Seismic fragility analysis for geostructures using ANN-based response surface
Abstract
Seismic fragility curve is an effective tool to predict the degree of damages to the structure probabilistically under seismic load. When the seismic fragility curve is to be prepared in general structures such as bridges or concrete structures, the seismic load is put as the random variable and then the fragility curve is established. However, in the case of the geostructures such as the cut slope and soil levee, there are uncertainties in the related geotechnical parameters. Therefore, they should be interpreted by considering the uncertainties. In this study, seismic fragility curves for levee and slope were prepared considering the uncertainty in the geotechnical parameter and using the pseudostatic analysis. For the probabilistic analysis, Monte Carlo Simulation(MCS) method was used based on the coefficient of variation(COV) provided from the previous studies. As far as MCS method is concerned, the number of simulation shall be increased to get a certain degree of reliability when the probability of failure is low. In this process, MCS method is unfavorable because it requires more time and expenses. To overcome these shortcomings, the response surface method using the artificial neural network(ANN) that improves the efficiency in preparing the fragility curve was applied. For the review of the applicability, the results were compared with the MCS-based fragility curves. In addition, fragility curves that depend on the variation of water level of levee were prepared using the ANN-based response surface. The results showed that the new method can get the fragility curve which is similar to the MCS-based fragility curve, and can be efficiently used to reduce the analysis time. Acknowledgements: This research was supported by the Korea Agency for Infrastructure Technology Advancement (KAIA) with funding from the Ministry of Land, Infrastructure and Transport of the Korean government (16SCIP-B065985-04).
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMNH41B1787P
- Keywords:
-
- 4326 Exposure;
- NATURAL HAZARDSDE: 4328 Risk;
- NATURAL HAZARDSDE: 4330 Vulnerability;
- NATURAL HAZARDSDE: 4337 Remote sensing and disasters;
- NATURAL HAZARDS