Management of inflow forecasting studies

Abstract

In hydroelectric systems, water inflow is important to coordinate a cascade and define the energy price. This paper presents a method for managing inflow forecasting studies with a specific module for advanced assessment. The main goal is to provide a structure that facilitates the analysis of water inflow prediction models. A case study has been applied to five mathematical models based on linear regression, artificial neural networks, and hydrologic simulation. These models present daily and monthly inflow forecasts for a set of hydroelectric plants and monitoring stations. The benefits of the proposed method are analyzed in four situations: water inflow prediction, performance evaluation of a specific model, research tool for inflow forecasting, and comparison tool for distinct models. The results show that implementation of the proposed method provides a useful tool for managing inflow forecasting studies and analyzing models. Therefore, it can assist researchers and engineering professionals alike by improving the quality of water inflow predictions.