Abstract :
This paper provides a review of some of the main methodological issues and techniques which have become innovative in addressing the problem of forecasting daily loads and prices in the new competitive power markets. Particular emphasis is placed upon computationally intensive methods, including variable segmentation, multiple modeling, combinations, and neural networks for forecasting the demand side, and strategic simulation using artificial agents for the supply side.
Keywords :
costing; electricity supply industry; load forecasting; neural nets; power system analysis computing; agent based simulation; artificial agents; combinations; competitive power markets; computationally intensive methods; daily loads forecasting; demand side forecasting; multiple modeling; neural networks; prices forecasting; strategic simulation; supply side; variable segmentation; Computational modeling; Costs; Demand forecasting; Economic forecasting; Load forecasting; Monopoly; Power markets; Predictive models; Uncertainty; Weather forecasting;
