One Day-Ahead Prognosis of Energy Demand Using Artificial Intelligence and Biometeorological Indices

Abstract

Nowadays demand side management has become an important issue. Managing the energy resources in an optimal manner has become imperative among energy planners and policy makers. An integrated energy management approach is essential for the sustainable development of any electricity grid. The main objective of this work is the development of a forecasting model in order to predict one day ahead the energy demand of Tilos Island, Greece. For this purpose, an artificial neural network (ANN) forecasting model was developed to predict the energy demand of the entire island region. Prediction concerns 24-hours ahead on an hourly basis, with the developed ANN model being fed with historical data of energy demand, historical data of solar irradiation and historical data of a biometeorological index known as Cooling Power Index. Results show that the proposed methodology gives a sufficient forecast of energy demand in order to design an automated energy demand information tool for end-users such as distribution and transmission system operators.

Tilos, the first autonomous renewable green island in Mediterranean: A Horizon 2020 project

“…Tests on batteries have then been performed, including initial characterization, calculation of the energy capacities and efficiency at different power rates, impact of the different environmental and internal battery temperatures, identification and tuning of the state of energy parameters. Based on measured data of local load demand [4], wind speed [1] and solar radiation [2,3], preliminary forecasting models have been developed for Tilos island, principally with the use of Artificial Neural Networks.…”