Deploying AI flexible algorithm in a Virtualized substation
In the energy landscape that DSOs find themselves in today, predicting demand and the generation capacity of the systems to which they distribute becomes essential. This is mainly due to the injection of renewable energy among the consumers of the network themselves, which in an increasing percentage are also becoming generators. Therefore, from the point of view of the distributor, it is vital to have a predictive system that avoids problems associated with imbalances between consumption and generation in the networks.
In this project, a digital twin of a transformer substation of Cuerva's network (DSO) using Barbara's technology (Edge AI Platform Developer) has been created. The installation of edge nodes with Barbara OS for monitoring and data processing for substation equipment, and Barbara Panel, which provides remote centralised management of all Edge nodes, enables full substation virtualization, Artificial Intelligence deployment, and the creation of its digital twin.
Furthermore, this MVP includes the development of a forecasting algorithm that is being retrained to be able to overcome the limitations described before and predict, with accuracy, the demand and production values of the consumers connected to the transformation centre.
The main goal of this project is to make accurate forecasts on the network’s future behaviour, taking into account, among other topics, problems related with the increasing integration of Distributed Energy Resources in LV networks.
The Use Case intend to distribute data analysis and decision-making among the different Edge Computing nodes, or distributed computing, located in adjacent TCs (e.g., in the same Medium Voltage ring) instead of sending this data to centralised servers (Cloud).
The objective is to enter a new market with great need through the design and development of a system that combines:
- IoT Nodes for distributed Edge Computing installable in TCs.
- An IoT platform for managing those nodes.
- Industrialised computing algorithms, mainly the following:
- Active and reactive power prediction on the consumer side in order to detect possible congestions and overvoltage events on the LV grid.
- Assess the flexibility assets in Cuerva´s grid and modelize them.
- Determine the flexibility needs of the system.
Cuerva’s algorithm and Barbara´s connectors have a proprietary licence, not an open source one. Those interested in the licence of our algorithms can reach us through Barbara Marketplace (https://market.barbaraiot.com/_home), or AIoD platform.
In this sense, the contribution to the AIoD platform is:
- Explanatory information on the operating requirements.
- Explanatory information on the type of service offered and its benefits.
- Video containing:
- Microservices deployed in the edge and Barbara Panel: platform that visualises the running connectors, databases, ingestor, running algorithm with logs.
- Influx queries: with the historical data captured and stored in the Influx Database.
- Grafana visualisation Panel: visualisation of faults, currents, overvoltages, line saturation, and flexibility algorithm prediction results.
Regarding the contribution to the AIoD Catalog:
- A sample of fully anonymised company data will be uploaded for free.
- Upload of generation and demand prediction algorithm for a fee. The purchase and licences will be made from the Barbara IoT Marketplace.
Finally for the contribution to the AIOD experiments:
- Contribution with network data and with the AI model for demand forecasting and generation.
- Technical support will be offered to members who want to implement the solution.