Work structure

WP1. User Requirements, Architectural Device and Specifications

Objective: Analysis and documentation of user requirements in two cycles, before implementation of each version of the overall system. Create detailed system specifications and complete architecture at the level of subsystems and for the overall system in two cycles that will lead to its smooth implementation in the EU3.

WP2. Research and Development of Machine Learning Techniques for Demand Prediction and RES Production

Objective: Research and develop innovative machine learning techniques based on learning with multiple targets and in deep neural networks aiming at low response time and high accuracy in continuous day-ahead load forecasting, generation forecasting of renewable energy sources and personalized short- and medium-term load forecasting at the end-user level low & medium voltage.

WP3. Software Parts and System Integration

Objective: Development of individual software components and system integration. Initially, a subsystem will be developed a for the collection and pre-processing of historical data, meteorological data, other ata (holidays, important events, etc.), intraday data, and telemetering data of low voltage consumers. At the same time, the techniques that will be developed and found optimal in WP2 will be integrated into subsystems for day-ahead load forecasting, intraday load forecasting and personalized load forecasting. Finally, all sub-systems will be integrated into an integrated system with a friendly environment
user.

WP4. Design and Implementation of Pilot Operation and System Evaluation

Objective: Control and evaluation of functionality and performance of the subsystems and the whole system based on defined objectives.