<?xml version="1.0" encoding="UTF-8"?>
<collection>
<dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:invenio="http://invenio-software.org/elements/1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd"><dc:identifier>doi:10.3390/app16041844</dc:identifier><dc:language>eng</dc:language><dc:creator>Fernández, Gregorio</dc:creator><dc:creator>Sanz Osorio, J. F.</dc:creator><dc:creator>Rocca, Roberto</dc:creator><dc:creator>Luengo-Baranguan, Luis</dc:creator><dc:creator>Torres, Miguel</dc:creator><dc:title>Practical considerations for the development of two-stage deterministic EMS (CLOUD–EDGE) to mitigate forecast error impact on the objective function</dc:title><dc:identifier>ART-2026-148328</dc:identifier><dc:description>The growing penetration of Distributed Energy Resources (DERs)—such as photovoltaic generation, battery energy storage, electric vehicles, hydrogen technologies and flexible loads—requires advanced Energy Management Systems (EMS) capable of coordinating their operation and leveraging controllability to optimize microgrid performance and enable flexibility provision to the grid. When the physical, electrical, and economic system model is properly defined, the main sources of performance degradation typically arise from forecast uncertainty and temporal discretization effects, which propagate into sub-optimal schedules and infeasible setpoints. This paper proposes and tests a two-stage deterministic EMS architecture featuring rolling-horizon planning at an upper layer and fast local setpoint adaptation at a lower layer, jointly to reduce the impact of forecast errors and other uncertainties on the objective function. The first stage can be deployed either on the edge or in the cloud, depending on computational requirements, whereas the second stage is executed locally, close to the physical assets, to ensure timely corrective action. In the simulated cloud-executed planning case, moving from hourly to 15 min granularity improves the objective value from −49.39€ to −72.12€, corresponding to an approximate 46% reduction in operating cost. In our case study, the proposed second-stage local adaptation can reduce the mean absolute error (MAE) of the EMS performance loss by approximately 50% compared with applying the first-stage schedule without local correction. Results show that this two-stage hierarchical EMS effectively limits objective-function degradation while preserving operational efficiency and robustness.</dc:description><dc:date>2026</dc:date><dc:source>http://zaguan.unizar.es/record/169398</dc:source><dc:doi>10.3390/app16041844</dc:doi><dc:identifier>http://zaguan.unizar.es/record/169398</dc:identifier><dc:identifier>oai:zaguan.unizar.es:169398</dc:identifier><dc:relation>info:eu-repo/grantAgreement/EC/HORIZON EUROPE/101096192 /EU/REplicable, interoperable, cross-sector solutions and Energy services for demand side FLEXibility markets</dc:relation><dc:identifier.citation>Applied Sciences (Switzerland) 16, 4 (2026), 1844 [31 pp.]</dc:identifier.citation><dc:rights>by</dc:rights><dc:rights>https://creativecommons.org/licenses/by/4.0/deed.es</dc:rights><dc:rights>info:eu-repo/semantics/openAccess</dc:rights></dc:dc>

</collection>