講演抄録/キーワード |
講演名 |
2013-10-18 13:55
極値降雨予測のためのニューラルネットワークモデル ○ジュナイダ スライマン・Darwis Herdianti・廣瀬英雄(九工大) R2013-65 |
抄録 |
(和) |
これまで降雨予測には,日毎,月毎,年毎というような単位で行なわれていたが,ここでは5日間での降雨の最大値のトレンドを観測データとしながら極値降雨を予測することを試みる.
ニューラルネットは最近水文学の分野で活発に用いられており適切な入力パラメータを選択することで予測精度を上げることができる.ここでは粒子群最適化をこのニューラルネットに加えることで最適化を図っている.従来の統計的方法であるARIMAによる結果と比較した結果,提案方法は良好な結果を得た. |
(英) |
Several days of precipitation can increase the magnitude of accumulated water in a basin. This can cause the lower area of community and housing over flooded with rainfall water in a short time. Many researchers are using precipitation data for forecasting the number of rainy days in daily, monthly and yearly. However, with a maximum 5-day precipitation, we can predict the magnitude of precipitation within a specified period for example in a month, that may identified as precipitation extremes. Therefore, this study describes a method to forecast the trend of maximum 5-day precipitation in the following month using a hybrid of artificial neural networks (ANN) and particle swarm optimization (PSO). It is important to analyze the trend of extreme precipitation for future prediction of high precipitations events in the area of interest. ANN is widely applied in the hydrology field due to its non-linearity ability to map a non-stationary and seasonal data. Here, we have compared ANN with seasonal autoregressive integrated moving average (ARIMA) to measure their performances in forecasting next month maximum 5-day precipitation. Prior to model development in ANN, the significant input lags are determined using linear correlation analysis (LCA) and stepwise regression method (SLR), respectively. Results showed that ANN method is feasible in forecasting precipitation extremes when it is trained with the particle swarm optimization. |
キーワード |
(和) |
ニューラルネット / 豪雨 / 粒子群最適化 / 極値降雨 / ARIMA / / / |
(英) |
artificial neural networks / particle swarm optimization / extreme precipitation / seasonal autoregressive integrated moving average / / / / |
文献情報 |
信学技報, vol. 113, no. 249, R2013-65, pp. 7-12, 2013年10月. |
資料番号 |
R2013-65 |
発行日 |
2013-10-11 (R) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
R2013-65 |