Optimization of Multi-User Personalized Demand Response for Virtual Power Plants Based on Federated Reinforcement Learning
Abstract
large fluctuations in user electricity consumption, high risks of data privacy, and insufficient adaptability of centralized optimization algorithms, which hinder the coordinated improvement of grid dispatch benefits and the personalized electricity needs of users. By using federated
reinforcement learning algorithms, the research constructs a demand response optimization system for virtual power plants suitable for multiple user situations. This system conducts local training and encrypted parameter combination to protect user data privacy while optimizing the
grid peak reduction and energy consumption with the improvement of user electricity usage satisfaction.
Keywords
Full Text:
PDFReferences
[1] Peng Z, Heng T. Hierarchical optimization of virtual power plants via sequential Game-Based Multi-Agent reinforcement learning [J].
Expert Systems With Applications, 2026, 310 131263-131263.
[2] Tian S, Xiao Q, Li T, et al. A Two-Layer User Energy Management Strategy for Virtual Power Plants Based on HG-Multi-Agent Reinforcement Learning [J]. Applied Sciences, 2025, 15 (12): 6713-6713.
[3] He S, Cui W, Li G, et al. Intelligent Scheduling of Virtual Power Plants Based on Deep Reinforcement Learning [J]. Computers, Materials & Continua, 2025, 84 (1): 861-886.
[4] Yao D J, Hao B W, Meng G Z, et al. Adaptive multi-agent reinforcement learning for dynamic pricing and distributed energy management in virtual power plant networks [J]. Journal of Electronic Science and Technology, 2025, 23 (1): 100290-100290.
DOI: http://dx.doi.org/10.70711/frim.v4i6.9686
Refbacks
- There are currently no refbacks.
