Structural Health Monitoring For Base Isolated Structures Using Parametric Models

1
Structural Health Monitoring For Base Isolated
Structures Using Parametric Models

• Gloria Shin
• University of California, Irvine
• Civil and Environmental Engineering

2
Outline

• Background
• General Procedure
• Approach
• Conclusions
• Future Work
• Acknowledgement

3
Structural Properties
4
System Identification

• Measured data U and Y
• where,
• u input data
• y output data

• u y

Dynamic Properties
5
General Procedure

Design the experiment and c collect data
Data
Polish and present data
Should data be f I filtered?
Data
Fit the model to . the data
Choice of model s structure
Model
Validate the m model
Model structure not OK
Data not OK
No
Can the model vbe accepted?
Yes
6
Raiousha at Keio University
7
Raiousha at Keio University
RF
Mass
2337
380
cm
7F
1904
380
6F
Superstructure 2838
1907
380
5F
1889
385
4F
2449
400
3F
2232
400
2F
2995
513
1F
6952
Isolation layer
Base
8
Types and Number of Base Isolation Used
9
Miyagi-Oki Earthquake

• May 26th, 2003
• 62428 p.m.
• Magnitude 7.0

y RFvv facc.
u BF a acc.
RAIOUSYA
10
Time VS. Acceleration
11
Polished and selected range of data
12
ARX and ARMAX

• Auto-Regression
• ARX
• ARMAX
• where

13
Long-hand equation

• ARX
• ARMAX

14
ARX Model
15
ARMAX model
16
Model Comparison

• ARX
• ARMAX

17
Model Good Enough?
18
Instrumental Variable Method

• ? Instrument or Instrumental Variable
• where a(e) transformation of e(t)
• ? parameters
• N number of finite time

19
Algorithm of IV method

• Estimate a model using a linear regression model
  of the form ARX
• Then, select correlation vector?(t)
• where L(q) a linear filter and
• x(t) simulated from input through
  a
• system
  where

20
Instrumental Variable Method
21
Conclusion
22
Future Work

• Experiment other available options to produce a
  better model
• Learn different types of models used in System
  Identification
• Cultivate engineering judgment of validating the
  model

23
Acknowledgements

• Dr. Maria Feng, University of California Irvine
• Dr. Akira Mita, Keio University
• Dr. Shirley Dyke, Washington University in St.
  Louis
• Dr. Richard Christenson, Colorado School of Mines
• Dr. Masato Abe, Tokyo University
• Dr. Yozo Fujino, Tokyo University
• National Science Foundation (NSF)
• Japan Society for Promotion of Science (JSPS)
• Masako Kamibayshi, Keio University