1.School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;2.State Grid Zhejiang Electric Power Research Institute, Hangzhou 310014, China

With the rapid development of renewable energy, the proportion of distributed generators and microgrids integrated into the distribution network continues to increase, and the simulation study of the distribution network becomes more difficult. Homology identification and equivalent modelling is a conventional method to reduce the order of the grid model. There have been many studies using homology theory to model the equivalence of distributed generators. However, most of the existing homology identification methods of distributed generators are suitable for droop or virtual synchronous control, and the universal homology identification method for power electronic devices under multiple control methods has not been fully studied. And most homology identification of distributed generators has not considered the impact of the network impedance on homology. Under this background, this paper proposes a method of homology identification based on relaxed variation coefficients, which can comprehensively quantify the homology degree of multi-state subsystems and more intuitively reflect the homology degree of each subsystem. First, to address the difficulty in determining the threshold of subgroups, this paper proposes to decide the optimal number of subgroups based on the maximum relaxed variation coefficient curve within the group. Then, the network model is further contracted by Kron simplification. Finally, the effectiveness of the proposed method is verified by an improved case of the IEEE 33-bus distribution system.