1.Key Laboratory of Smart Grid of Ministry of Education (Tianjin University), Tianjin 300072, China;2.Electric Power Science Research Institute of Hainan Power Grid Co., Ltd., Haikou 570105, China;3.State Power Investment Company Qinghai Photovoltaic Industry Innovation Center Co., Ltd., Xining 810008, China

When the high proportion of converters are connected to the weak grid, it is easy to produce the low-frequency dynamic phenomenon dominated by the interaction influence of “outer loop control-phase-locked loop (PLL)-weak-grid”. Because of technical security and other reasons, it is difficult to completely obtain the line parameters and converter control parameters on the network side, which makes it difficult to construct the low-frequency dynamic stability analysis model. Therefore, a data-driven parameter identification method for low-frequency dynamic stability analysis model is proposed. Firstly, the network parameter matrix is identified by the least square regression algorithm. Secondly, the system matrix is established, and the dynamic mode decomposition algorithm based on singular value decomposition is used to reconstruct it and replace it with the corresponding control link. By solving a set of linear equations, the control parameters of the outer loop and PLL of all converters can be obtained simultaneously. Then, combined with the device-side model, the complete small signal stability analysis model of a system is constructed. Finally, a case with two renewable energy stations connecting to the system is built in PSCAD/EMTDC to verify the accuracy of parameter identification and the effectiveness of the model, and to further analyze the low-frequency dynamic stability of the system.