Title :
Climate variables and future hydrological scenarios: understanding causes, predicting consequences
Author :
Szczupak, Jacques ; De Macedo, Luiz H. ; Sallas, Edgard ; Pinto, Leontina
Author_Institution :
Electr. Electron. Dept., PUC-RIO, Brazil
Abstract :
This work presents a new model for future hydrological inflow forecast. Instead of trying to repeat the past, the proposed approach targets the future prediction based on climatological information able to explain hydrological behavior of the desired variables. A neural network, customized to extract sensitive data information, "translates" climatological forecasts into future hydrological scenarios.
Keywords :
hydroelectric power stations; neural nets; power engineering computing; climatological forecast; hydroelectric power system; hydrological inflow forecast; hydrological scenarios; neural network; Contracts; Data mining; Floods; Hydroelectric power generation; Load forecasting; Neural networks; Portfolios; Predictive models; Student members; Uncertainty;
Conference_Titel :
Probabilistic Methods Applied to Power Systems, 2004 International Conference on
Print_ISBN :
0-9761319-1-9
