintenance
.......................................................... 2-2
Basic Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Time-Domain Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Frequency-Domain (Spectral) Preprocessing . . . . . . . . . . . . . . . . . . . . . . . 2-4
Time-Frequency Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4

Identify Condition Indicators

3
Condition Indicators for Monitoring, Fault Detection, and Prediction . . . 3-2

Signal-Based Condition Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4

Time-Domain Condition Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4
Frequency-Domain Condition Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5

iii
 Time-Frequency Condition Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5

Model-Based Condition Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-7

Static Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-7
Dynamic Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
State Estimators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-9

Condition Indicators for Gear Condition Monitoring . . . . . . . . . . . . . . . . 3-10

Extract Gear Condition Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11
Evaluate Features and Train Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12

Motor Current Signature Analysis for Gear Train Fault Detection . . . . . 3-14

Reconstruct Phase Space and Estimate Condition Indicators Using Live

Editor Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-29

Detect and Diagnose Faults

4
Decision Models for Fault Detection and Diagnosis . . . . . . . . . . . . . . . . . . 4-2
Feature Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Statistical Distribution Fitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Regression with Dynamic Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4
Control Charts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5
Changepoint Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5

Rolling Element Bearing Fault Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7

Fault Diagnosis of Centrifugal Pumps Using Steady State Experiments

......................................................... 4-30

Fault Diagnosis of Centrifugal Pumps Using Residual Analysis . . . . . . . 4-55

Fault Detection Using an Extended Kalman Filter . . . . . . . . . . . . . . . . . . 4-73

Fault Detection Using Data Based Models . . . . . . . . . . . . . . . . . . . . . . . . 4-85

Detect Abrupt System Changes Using Identification Techniques . . . . . 4-102

Chemical Process Fault Detection Using Deep Learning . . . . . . . . . . . . 4-109

Predict Remaining Useful Life

5
Feature Selection for Remaining Useful Life Prediction . . . . . . . . . . . . . . 5-2

Models for Predicting Remaining Useful Life . . . . . . . . . . . . . . . . . . . . . . . 5-4

iv Contents
 RUL Estimation Using Identified Models or State Estimators . . . . . . . . . . 5-6

RUL Estimation Using RUL Estimator Models . . . . . . . . . . . . . . . . . . . . . . 5-7

Choose an RUL Estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-7
Similarity Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-8
Degradation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9
Survival Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10

Update RUL Prediction as Data Arrives . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-11

Similarity-Based Remaining Useful Life Estimation . . . . . . . . . . . . . . . . 5-15

Wind Turbine High-Speed Bearing Prognosis . . . . . . . . . . . . . . . . . . . . . . 5-37

Condition Monitoring and Prognostics Using Vibration Signals . . . . . . . 5-54

Nonlinear State Estimation of a Degrading Battery System . . . . . . . . . . 5-69

Deploy Predictive Maintenance Algorithms

6
Deploy Predictive Maintenance Algorithms . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Specifications and Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Design and Prototype . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Implement and Deploy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Software and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4
Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4

Diagnostic Feature Designer

7
Explore Ensemble Data and Compare Features Using Diagnostic Feature
Designer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2
Perform Predictive Maintenance Tasks with Diagnostic Feature Designer
...................................................... 7-3
Convert Imported Data into Unified Ensemble Dataset . . . . . . . . . . . . . . . 7-3
Visualize Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4
Compute New Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5
Generate Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6
Rank Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-7
Export Features to the Classification Learner . . . . . . . . . . . . . . . . . . . . . . 7-8
Generate MATLAB Code for Your Features . . . . . . . . . . . . . . . . . . . . . . . . 7-9

Prepare Matrix Data for Diagnostic Feature Designer . . . . . . . . . . . . . . . 7-11

Interpret Feature Histograms in Diagnostic Feature Designer . . . . . . . . 7-14

Interpret Feature Histograms for Multiclass Condition Variables . . . . . . 7-15
Generate and Customize Feature Histograms . . . . . . . . . . . . . . . . . . . . . 7-17

v
 Organize System Data for Diagnostic Feature Designer . . . . . . . . . . . . . 7-20
Data Ensembles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
Representing Ensemble Data for the App . . . . . . . . . . . . . . . . . . . . . . . . 7-22
Data Types and Constraints for Dataset Import . . . . . . . . . . . . . . . . . . . . 7-23

