Long-term hydropower scheduling using model predictive control approach with hybrid monthly-annual inflow forecasting

In this paper, a hybrid monthly-annual inflow forecasting approach is proposed and tested within a model predictive control framework for the long-term hydropower scheduling (LTHS). The inflow forecasts are provided on a monthly basis for a short horizon (close to present) and on an annual basis for the remaining optimization horizon, up to three years. The tests are conducted in a simulation environment with historical inflows for single reservoir hydrothermal systems. Results are compared with those using a monthly inflow forecasting approach and that from traditional stochastic dynamic programming approach, showing that the hybrid model is a promising approach to be used in the decision making process on LTHS problems.