Householder-based Prony Method For Identification Of Low-frequency Oscillation In Power System

With the development of interconnection planning and construction of cross-regional power grids,according to the current power planning and actual operation issues,this has put forward new requirements for containing low-frequency oscillations,and it is still necessary to study low-frequency oscillations.Due to the error in the Prony toolbox that comes with the grid operation analysis software in actual work,this article has carried out the following work to address the shortcomings of the traditional Prony algorithm:The mathematical models of traditional Prony algorithm and SDM-Prony algorithm are introduced,and the implementation principles,advantages and disadvantages are analyzed.Through the specific calculation examples,the f itting effects of the two Prony algorithms under ideal conditions and noisy interference are verified.It is verified that the SDM-Prony algorithm is better than the traditional Prony algorithm,but the comprehensive analysis of the two algorithms is not h igh in calculation accuracy.In practice,there will be a certain deviation in the calculation,and there is a problem of insensitivity to low-frequency oscillation parameter estimation.Therefore,it is urgent to find a method suitable for improving the calculation accuracy of the higher-order matrix,and further improve the Prony algorithm.In order to solve the effect of the rounding error generated in the least squares calculation on the calculation accuracy,the Householder transform is introduced into the traditional Prony algorithm least squares calculation characteristic equation coefficients,and the Prony method based on the Householder transform is used to identify the low frequency oscillation mode of the power system.And the adaptive search is introduced into the least squares solution eigenvalues to determine the initial point and initial search direction,so that the iterative calculation process has a certain self-adaptation.The algorithm effectively reduces the order of the matrix,reduc es the iteration amount of the system through symmetric transformation,improves the operation speed of the Prony algorithm,and further improves the calculation accuracy.Through a specific calculation example,the ideal signal and the white noise signal with different signal-to-noise ratios are compared and analyzed,and the actual three-phase short-circuit fault in a certain area is used as an example for simulation calculation.The SDM-Prony algorithm,the new algorithm in this paper and the Prony toolbox included in the power system analysis software are compared and analyzed.Finally,the results are verified by calculating the parameters with the implicit restart algorithm of the small interference program.It is proved that the proposed new algorithm can accurately identify the parameters of the oscillating signal,which is theoretically reasonable and has certain feasibility and reliability.It can be considered for practical calculation and analysis to provide reference and help for grid mode calculation and planning.