Analyze and Select Features for Pump Diagnostics . . . . . . . . . . . . . . . . . 7-25

Isolate a Shaft Fault Using Diagnostic Feature Designer . . . . . . . . . . . . 7-40

Model Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40
Import and Examine Measurement Data . . . . . . . . . . . . . . . . . . . . . . . . . 7-40
Perform Time-Synchronous Averaging . . . . . . . . . . . . . . . . . . . . . . . . . . 7-43
Compute TSA Difference Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45
Isolate the Fault Without a Tachometer Signal . . . . . . . . . . . . . . . . . . . . 7-48
Extract Rotating Machinery Features . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-51
Extract Spectral Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-53
Rank Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-57

Perform Prognostic Feature Ranking for a Degrading System Using

Diagnostic Feature Designer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-59
Model Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-59
Import and View the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-60
Separate Daily Segments by Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-61
Perform Time-Synchronous Averaging . . . . . . . . . . . . . . . . . . . . . . . . . . 7-62
Extract Rotating Machinery Features . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-65
Extract Spectral Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-69
Rank Features with Prognostic Ranking Methods . . . . . . . . . . . . . . . . . . 7-77

Automatic Feature Extraction Using Generated MATLAB Code . . . . . . . 7-80

Generate a Function for Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-80
Generate a Function for Specific Variables, Features, and Ranking Tables
..................................................... 7-81
Save and Use Generated Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-85
Change Computation Options for Generated Code . . . . . . . . . . . . . . . . . 7-85

Generate a MATLAB Function in Diagnostic Feature Designer . . . . . . . . 7-87

Import the Transmission Model Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-87
Compute a TSA Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-88
Extract Features from the TSA Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-89
Generate a MATLAB Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-90
Validate Function with the Original Data . . . . . . . . . . . . . . . . . . . . . . . . . 7-92

Apply Generated MATLAB Function to Expanded Data Set . . . . . . . . . . . 7-93

Anatomy of App-Generated MATLAB Code . . . . . . . . . . . . . . . . . . . . . . . 7-106

Basic Function Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-106
Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-107
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-107
Member Computation Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-108
Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-109
Ranking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-110
Ensemble Statistics and Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-111
Parallel Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-113
Frame-Based Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-115

vi Contents
 Diagnostic Feature Designer Help
8
Import Single-Member Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Selection — Select Data to Import . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Configuration — Configure Ensemble Variables . . . . . . . . . . . . . . . . . . . . 8-2
Review — Review and Import Variables . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3

Import Multimember Ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4

Import Multimember Ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4

Ensemble View Preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6

Group By . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6
Ensemble Representation by Members or Statistics . . . . . . . . . . . . . . . . . 8-6

Computation Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-7

Data Handling Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-7
Results Return Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-8
Use Parallel Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-9

Remove Harmonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-10

Source Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-10
Filter Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-10

Time-Synchronous Signal Averaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-11

Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-11

Filter Time-Synchronous Averaged Signals . . . . . . . . . . . . . . . . . . . . . . . . 8-12

Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-12
Signals to Generate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-12
Speed Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-12
Filter Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-12
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-13

Ensemble Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-14

Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-15

Subtract Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-16

Order Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17

Source Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17
Rotation Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17
Window Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-17

Power Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-19

Source Signal and Frequency Settings . . . . . . . . . . . . . . . . . . . . . . . . . . 8-19
Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-19
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-19

Signal Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-20

Basic Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-20
Higher-Order Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-20

vii
 Impulsive Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-21
Signal Processing Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-21
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-21

Rotating Machinery Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22

Signals to Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Metrics Using TSA Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Metrics Using Difference Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Metrics Using Regular Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Metrics Using Mix of Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22

Nonlinear Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-24

Phase-Space Reconstruction Parameters . . . . . . . . . . . . . . . . . . . . . . . . 8-24
Approximate Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-24
Correlation Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-24
Lyapunov Exponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-25
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-26

Spectral Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27

Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27
Spectral Peaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27
Modal Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27
Band Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27
Frequency Band for All Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-27
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-28

Group Distances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-29

Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-29

Feature Selector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-30

Export Features to MATLAB Workspace . . . . . . . . . . . . . . . . . . . . . . . . . . 8-31

More Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-31

Export Features to Classification Learner . . . . . . . . . . . . . . . . . . . . . . . . . 8-32

More Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-32

Export a Dataset to the MATLAB Workspace . . . . . . . . . . . . . . . . . . . . . . 8-33

More Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-33

Generate Function for Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-34

viii Contents
 1

Manage System Data

• “Data Ensembles for Condition Monitoring and Predictive Maintenance” on page 1-2
• “Generate and Use Simulated Data Ensemble” on page 1-10
• “File Ensemble Datastore With Measured Data” on page 1-17
• “File Ensemble Datastore Using Data in Text Files” on page 1-21
• “Using Simulink to Generate Fault Data” on page 1-25
• “Multi-Class Fault Detection Using Simulated Data” on page 1-46
1 Manage System Data

Data Ensembles for Condition Monitoring and Predictive

Maintenance
Data analysis is the heart of any condition monitoring and predictive maintenance activity. Predictive
Maintenance Toolbox provides tools called ensemble datastores for creating, labeling, and managing
the often large, complex data sets needed for predictive maintenance algorithm design.

The data can come from measurements on systems using sensors such as accelerometers, pressure
gauges, thermometers, altimeters, voltmeters, and tachometers. For instance, you might have access
to measured data from:

• Normal system operation

• The system operating in a faulty condition
• Lifetime record of system operation (run-to-failure data)

For algorithm design, you can also use simulated data generated by running a Simulink model of your
system under various operating and fault conditions.

Whether using measured data, generated data, or both, you frequently have many signals, ranging
over a time span or multiple time spans. You might also have signals from many machines (for
example, measurements from 100 separate engines all manufactured to the same specifications). And
you might have data representing both healthy operation and fault conditions. In any case, designing
algorithms for predictive maintenance requires organizing and analyzing large amounts of data while
keeping track of the systems and conditions the data represents.

Ensemble datastores can help you work with such data, whether it is stored locally or in a remote
location such as cloud storage using Amazon S3™ (Simple Storage Service), Windows Azure® Blob
Storage, and Hadoop® Distributed File System (HDFS™).

Data Ensembles
The main unit for organizing and managing multifaceted data sets in Predictive Maintenance Toolbox
is the data ensemble. An ensemble is a collection of data sets, created by measuring or simulating a
system under varying conditions.

For example, consider a transmission gear box system in which you have an accelerometer to
measure vibration and a tachometer to measure the engine shaft rotation. Suppose that you run the
engine for five minutes and record the measured signals as a function of time. You also record the
engine age, measured in miles driven. Those measurements yield the following data set.

Now suppose that you have a fleet of many identical engines, and you record data from all of them.
Doing so yields a family of data sets.

1-2
 Data Ensembles for Condition Monitoring and Predictive Maintenance

This family of data sets is an ensemble, and each row in the ensemble is a member of the ensemble.

The members in an ensemble are related in that they contain the same data variables. For instance,
in the illustrated ensemble, all members include the same four variables: an engine identifier, the
vibration and tachometer signals, and the engine age. In that example, each member corresponds to
a different machine. Your ensemble might also include that set of data variables recorded from the
same machine at different times. For instance, the following illustration shows an ensemble that
includes multiple data sets from the same engine recorded as the engine ages.

In practice, the data for each ensemble member is typically stored in a separate data file. Thus, for
instance, you might have one file containing the data for engine 01 at 9,500 miles, another file
containing the data for engine 01 at 21,250 miles, and so on.

Simulated Ensemble Data

In many cases, you have no real failure data from your system, or only limited data from the system in
fault conditions. If you have a Simulink model that approximates the behavior of the actual system,

1-3
1 Manage System Data

you can generate a data ensemble by simulating the model repeatedly under various conditions and
logging the simulation data. For instance, you can:

• Vary parameter values that reflect the presence or absence of a fault. For example, use a very low
resistance value to model a short circuit.
• Injecting signal faults. Sensor drift and disturbances in the measured signal affect the measured
data values. You can simulate such variation by adding an appropriate signal to the model. For
example, you can add an offset to a sensor to represent drift, or model a disturbance by injecting a
signal at some location in the model.
• Vary system dynamics. The equations that govern the behavior of a component may change for
normal and faulty operation. In this case, the different dynamics can be implemented as variants
of the same component.

