Flux and Level Prediction based on An Wavelet Neural Network Flood Model

摘要：

This paper uses wavelet neural networks for flood prediction. It presents an flood prediction model and give an rapid algorithm. The water flux and level are used as input and output variables in the prediction model. The analysis of time-frequency characteristic of wavelet transformation is given. The prediction precision is improved by combining low frequency feature vector with high frequency ones. High frequencies of signals, which are middle or low numbers, are decomposed into small scales in wavelet space in flood flux and level analyses, and low frequencies of signals, which are large numbers, are decomposed into big scales. The model developed in this paper provided a new procedure for flood prediction. The experiment shows that the application of wavelet neural networks in flood prediction can give more accurate results.

关键词： Wavelet neural network Flood prediction model Flood flux and level

作者: Zhang Shaozhong Yuan Juqin

作者单位: Institute of Electronics and InformationZhejiang Wanli UniversityNingbo, Zhejiang 315100, China

会议类型: 国际会议

会议名称: 2010 Third International Symposium on Knowledge Acquisition and Modeling(第三届知识获取与建模国际研讨会 KAN 2010)

会议地点: 武汉

会议语种:英文

页码: 67-70

在线出版日期: 2010-10-20（万方平台首次上网日期，不代表论文的发表时间）