Advanced Power Generation & Demand Forecasting Considering the Complete Energy Matrix using an Artificial Neural Network

This paper introduces the application of an Artificial Neural Network to perform the power generation & demand forecasting of a regional electricity network. The main technical contribution of this article is the consideration of many variables (35) that involve the complete energy matrix information to calculate the generated and demanded power accurately. The final results showed that the related correlation factor of the net generation forecast and net demand forecast using Artificial Neural Network are very adequate (greater than 0.95). Furthermore, the percentage of error between the predicted value and the actual power generation and demand values is reduced (up to 0.22 %). Finally, the results of the long-term forecast (up to 30 days) also proved to be adequate considering the correlation factor and percentage of error associated with the power generation and demand forecast.