Short-Term Forecasting of the Occurrence Time of Strong Wind Speed during a Typhoon based on LSTM for Sea-Crossing Bridge Operation
|
|
|||||||||||
Bibliographic Details
| Author(s): |
Jae-Yeong Lim
(Department of Civil and Environmental Engineering, Seoul National University, South Korea)
Se-Jin Kim (Department of Civil and Environmental Engineering, Seoul National University, South Korea) Ho-Kyung Kim 
(Department of Civil and Environmental Engineering, Seoul National University, South Korea)
|
||||
|---|---|---|---|---|---|
| Medium: | conference paper | ||||
| Language(s): | English | ||||
| Conference: | IABSE Conference: Risk Intelligence of Infrastructures, Seoul, South Korea, 9-10 November 2020 | ||||
| Published in: | IABSE Conference Seoul 2020 | ||||
|
|||||
| Page(s): | 230-236 | ||||
| Total no. of pages: | 7 | ||||
| DOI: | 10.2749/seoul.2020.230 | ||||
| Abstract: |
Vehicles running on sea-crossing bridges are vulnerable to strong winds with instantaneous speeds of over 20 m/s. Bridge operators should secure the safety of the bridge users by limiting vehicle speeds or restricting the traffic when wind speed measured on the bridges exceeds a certain threshold value. To guarantee the safety of the bridge users during typhoons, an accurate forecasting of the strong winds would be essential. In this study, an Artificial Neural Network (ANN) was considered to model the occurrence characteristics of the strong wind speed at the sea- crossing bridge during typhoons. The Long Short-Term Memory (LSTM), which is generally used in the time-series analysis, was applied. This research utilized 16 years of wind speed data acquired by sensors located on a suspension bridge in South Korea and Best Track data of typhoons from the Regional Specialized Meteorological Center (RSMC) in Tokyo. |
||||
| Keywords: |
Strong Winds Prediction K-means clustering Artificial Neural Network LSTM Strong winds Warning System
|
||||