EnergyTwin: A Multi-Agent System for Simulating and Coordinating Energy Microgrids

EnergyTwin: A Multi-Agent System for Simulating and Coordinating Energy Microgrids Jakub Muszynski1[0009-0000-2797-6044], Ignacy Walużenicz1[0009-0008-6269-362X], Patryk Zan1[0009-0002-6772-8111], Zofia Wrona1[0009-0008-2863-5557], Maria Ganzha1[0000-0001-7714-4844], Marcin Paprzycki2,3[0000-0002-8069-2152], and Costin Bădică4[0000-0001-8480-9867] 1 Warsaw University of Technology, Warsaw, Poland, jakub.muszynski2.stud@pw.edu.pl, ignacy.waluzenicz.stud@pw.edu.pl patryk.zan.stud@pw.edu.pl, zofia.wrona.dokt@pw.edu.pl, maria.ganzha@pw.edu.pl 2 Systems research Institute Polish Academy of Sciences, Warsaw, Poland 3 University of Technology and Arts, Warsaw, Poland, marcin.paprzycki@ibspan.waw.pl 4 University of Craiova, Craiova, Romania, costin.badica@edu.ucv.ro Abstract. Microgrids are deployed to reduce purchased grid energy, limit exposure to volatile tariffs, and ensure service continuity during disturbances. This requires coordinating heterogeneous distributed en- ergy resources across multiple time scales and under variable conditions. Among existing tools, typically, power-system simulators capture physical behaviour but assume centralized control, while multi-agent frameworks model decentralized decision-making but represent energy with no physical grounding. In this context, the EnergyTwin is introduced, an agent-based microgrid simulation environment that couples physically grounded mod- els with forecast-informed, rolling-horizon planning, and negotiations. Each asset is modeled as an agent, interacting with a central agent that obtains forecasts, formulates predictions, and allocates energy through contract-based interactions. EnergyTwin targets tertiary-layer decision making and is extensible for digital-twin use. Its feasibility was evalu- ated in a university campus microgrid scenario where multiple planning strategies were compared. Achieved results show that forecast-driven rolling-horizon planning increases local energy self-sufficiency, maintains higher battery reserves, and reduces exposure to low-resilience operat- ing states. They demonstrate also potential of EnergyTwin as platform supporting research on resilient, negotiation-driven microgrids. Keywords: Microgrids · Multi-agent systems · Energy management · Physical modelling · Distributed optimization arXiv:2511.20590v1 [cs.MA] 25 Nov 2025 2 Muszynski et al. 1 Introduction Modern energy distribution systems evolve. Large consumers, e.g. university campuses, industrial parks, or research facilities are expected to meet sustainabil- ity targets (e.g., actual reductions in Scope 2 emissions [37] and higher on-site renewable penetration), limit exposure to volatile energy prices, and demonstrate continuity of service during power disturbances. Meeting these obligations typi- cally involves: (i) increasing the share of local, low-carbon generation, e.g. via photovoltaic (PV) arrays, (ii) electrifying new end uses, e.g. transportation, and (iii) adding controllable flexibility in the form of battery storage and dispatchable loads. Microgrids – electrically bounded subsystems that can coordinate these assets, operate independently from the upstream utility, making them a practical vehicle for achieving cost, carbon, and resilience objectives simultaneously. Let us consider a university campus: a network of lecture halls, research laboratories with sensitive equipment, student dormitories, and athletic facilities. This merits introduction of microgrid, to meet sustainability goals and enhance resilience to power grid disturbances. Here, rooftops of campus buildings are fitted with PV arrays, a central battery energy storage system (BESS) is installed to store excess solar power, and a fleet of electric vehicle (EV) chargers is deployed. To maximize the energy self-sufficiency, this system must meet the operational objectives, e.g.: (1) minimize electricity costs, (2) reduce use of external energy, and (3) ensure that critical facilities (e.g. laboratories and data center) remain powered during upstream contingencies (e.g. feeder outages, voltage sags, etc., creating short-term instability or price spikes). This requires, among others, to dynamically (1) arbitrage time-of-use tariffs and shaving peaks, and (2) maximize on-site renewable energy self-consumption. This straightforward scenario encapsulates the complexity and promise of microgrids [28,29]. Obviously, the “campus energy manager” must coordinate a heterogeneous collection of Distributed Energy Resources (DERs), each with own operational characteristics and constraints. Such coordination spans multiple, often conflicting, timescales. (1) Inverters connected to PVs and battery systems must react in milliseconds, to maintain local voltage and frequency stability [4]. (2) BESS charging and discharging schedule must be optimized over minutes and hours, to align with solar production forecasts and fluctuating electricity prices [20,3]. (3) EV charging must be managed to avoid overwhelmin

…(Full text truncated)…

This content is AI-processed based on ArXiv data.