물 자연 그리고 사람 - 물로 더 행복한 세상을 만들어가겠습니다.
HOMER&D성과논문실적

논문실적

Enhancing Hourly Reservoir Inflow Prediction Performance in Deep Learning using Improved Inflow Data Preprocessing - Integration of Event Identification and Exponential Smoothing?LSTM (IE2S-LSTM) 게시글의 제목, 학술지명, 저자, 발행일, 작성내용을 보여줌
Enhancing Hourly Reservoir Inflow Prediction Performance in Deep Learning using Improved Inflow Data Preprocessing - Integration of Event Identification and Exponential Smoothing?LSTM (IE2S-LSTM)
학술지명 iEMSS 2024 저자 최영돈
발표일 2024-06-24

Recently, the performance of LSTMs in analyzing rainfall-runoff has been outperforming physically based models in hydrology. However, data quality is still important to further improve performance of LSTM. In this study, we applied the Integration of Event Identification & Exponential Smoothing (IE2S) method to remove fluctuation of inflow which is still estimated by Simple Water Balance method. Then, preprocessed inflow was trained using LSTM and this approach proved improvement of reliability and variability for hourly inflow prediction. This research presents not only new deep learning methods, but also data-centric research is critical to improve performance of deep learning as an example of rainfall-runoff analysis.

목록