EUVIMEDEUVIMED entdecken
Crossref

Proximal Policy Optimization Method for Enterprise Energy Management

Nazarii Luchyt, Mykola Medykovskyy

Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì · 22.06.2026 · 19 · 154-168

DOI: 10.23939/sisn2026.19.154 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì, online erschienen am 17.05.2026. Verfasst von Nazarii Luchyt und 1 weiteren Autor:innen. Die Arbeit nennt 51 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The article presents the results of research on the multi-criteria optimization problem in enterprise energy systems, aimed at improving the efficiency of energy-flow balancing. The results can be applied to develop adaptive control systems for hybrid industrial microgrids, as their operation modes are characterized by high stochasticity in renewable generation and consumption. The study models an environment as a Markov Decision Process to account for nonlinear constraints and internal interconnections among system elements. The study proposes an improvement to the Proximal Policy Optimization method within the Deep Reinforcement Learning paradigm, using an Actor-Critic architecture to implement an intelligent agent that serves as an automated energy management system. The method incorporates linear time interval scaling, which allows the model to adapt standard hourly equipment ratings to varying time steps. This facilitates the investigation of the impact of time discretization on decision-making quality while directly utilizing these standard parameters. Computer modeling methods were used to assess the impact of parameters on agent training stability. Experimental results demonstrate that reducing the stable steps window size shifts the training priority to active regulation modes, whereas the implementation of a deadband function effectively mitigates decision noise. The research results confirm the stabilization of operational modes and the optimization of the energy resource management strategy, which guarantees the fulfillment of energy balance conditions under typical load profiles. The proposed framework establishes a foundation for integrating economic criteria and optimizing energy storage capacity.

Zitieren

APA

Luchyt, N., & Medykovskyy, M. (2026). Proximal Policy Optimization Method for Enterprise Energy Management. Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì, 19, 154-168. https://doi.org/10.23939/sisn2026.19.154

Vancouver

Luchyt N, Medykovskyy M. Proximal Policy Optimization Method for Enterprise Energy Management. Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì. 2026;19:154-168. doi:10.23939/sisn2026.19.154

BibTeX
@article{Luchyt2026,
  author = {Luchyt, Nazarii and Medykovskyy, Mykola},
  title = {Proximal Policy Optimization Method for Enterprise Energy Management},
  journal = {Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì},
  year = {2026},
  volume = {19},
  pages = {154--168},
  doi = {10.23939/sisn2026.19.154}
}
Kostenlos starten →