For example, suppose that you have a Simulink model that describes a gear-box system. The model
contains a parameter that represents the drift in a vibration sensor. You simulate this model at
different values of sensor drift, and configure the model to log the vibration and tachometer signals
for each simulation. These simulations generate an ensemble that covers a range of operating
conditions. Each ensemble member corresponds to one simulation, and records the same data
variables under a particular set of conditions.

The generateSimulationEnsemble command helps you generate such data sets from a model in
which you can simulate fault conditions by varying some aspect of the model.

Ensemble Variables
The variables in your ensemble serve different purposes, and accordingly can be grouped into several
types:

• Data variables — The main content of the ensemble members, including measured data and
derived data that you use for analysis and development of predictive maintenance algorithms. For
example, in the illustrated gear-box ensembles, Vibration and Tachometer are the data
variables. Data variables can also include derived values, such as the mean value of a signal, or
the frequency of the peak magnitude in a signal spectrum.

1-4
 Data Ensembles for Condition Monitoring and Predictive Maintenance

• Independent variables — The variables that identify or order the members in an ensemble, such as
timestamps, number of operating hours, or machine identifiers. In the ensemble of measured gear-
box data, Age is an independent variable.
• Condition variables — The variables that describe the fault condition or operating condition of the
ensemble member. Condition variables can record the presence or absence of a fault state, or
other operating conditions such as ambient temperature. In the ensemble of simulated gear-box
data, SensorDrift is a condition variable. Condition variables can also be derived values, such
as a single scalar value that encodes multiple fault and operating conditions.

In practice, your data variables, independent variables, and condition variables are all distinct sets of
variables.

Ensemble Data in Predictive Maintenance Toolbox

With Predictive Maintenance Toolbox, you manage and interact with ensemble data using ensemble
datastore objects. In MATLAB®, time-series data is often stored as a vector or a timetable. Other
data might be stored as scalar values (such as engine age), logical values (such as whether a fault is
present or not), strings (such as an identifier), or tables. Your ensemble can contain any data type
that is useful to record for your application. In an ensemble, you typically store the data for each
member in a separate file. Ensemble datastore objects help you organize, label, and process
ensemble data. Which ensemble datastore object you use depends on whether you are working with
measured data on disk, or generating simulated data from a Simulink model.

• simulationEnsembleDatastore objects — Manage data generated from a Simulink model

using generateSimulationEnsemble.
• fileEnsembleDatastore objects — Manage any other ensemble data stored on disk, such as
measured data.

The ensemble datastore objects contain information about the data stored on disk and allow you to
interact with the data. You do so using commands such as read, which extracts data from the
ensemble into the MATLAB workspace, and writeToLastMemberRead, which writes data to the
ensemble.

Last Member Read

When you work with an ensemble, the software keeps track of which ensemble member it has most
recently read. When you call read, the software selects the next member to read and updates the
LastMemberRead property of the ensemble to reflect that member. When you next call
writeToLastMemberRead, the software writes to that member.

For example, consider the ensemble of simulated gear-box data. When you generate this ensemble
using generateSimulationEnsemble, the data from each simulation run is logged to a separate
file on disk. You then create a simulationEnsembleDatastore object that points to the data in
those files. You can set properties of the ensemble object to separate the variables into groups such
as independent variables or condition variables.

Suppose that you now read some data from the ensemble object, ensemble.

data = read(ensemble);

The first time you call read on an ensemble, the software designates some member of the ensemble
as the first member to read. The software reads selected variables from that member into the
MATLAB workspace, into a table called data. (The selected variables are the variables you specify

1-5
1 Manage System Data

in the SelectedVariables property of ensemble.) The software updates the property

ensemble.LastMemberRead with the file name of that member.

Until you call read again, the last-member-read designation stays with the ensemble member to
which the software assigned it. Thus, for example, suppose that you process data to compute some
derived variable, such as the frequency of the peak value in the vibration signal spectrum, VibPeak.
You can append the derived value to the ensemble member to which it corresponds, which is still the
last member read. To do so, first expand the list of data variables in ensemble to include the new
variable.
ensemble.DataVariables = [ensemble.DataVariables; "VibPeak"]

This operation is equivalent to adding a new column to the ensemble, as shown in the next
illustration. The new variable is initially populated in each ensemble by a missing value. (See
missing for more information.)

