This paper presents a data processing technique for the analysis of output waveforms from transient stability programs. The method is an autoregressive process based on an extension of Prony's method. Using this approach an algorithm is proposed to determine the main parameters that identify a particular waveform. These parameters include the amplitude, frequency, damping constant and phase angle of each mode of oscillation in the swing curve. The algorithm incorporates these variables in the solution and there is no need for further numerical analysis. The effectiveness of the algorithm is verified by tests on both artificial waveforms and actual swing curves from a stability program. In the cases tested, the method has demonstrated satisfactory results with an acceptable degree of accuracy. The computational complexity, necessary precautions, other applications and improvements are also presented.
Read full abstract