Three Strategies for Coordinating Multiple Resources for Frequency Control in Power Systems Considering Diverse Computing and Communication Capabilities

Wenduo Sun,Wei Gu,Yiyuan Chen,Jiandong Wo,Xiang Ma,Xuan Fang

DOI: https://doi.org/10.1109/psgec58411.2023.10255912

2023-01-01

Abstract:This paper presents three distinct strategies to improve the power grid frequency control with varying computing and communication capabilities of the distributed resources, such as wind power, photovoltaics (PV), and energy storage (ES). The first strategy, termed traditional virtual droop and inertial control, is suitable for resources with limited computing and communication capabilities. In situations where distributed resources possess strong communication but weak computing capabilities, the second strategy combines a virtual synchronous generator model with a proportional power sharing (PPS) strategy to propose the expected power and allocate it based on adjustable capacity. The third strategy caters to distributed resources with strong computing but weak communication capabilities, employing a control mechanism based on multi-agent deep reinforcement learning (MA-DRL). The effectiveness of the proposed strategies is demonstrated through simulations of a two-region interconnected load frequency control system on the Matlab/Simulink platform, which maintains stable frequency even as the proportion of renewable energy increases.