A neural network operator oriented short-term and online load forecasting environment

This paper owes its origins to a project, still in progress at ENEL/ARC, which aims to investigate the application of artificial intelligence techniques and eventually to check their positive contribution in the field of short-term load forecasting. In particular, this article focuses on the construction problems of an integrated tool specifically designed to meet the needs of utility forecasters and power system operators. Even if the use of artificial neural networks for short-term forecasting had already been stressed in the past and no longer represents an innovative solution, the authors believe that their use for online forecasting in an adaptive and reliable calculation structure still represents a new subject of interest.