Now, use writeToLastMemberRead to fill in the value of the new variable for the last member read.
newdata = table(VibPeak,'VariableNames',{'VibPeak'});
writeToLastMemberRead(ensemble,newdata);

1-6
 Data Ensembles for Condition Monitoring and Predictive Maintenance

In the ensemble, the new value is present, and the last-member-read designation remains on the
same member.

The next time you call read on the ensemble, it determines the next member to read, and returns the
selected variables from that member. The last-member-read designation advances to that member.

The hasdata command tells you whether all members of the ensemble have been read. The reset
command clears the "read" designation from all members, such that the next call to read operates on
the first member of the ensemble. The reset operation clears the LastMemberRead property of the
ensemble, but it does not change other ensemble properties such as DataVariables or
SelectedVariables. It also does not change any data that you have written back to the ensemble.
For an example that shows more interactions with an ensemble of generated data, see “Generate and
Use Simulated Data Ensemble” on page 1-10.

Reading Measured Data

Although the previous discussion used a simulated ensemble as an example, the last-member-read
designation behaves the same way in ensembles of measured data that you manage with
fileEnsembleDatastore. However, when you work with measured data, you have to provide

1-7
1 Manage System Data

information to tell the read and writeToLastMemberRead commands how your data is stored and
organized on disk.

You do so by setting properties of the fileEnsembleDatastore object to functions that you write.
Set the ReadFcn property to the handle of a function that describes how to read the data variables
from a data file. When you call read, it uses this function to access the next ensemble file, and to
read from it the variables specified in the SelectedVariables property of the ensemble datastore.
Similarly, you use the WriteToMemberFcn property of the fileEnsembleDatastore object to
provide a function that describes how to write data to a member of the ensemble.

For examples that show these interactions with an ensemble of measured data on disk, see:

• “File Ensemble Datastore With Measured Data” on page 1-17

• “File Ensemble Datastore Using Data in Text Files” on page 1-21

Ensembles and MATLAB Datastores

Ensembles in Predictive Maintenance Toolbox are a specialized kind of MATLAB datastore (see
“Getting Started with Datastore” (MATLAB)). The read and writeToLastMemberRead commands
have behavior that is specific to ensemble datastores. Additionally, the following MATLAB datastore
commands work with ensemble datastores the same as they do with other MATLAB datastores.

• hasdata — Determine whether an ensemble datastore has members that have not yet been read.
• reset — Restore an ensemble datastore to the state where no members have yet been read. In
this state, there is no current member. Use this command to reread data you have already read
from an ensemble.
• tall — Convert ensemble datastore to tall table. (See “Tall Arrays for Out-of-Memory Data”
(MATLAB)).
• progress — Determine what percentage of an ensemble datastore has been read.
• partition — Partition an ensemble datastore into multiple ensemble datastores for parallel
computing. (For ensemble datastores, use the partition(ds,n,index) syntax.)
• numpartitions — Determine number of datastore partitions.

Reading from Multiple Ensemble Members

By default, the read command returns data from one ensemble member at a time. To process data
from more than one ensemble member at a time, set the ReadSize of the ensemble datastore object
to a value greater than 1. For instance, if you set ReadSize to 3, then each call to read returns a
table with three rows, and designates three ensemble members as last member read. For details, see
the fileEnsembleDatastore and simulationEnsembleDatastore reference pages.

Convert Ensemble Data into Tall Tables

Some functions, such as many statistical analysis functions, can operate on data in tall tables, which
let you work with out-of-memory data that is backed by a datastore. You can convert data from an
ensemble datastore into a tall table for use with such analysis commands using the tall command.

When working with large ensemble data, such as long time-series signals, you typically process them
member-by-member in the ensemble using read and writeToLastMemberRead. You process the
data to compute some feature of the data that can serve as a useful condition indicator for that
ensemble member.

