Partitioned computation of critical eigenvalues corresponding to spontaneous oscillations for power systems

With the integration of renewable resources into power grid, the partition management model emerges with high requirements of security and privacy. To satisfy this demand when analyzing the small signal stability of the partitioned power system, a partitioned computation method for critical eigenvalues related to the specified generators involving synchronous generators and inverters is proposed in this paper, which is based on analysis of essentially spontaneous oscillations in power system (AESOPS). First, the principle of system partition is introduced, to identify internal and boundary nodes for each region. Then, the partitioned modeling of the power network equations is constructed to adapt AESOPS for computing critical eigenvalues. Accordingly, the information of internal nodes is transformed into equivalent boundary node admittance and current information. Subsequently, boundary node voltages are computed and converted to internal voltage information. Meanwhile, critical eigenvalues related to the specified generators are calculated iteratively. Finally, the effectiveness and accuracy of the proposed method are verified by the improved 3-machine 9-bus system with integration of GFM inverter and a real-life power system. This method requires interaction with only a small amount of equivalent information corresponding to the specified generators among regions, rather than detailed system information of each region, which avoids handling the full model data of the entire system.