Development of Brazilian multi region short-term load forecasting model considering climate variables weighting in ANN model

The short-term load forecasting study is characterized as an estimative of the consumption pattern ranging from a day to a few months ahead, related to the operation planning, mainly in a Unit Commitment and power supply strategies, for example. Improve load forecasting models, mainly about errors reduction is so important, improving the energy efficiency, reducing energy losses and increasing financial optimizations. The objective of this work is develop a model to multi region short-term load forecasting, to next day hour forecast and next 7 days. Will be considered the effects of the weighted average application on weather variables, to correct the large extension pattern of Multi Regions, and a new approach of power demand patterns recognition, as an input variable of forecast model. Will be proposed a model based on Artificial Neural Networks. Correlation studies will be presented between the input variables and a MAPE comparison to load forecasting results.