As a result of the FLEXERGY project, a set of new software modules was developed.
Machine learning technology was used by Efacec to develop a load forecast module. This module uses Neural Network (Multi-Layer Perceptron) with hyper-parameters tuning.
In addition, machine learning technology was also used by Efacec to implement a PV generation forecast module. This module uses decision trees based on extreme gradient boost algorithms.
INESC TEC played a relevant role in the project, as their researchers developed an API providing features for optimal dispatching of controllable assets, using mathematical based optimization algorithms such as mixed integer linear programming.
That API was integrated by Efacec in the application server, the ES Manager.
The developed software modules provide features which enable the following use cases:
• Technical and economic optimization of hybrid park (PV/Wind/Storage) supported by
Battery Energy Storage System (BESS)
• BESS as a buffer for Electric Vehicles (EVs) integration
• Holistic optimization of microgrids integrating BESS
• Maximization of BESS useful time through Life Cycle Assessment (LCA)
The ES Manager provides user friendly human-machine interface (HMI) features. Some HMI examples are shown below.