Author(s): Aiju You; Bingyao Lin; Youcheng Xu
Linked Author(s):
Keywords: Qiantang Estuary; Self-tuning forecastor; Neural network; River stage real-time forecasting
Abstract: It is difficult to establish river stage real-time forecasting model of Qiantang Estuary in hydrologic or hydraulic way because the riverbed changes violently but the necessary data of river configuration are limited. Based on system theory, this paper develops self-tuning forecastor and recursive neural network models, taking the upstream flow and the downstream tide as the input, and river stage of five forecasted stations as the output. Simulation result shows that the prediction accuracy is different along with the different stations and the different causes of high water-level formation.
Year: 2002