|
| |
Structural Health Monitoring Using Statistical Pattern
Recognition
COURSE GOALS
Upon completion of this course, the student will be able to:
 | Describe structural health monitoring as a problem in statistical pattern
recognition |
| Describe and classify the primary methods of
structural health monitoring, with
their associated advantages and disadvantages |
| Describe the historical and current real-world applications of damage identification in
the aerospace, civil, and mechanical engineering fields |
| Discuss the primary practical implementation issues, including relevance of baseline
measurements, importance of measurement statistics, and aspects of comparative studies
between methods |
COURSE OUTLINE
(Note: All course instruction in English only.
Course outline subject to change without prior notice)
Introduction
| Motivation for Structural Health Monitoring (SHM) |
| Statistical pattern recognition (SPR) paradigm |
| Sensing issues for SHM |
| Fundamental axioms of SHM |
Historical Overview
| Discipline specific applications (aerospace, civil, rotating
machinery, offshore oil platforms) |
| Damage detection methods review (modal parameters, model updating
techniques) |
| Impact of other technologies on SHM |
Operational Evaluation
| Define system specific damage |
| Evaluate environmental/operational conditions |
Active SHM Sensing Technologies
| Lamb wave propagation/Impedance method |
| Time reversal acoustics |
| Sensor self-diagnostics |
| Active-sensing hardware development |
| Hardware/software integration |
Emerging SHM Sensing Technologies
| Sensing system design issues |
| Fiber optic sensing |
| Active versus passive sensing |
| Embedded Computing |
| Energy Harvesting |
Feature extraction
| Feature selection criteria |
| Limitations of commonly used features |
| Time series analysis & state-space representation |
| Frequency domain analysis |
| Features based on nonlinear analysis |
Introduction to Statistical Inference
| Supervised/unsupervised learning |
| Group classification |
| Regression modeling |
Basic Statistical Tools
| Statistical moments |
| Probability distributions and density estimation |
| Fisher’s discriminant |
| Principal component analysis |
Unsupervised Learning Methods
| Hypothesis testing |
| Statistical probability ratio test |
| Statistical process control |
| Outlier analysis |
Supervised Learning Methods
| Neural networks |
| Support vector machines |
| Clustering |
| Regression analysis |
Data Normalization
| Influence of environmental/operational variability |
| Test modification |
| Modeling of environmental effects |
| Auto-associative neural networks |
Examples/Applications
| I-40 bridge |
| Bridge concrete column |
| Three story shear building model |
| Light rail system |
| Fast patrol boat |
Software Demonstration
Concluding Remarks
Course evaluation / Informal discussions
|
|
