| Title |
Prediction of Floating Debris for the Dam Basin Using the Artificial Neural Network |
| Authors |
최성욱(Choi, Seongwook) ; 김준성(Kim, Junsung) ; 강형식(Kang, Hyeongsik) |
| DOI |
https://doi.org/10.12652/Ksce.2025.45.2.0173 |
| Keywords |
인공신경망; 유송잡물; 영향인자; 댐 유역; 차단망 Artificial neural network; Floating debris; Influence factor; Dam basin; Barrier boom |
| Abstract |
Artificial intelligence has been increasingly applied across various academic fields and can be effectively utilized to predict the amount of floating debris generated in watersheds. An artificial neural network model was developed in this study to predict the accumulation of floating debris in dam basin. The annual floating debris collection data was used to estimate the floating debris accumulation. Key influencing factors for floating debris generation were identified by analyzing the mechanisms of floating debris occurrence. The final input variables for the artificial neural network model were selected using Pearson correlation coefficient analysis and included annual average inflow, annual maximum flood discharge, and watershed area. The artificial neural network model was trained and validated using floating debris data and various hydrological data corresponding to the key influencing factors. Model performance was confirmed through correlation analysis, ensuring adequate training and validation. However, the prediction performance was significantly lower in certain dam watersheds, primarily due to the low reliability of data collected during periods of minimal debris accumulation. A comparative analysis with an existing linear regression model was also conducted, and methods to improve the prediction accuracy of floating debris accumulation were explored. This study demonstrates that the artificial neural network model can reliably predict floating debris accumulation given an extensive dataset on floating debris generation. |