This paper studies the technique using neural network in order to estimate recession characteristics of river flow rate into a dam after rainfall. The rainfall is classified by similarity such as accumulation amount of rainfall, rainfall intensity, and base flow to the peak rainfall by using the cluster analysis. The training of the neural network is carried out by using the one out of the rainfalls of the similar group. The one data of remainder is used to asses the performance of the neural network on the accuracy of the estimated RTC of the river flow rate. The estimated error by cluster analysis becomes small.
J-STAGEがリニューアルされました！https://www.jstage.jst.go.jp/browse/-char/ja/
