Distributed new energy aggregation equivalent method based on adaptive reinforcement learning algorithm

Abstract This paper proposes a distributed new energy aggregation equivalent modeling method based on multi-agent collaborative adaptive reinforcement learning. Firstly, an adaptive reinforcement learning algorithm based on multi-agent collaboration is proposed to optimize the learning rate and behaviour search strategy. Then, the IFORL algorithm is used to determine the weight of each generalized load model, and the dynamic equivalent model of direct-drive wind turbine and photovoltaic power generation is constructed. Finally, the model is built in the simulation software and verified by simulation. The results show that the proposed method can identify the dynamic equivalent model parameters of distributed new energy, which is of great significance to improve the accuracy of distributed new energy modeling.