Full Length Research Paper
Abstract
Many of the activities associated with the planning and operation of water resource systems require forecasts of future events. For the hydrologic component that forms the input for water resource systems, there is a need for both short and long term forecasts of stream flow events, in order to optimize the real-time operation of the system or to plan for future expansion. For this historical inflow series from Sewa hydroelectric Project Stage-II which is a run-of-the river project has been used. For model development, 16 years historical inflows data of the river out of available 18 years inflow data has been used and the Artificial Neural Network Model has been trained to predict 2 years inflows. In order to accomplish this task, historical inflow series is employed for training, validating and testing with three different proportions of ratio 60:20:20, 80:10:10 and 90:05:05 were analyzed. The analysis of this study demonstrates the ability of neural network prediction model, to forecast quite accurately ten days inflows of two years ahead and generate synthetic series of ten days inflows that preserve the key statistics of the historical ten days inflows which in a way helps in effective utilization of available water, especially in a multipurpose context.
Key words: Artificial neural network, Levenberg-Marquardt algorithm, stream flow forecasting.
Copyright © 2024 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0