AI Flex in the Edge: advanced algorithms for consumption prediction and generation

The motivation for this project lies in the need to anticipate and avoid problems related to imbalances between consumption and generation.

Context

The main objective of this project is the implementation of advanced algorithms for the prediction of the consumption and generation of a low voltage network, directly on edge nodes deployed in the infrastructure of a distribution company such as Cuerva.  

Due to the growing need for a predictive system that avoids problems associated with imbalances between consumption and generation in networks, especially since the increase in the injection of energy by consumers such as Distributed Energy Resources, Cuerva Energía and Barbara are collaborating on this project towards digitalisation by applying artificial intelligence.

Objective

The deployment of nodes on the Edge and the virtualisation of transformation centres will allow us to advance in the smart grid strategy promoted by Cuerva, with the following specific objectives:

• Data exploitation and processing for advanced algorithm training.
• Study of the state of sensorisation of the system.
• Development of Edge Computing algorithms using advanced techniques for: Prediction of demand and generation, prediction of network problems associated with this demand/generation and identification of potential actions on Flexible Service Providers.
  
• Integration in relevant environment of the Edge node technology offered by Barbara IOT.

Cuerva´s Role

From Cuerva's experience in leading the digitisation of the network, using control and monitoring techniques of its elements, Cuerva has advanced in the digitisation of strategic points. In Láchar (Granada) the network is highly digitised and monitored with the implementation of Barbara's IoT technology at the Edge. With these systems installed, an improvement is proposed through:

• The inclusion of new Raven algorithms that will allow the DSO not only to fully understand the behaviour of the network, but also to predict future problems in the short term.
• Propose practical solutions to these problems through the actions of various Flexibility Service Providers (FSPs). 
• The integration of these algorithms and decisions into a replicable framework will enable other European DSOs to improve the state of their networks.