
Leader
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum.
Hosting Facilities
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum.
AIMS
Creation of Insights
AI-enabled decision support tool for optimal microgrid sizing and energy control
Challenge: Energy system Digital Twin decision support tool
AIMS is a holistic, AI-enabled decision support tool for grid operators, offering both sizing insights for microgrid components, such as EV chargers, batteries, and PV systems, and energy control for flexible assets. It utilizes Reinforcement Learning agents and deep learning forecasting models to deliver actionable recommendations that reduce energy costs and improve renewable self-consumption. Integrating with both AID4SME playgrounds, the Digital Twin for simulation-based sizing and the real microgrid playground for live validation, AIMS unifies long-term planning and real-time operations into a single, intuitive UI developed by Plegma Labs.
Leader
Hosting Facilities
The AID4SME playground is hosted by ELES, a Slovenian transmission and distribution system operator and member of the AID4SME consortium. ELES provides an actual microgrid setup with historical measurements, along with pricing scenarios and optimization objectives reflecting real conditions on the Slovenian electro-energy market. The three validation scenarios are: Slovenian electrical energy pricing, spot-market energy pricing, and grid balancing service.
Background
Electric mobility is accelerating across Europe, reshaping demand at the grid’s edge, with Electric Vehicles (EVs), and specifically electric cars, already accounting for 15.6% of new EU registrations. In parallel, buildings account for ~40% of all energy consumed in the EU, with approximately half of EU gas consumption being attributed to them, further enforcing the need for optimizing an already stressed electricity grid before expanding the EV charging infrastructure. Looking ahead, Europe’s electricity demand from EV charging alone is projected to rise from 9 TWh (2021) to 165 TWh by 2030, accounting for ~6% of total EU electricity consumption, an increase that local networks and grid operators must plan and operate for with precision. Hence, AIMS targets specific challenges that grid operators face:
- Challenge #1 – Microgrid sizing under uncertainty: Determine the optimal configuration and capacities of microgrid components (i.e., generation, storage, and EV charging) when demand, renewable output, and prices have significant uncertainty due to their dynamic nature, and when investment decisions must balance Capital Expenditures (CAPEX), Operational Expenditures (OPEX), and grid stability.
- Challenge #2 – Operational coordination of flexibility: Real-time adjustment of flexible assets, such as building loads, EV charging, and batteries, so that energy costs are minimized without violating comfort or network constraints.
- Challenge #3 – End-to-end decision support: A holistic operator-centric tool that unifies long-term design exploration with short-term operational support is needed, leveraging a digital twin for year-round evaluation and validating performance with real measurements so simulated gains translate to field impact.
Solution
AIMS constitutes a holistic AI-enabled decision support tool for grid operators that offers both sizing insights related to microgrid components, such as EV chargers, as well as real-time energy control for a microgrid’s flexible assets. It uses Reinforcement Learning agents that produce actionable recommendations that reduce energy costs and size the integration of additional EV charging without destabilizing microgrids. AIMS also integrates high-granularity deep learning time-series forecasting models, such as RNNs and Transformers, to predict energy needs and production within microgrids. The holistic decision support tool incorporates multiple AI-powered modules within two data flows: planning and operation. Regarding planning, candidate microgrid setups are evaluated with different generation, storage, conversion, and EV charging capacities by utilizing the Digital Twin over full-year horizons. Regarding operation, the tool consumes short-term forecasts regarding microgrid energy needs and production and computes actionable energy control setpoints for flexible assets while respecting comfort and network limits. The UI will present both outcomes to grid operators.
Objectives
- Develop a microgrid decision support sizing module utilizing AI algorithms for optimal sizing of microgrid components(e.g.EV chargers), using the AID4SME DT playground.
- Develop a real-time microgrid energy control AI agent that provides optimal adjustments of flexible asset operation mode (e.g. building loads, EV charging, batteries), using high-granularity predictions of energy needs and production.
- Combine the developed sizing and control agents and deep learning forecasters into an AI-enabled decision support tool for grid operators, with intuitive UI for monitoring and insights.
- Demonstrate the developed tool with data from the AID4SME microgrid playground.
KPIs achieved
Development
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum. Sed sit amet consequat ex. Pellentesque id odio felis. Etiam tellus dui, maximus non vehicula ut, suscipit vel ex. Maecenas quis lorem ut sem aliquet fermentum. Duis vel nulla hendrerit augue aliquet posuere. In erat arcu, accumsan ut rhoncus in, fringilla at elit. Morbi non gravida ligula. Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum. Sed sit amet consequat ex. Pellentesque id odio felis. Etiam tellus dui, maximus non vehicula ut, suscipit vel ex. Maecenas quis lorem ut sem aliquet fermentum. Duis vel nulla hendrerit augue aliquet posuere. In erat arcu, accumsan ut rhoncus in, fringilla at elit. Morbi non gravida ligula.
Feedback
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum. Sed sit amet consequat ex. Pellentesque id odio felis. Etiam tellus dui, maximus non vehicula ut, suscipit vel ex. Maecenas quis lorem ut sem aliquet fermentum.
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum. Sed sit amet consequat ex. Pellentesque id odio felis. Etiam tellus dui, maximus non vehicula ut, suscipit vel ex. Maecenas quis lorem ut sem aliquet fermentum.
Mauris et nulla tellus. Nulla consectetur placerat nunc placerat bibendum. Sed sit amet consequat ex. Pellentesque id odio felis. Etiam tellus dui, maximus non vehicula ut, suscipit vel ex. Maecenas quis lorem ut sem aliquet fermentum.