Electric Power Systems Research, Год журнала: 2024, Номер 237, С. 111007 - 111007
Опубликована: Авг. 31, 2024
Язык: Английский
Electric Power Systems Research, Год журнала: 2024, Номер 237, С. 111007 - 111007
Опубликована: Авг. 31, 2024
Язык: Английский
International Journal of Electrical Power & Energy Systems, Год журнала: 2024, Номер 159, С. 110022 - 110022
Опубликована: Май 27, 2024
Deep Reinforcement Learning (DRL) is effective in solving complex, non-linear optimization problems, which particularly relevant energy management within Integrated Energy Systems (IESs). However, DRL approaches conventionally focus on single-objective policy learning, inadequate for the multi-objective tasks commonly encountered IESs management. To improve this, these typically combine multi-objectives, such as operating cost objective and safety into a single reward function using scalarization techniques. This reduces fidelity interpretability of space limits its applicability to wide range address challenges, this paper presents novel framework called Multi-Agent Multi-Objective (MAMODRL). combines value decomposition gradient methods achieve Pareto-optimal solution. The initially formulated Markov decision process. Then, an advanced MAMODRL architecture developed, includes networks facilitate optimization. Finally, based definition dominance, Pareto frontier approximated A case study suggests that proposed approach ensure safe operation system, threshold set at forming with conditions. Compared traditional approaches, more flexible, interpretable, capable making multi-dimensional decisions.
Язык: Английский
Процитировано
4Applied Energy, Год журнала: 2024, Номер 379, С. 124902 - 124902
Опубликована: Ноя. 22, 2024
Язык: Английский
Процитировано
4International Journal of Hydrogen Energy, Год журнала: 2025, Номер unknown
Опубликована: Фев. 1, 2025
Язык: Английский
Процитировано
0Energy, Год журнала: 2024, Номер 295, С. 130996 - 130996
Опубликована: Март 16, 2024
Язык: Английский
Процитировано
3Electric Power Systems Research, Год журнала: 2024, Номер 237, С. 111007 - 111007
Опубликована: Авг. 31, 2024
Язык: Английский
Процитировано
3