We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Game Theory Guided Data-Driven Multi-Entity Distribution Network Optimal Strategy.
- Authors
Xingmou Liu; Yuan Zuo; Ning Yang; Yao Xiao; Jadoon, Ammd
- Abstract
As the penetration of renewable energy sources (RES) continues to increase, more and more microgrids (MG) are interconnected with distribution system operators (DSO). To reduce system operational costs between MGs and DSOs, it is necessary to develop certain optimization strategies. This article proposes an optimized collaborative framework for modeling multi-entity distribution networks. In this model, DSOs are placed at the upper level to formulate policies, while MGs are at the lower level to respond in real-time to these policies. Furthermore, the multi-agent relationships in the model are described using Stackelberg game mechanisms, enhancing economic efficiency through dynamic gaming. Additionally, a data-driven multi-agent twin-delayed deep deterministic policy gradient (MATD3) algorithm is investigated to simulate the gaming process and improve the overall model's non-linear optimization capabilities. Considering that the simulation process can lead to violations of the energy storage system capacity constraints, a physics-based model is designed within the framework to ensure the safety of energy storage systems (ESS). Finally, compared to the MADDPG and penalty function methods, the proposed approach reduces the operational costs by 19.05%.
- Subjects
GAME theory; RADIAL distribution function; RENEWABLE energy sources; ENERGY storage; DEEP reinforcement learning; REINFORCEMENT learning
- Publication
Engineering Letters, 2024, Vol 32, Issue 4, p713
- ISSN
1816-093X
- Publication type
Article