The Modeling And Algorithm Of Eigenvalue Analysis For SSO In Complex Power System

As the reverse distribution of energy resources and the needs in our country,long-distance and high capacity electricity transmission is widely applied, however, theseries capacitor compensation technology and HVDC technology, which are both used inlong-distance and high capacity electricity transmission, will cause the subsynchronousoscillation(SSO). In recent years, the SSO has happened in several power plants in ourcountry, so it has a very important significance to analysis the power system which the SSOmight be happened in. Eigenvalue analysis method is a classical method to analysis thesmall disturbance stability, so it is fit for SSO. However, eigenvalue analysis method hassome limitations in complex power system.Using eigenvalue analysis method to analysis SSO is different from low frequencyoscillation analysis in the aspect of each component’s mathematical modeling, so there isno generic analysis software can be used. This thesis carried a work of eigenvalue analysismodeling and algorithm for the SSO in complex power system. For the purpose of applyingeigenvalue analysis method in any power system, a algorithm is designed. This algorithm isbased on the modular modeling, so it is also fit for software development.Firstly, Modular modeling of every element is carried on in this thesis, includinggenerator, transformer, transmission line and some controller. The model of networkelements cannot use quasi-steady-state model, which is fit for low-frequency oscillation,and electromagnetic transient process of every element is need to be concerned.On this basis of modeling, an algorithm of the formation of the state space is putforward, which is fit for eigenvalue analysis for SSO in complex power system. Thisalgorithm is based on modular modeling, concerning the coordinate transformation ofgenerator and system, and fit for any power system. For the purpose of realizing thealgorithm, this thesis carried a research on the selection rules of state variables and theformation rules of normal tree, and how these rules can be realized in software.Lastly, this algorithm is applied in two examples. We compare the eigenvalueanalysis results with time-domain simulation results to prove the the algorithm.