A Microforecasting Module for Energy Consumption in Smart Grids

This work describes the module dedicated to the energy demand microforecasting developed within the 'Energy Router' research project and presents results obtained on pilot studies. This project aims to give prosumers a control and optimization device that allows them to schedule their own electric consumption in a demand response framework. The control of resources is performed in two stages, according to a hierarchical control structure. A cloud-based computation platform provides global control functions based on predictive control whereas a closed-loop local device controls all field components and actuators. Demand forecasts are available through the microforecasting module hereby presented. The module is designed to be flexible, adaptive, and to provide data with small time resolution. It includes alternative forecasting techniques, such as exponential smoothing, ARIMA and neural networks. Test results based on a data acquisition campaign are presented.