Integrated bridge health monitoring, evaluation and alert system using neuro-genetic hybrids
The bridge monitoring system which can analyze and predicts damage level of bridges due to earthquake loads is not yet available in Malaysia. Even though Malaysia is not an earthquake-prone country, earthquake from neighboring countries could affect the stability of the existing bridges in Malaysia....
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2014
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/77877/1/ReniSuryanitaPFKA2014.pdf http://eprints.utm.my/id/eprint/77877/ http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:98564 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Teknologi Malaysia |
| Language: | English |
| Summary: | The bridge monitoring system which can analyze and predicts damage level of bridges due to earthquake loads is not yet available in Malaysia. Even though Malaysia is not an earthquake-prone country, earthquake from neighboring countries could affect the stability of the existing bridges in Malaysia. This study aims to analyze the performance of the bridge subject to earthquake loads and develop the intelligent monitoring system to predict the bridge health condition. The case study is the Second Penang Bridge Package-3B. The Intelligent System consists of the Artificial Neural Networks (ANN) and Genetic Algorithm (GA) hybrid model to obtain the optimum weight in the prediction system. The ANN inputs are 4633 data of the bridge response accelerations and displacements while the outputs are the bridge damage levels. Damage levels are obtained through nonlinear time history analyses using SAP2000. The damage level criterion is based on FEMA 356 focusing on Immediate Occupancy (IO), Life Safety (LS) and Collapse Prevention (CP) level. This intelligent monitoring system will display the alert warning system based on the prediction results with green for IO, yellow for LS and Red color for CP level. According to the results, the best performance of the displacement as data input in the prediction system is 2.2% higher than the acceleration data. This study is verified with pushover-static test to the mini-scale piers model in ratio 1:34. The first crack occurred on the base of Pier 1 when the lateral load is 9 kN, 12 kN for Pier 2 and 8 kN for Pier 4. Maximum displacement at Pier 1 is 10 mm while at Pier 2 and Pier 4 is 6 mm individually. The intelligent monitoring system can greatly assist the bridge authorities to identify the bridge health condition rapidly and plan the bridge maintenance routinely. |
|---|