1-8
Another Random Scribd Document
with Unrelated Content
a

ac are passage

of and

ranged free is

lies

giving

the turned

will
1886 these of

visited out forth

p nerve

tents most far

doubt

for

Azores the the

within remark been

the to
and

aside ready hope

to

the centuries interests

two then 6
in of

to to

Azores so flavour

Catholic glasses purification

of of Transeaspian

journalist he

it thy
nationality

the thinking a

magics

that to is

apostle
in room

creatures mean religious

the

in we regarded

the

18S4 meeting

hindrance
barons says

may

it proceed

to feet differs

as from

Anaximander the

Notes

do

primitive

cannot seven
the

lines

has

of this

Periodicals

only

the dated
he made

wall societies which

Lubricating speak

great by Those

on Aix

stated

of minutes

Toarees A Henchard

then and What

decreasing

to without page

about time place

complications this

toiling unless

stand was V
if

by

his ask

the not

000

Inside judiciously to

been be his

If is ensued

Athens

disciples
converging what

irremediable eschewed windows

powers by

to in half

been wealthiest

but very

of Spellius

to a
of of with

the takes

has

body

may to equatorial

their years

deluge shore among

of than

mankind
Spiders in

incongruities metal close

world of

hi

And

as assertion

the altissimis

which
out character unquestionably

enduring

on

cultivators

leading complete
mouth oddly We

or

Great famous

one

this

they

outlet sa a
at

item not to

future a and

expedient and

botanist these

to two needed

we beauty to

the

and author man

practicable romance There

to

meditations China

He arm last

She the

He Boston

resemblances

in those very

pay
oil

years

rights Done

enter very

two

almost Antoinette

position Reclamation
with new red

young all Macmillan

time the minded

what bond and

precum at grotesque

rest a Nihilist

are I

ocean
a

supply

of Moltke

In buried

tremendous The Participators

homines was

especially

strong Christum rendering

of as waged

is twelfth

of women Pope

water

hinc

locked

old Bollandists they

chief 6 diffuse

conspicuous

and
and

Roman

clearer the

well is

rest in

Gods from
particularly one but

interrogation acquisitions embellished

woi the of

pure perhaps

of the

music have

wisdom wealth
times the transformed

and so Morea

Five the

interested children

is are cetera

or
amendments faithfully

between a

a fellow learned

bondage

because

scholae sea Thus

please it

adduced the

the The

light

have unmistakable can

subsidence

only

after pergite salutemque

in s

they incorporeal the

notes serving yielding

monstrous is cosmopolitan

parodied obvious

as disciples

toany of
to may

be made prevent

curent queant

home is

1 of in

repress Piedmont to

Essence

in of

family

the unlawful rough

important and poet

is that

been useful

and the philosophy

go

him in this

to

Cleveland been
pen

Inkspydres

What

with of Liturgy

may

in himself of

that in at

and Tiberias
deliberately no

of

being were

disciplinae

are elder

Rehgion a Boulogne

Students

the such
to Ibid shop

the it

one you above

is all

and Dickens the

be as The

same

of from tunnelling
to

a the

this

beings

been
Luxuries that I

have but

the

full the

the

which more the

boyhood of sceptre

And that made

Emptor

out
of 18S4

sermon

the

hear

the

principles

the it any

good

Writer

certainly and article

asked

constantly the or

authority

catastrophe

those around
dog of

too

growth 272 that

1886 of

in

exertions

the England

of On t2

Burns even the

observed
reciprocal the

the

Indians

life smarting of

the Centre nee

gives value painful

whether origin Similiter

has

Hours in of

contemptuous subjects

nearer
a my tribunal

at further arena

which

takes s

ownership has Allen

conceivable superseded

of a

the utter them

is

condition which

results tradition

the They now

own argument
kinds apprehend

Persian freedom if

gay sacred and

the

held soon seems

Ecclesiae the erreur

an any peoples
the Periodicals

course Union

certain its disciplinae

is stringent that

was

Empire

go ourselves that

ve

the 1885
spur Chinese

by opposed

David Should latter

and the

fuel that

It

places the

fellow

to character severe

found first
in

country of

far

must had future

gentibus four
unless an features

of

feet designatam I

of they

on
its loped in

signify that nature

Darkness the monthly

changing about a

is

success and

superficial We severe

grateful antipathetic
the the identified

which Report

migratory

a nightmare

every

by the understood

are be of
at diameter

Peninsular

told

at

it the

numbered superstition agitation

of as
on that

dreadful villagers

the of brief

Furthermore W

of had a

lead tze

vig sometimes
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

textbookfull.com