Small Signal Stability Analysis Method Research Based On Automated Frequency Domain Decomposition

With the gradual process of marketization of power industry,the scale and complexity of interconnected smart power grids are increasing continuously,the continuous application of fast excitation technology.And a heavy loading and long-distance characteristics of the power transmission has become a distinctive feature of modern smart power grids.The dynamic process of the power system is becoming more and more complicated,which make safety and stability issues become highlight.The underlying weak damping modes has become one of the outstanding problems affecting the safe operation of modern power system.Therefore,timely waring it and providing base data for dispatcher is of significance for ensuring normal operation of modern power system.Based on the random response data,this paper studies a small signal stability evaluation method under ambient excitation.The specific research contents are as follows:First of all,the steady state signal and the dynamic signal are compared and analyzed from the spectrum density point of view.A linear power system state equation is constructed by using the classical second order generator model.Theoretically,the time domain and frequency domain expressions of state variables are derived by mathematical analytic methods.The existence of electromechanical oscillation characteristics of power system under random response is proved mathematically.It provides a theoretical basis for identifying the electromechanical oscillation characteristics of the system by means of steady-state random signals.Secondly,because the automatic frequency domain decomposition method is mainly derived from the peak value method and the basic principle of frequency domain decomposition method.Therefore,the basic principle of peak value method is studied,and the basic theory of frequency domain decomposition method is deduced.Under the circumstance of environmental excitation,the identification results of the two algorithms are compared and analyzed.Compared with peak value method,frequency domain decomposition method is more accurate.But both approaches have limitations.It provides a theoretical basis for the further improvement of the new algorithm.Finally,a method for parameter identification of low-frequency oscillations is proposed On the basic principle of FDD algorithm,improved periodic graph method is added to estimate power spectral density matrix.At the same time,automatic peak recognition method is added to realize automatic recognition of feature frequency.The least square method was used to fit the damping ratio.The nonlinear power data is analyzed theoretically.The smoothing prior method is studied and the regularization parameters are determined.Thus,a low-frequency oscillation online identification scheme of power system is developed.AFDD algorithm is compared with SSAT method and stochastic subspace identification method.AFDD algorithm is accurate and robust.The weak damping mode of power system can be found in time.It provides stable data platform foundation for dispatcher to take restraining measures.The validity and accuracy of the proposed method are demonstrated by the IEEE4 machine 2 region system and IEEE16 machine 5 region system.