Download - Vibration Based Fuzzy-Neural System for Structural Health Monitoring Lakshmanan Meyyappan (Laks)
Vibration Based Fuzzy-Neural System for Structural Health Monitoring
Lakshmanan Meyyappan (Laks)
1. Objectives
The main goal is to develop a practical real-time structural health monitoring system using smart systems engineering concepts and tools.
2. Overall System
Vibration Data Signal Processing, Feature Extraction and Data Cleansing
Fuzzy Logic Detection System
Possible Damage?
Bridge is perfect
Neural Network Prediction System
Damaged?
Damage Value
Small Damage
Medium Damage
Large Damage
YES
NO
YES
NO
3.1.1 Vibration Signatures
Advantages: NDE Technique Global Analysis Normal Operation of the
Structure Small Reliable Less Expensive (both initial
and operating costs) Sensitive
Disadvantages: Unsupervised Learning
Mode Data Accuracy (Potential
problem with any type of data)
3.3 Experiment
Teardrop
Bridge
4. Damage Detection
For simplicity of explanation the data collected with the sensors attached to the above five locations are used.
4. Damage Detection
Relationship between the members remains the same that is member 3 has the highest power spectrum value in all of the above cases followed by member 1, 5, 4 and 2 respectively
5. Fuzzy Logic Decision System
Goal: To take power spectrum values of various members as input and predict a possible damage
Method: Fuzzy Ranking System
5.1 Fuzzy Ranking System
Fuzzy Ranking based on Fuzzy Integral values calculated using the formula:
where a, b, c are the vertices of the triangular membership functions
Alpha is the index of optimism and it varies between 0 and 1
abcFTI )1(2
1)(
6. Neural Network Prediction System
Goal: To make the final prediction on the condition of the bridge
Inputs:
Fuzzy logic system output
Speed of the vehicle ( Speed Gun output)
6. Neural Network Prediction System
Input : 100 Data Points (speed) Target : 100 Data Points (Power
Spectrum Peak Value) Algorithm : Back Propagation (LM Method) Layers : 2 Layers [15 1] Transfer
Functions : [Tansig Purelin] Error Rate : 1e-8 Max Epochs : 1500
6. Neural Network Prediction System
T A N S I G
b1
+
IW 1,1
P U R E L I N
b2
+
LW 1,1
1 1
Output Layer
a1 = tansig (IW 1,1 p1 + b1) a2 = purelin (LW 2,1 a1 + b2)
100 15 X 1
15 X 100
100 X 1
15 X 1
15 100 X 1
100 X 15
15 X 1
100 X 1
p1
n1
a1
n2
a3 = y
100 X 1
Hidden Layer Input