Mode identification method for power system low frequency oscillation based on fastica and tls-esprit

Abstract

The invention discloses a low-frequency oscillation mode identification method of a power system based on FastICA and TLS-ESPRIT, obtains a low-frequency oscillation signal and performs zero-mean and whitening processing; constructs a separation matrix and calculates the inverse matrix of the separation matrix; performs an iterative cycle to update the separation matrix; After it is judged to be converged, the signal recovered by FastICA will be used as the new dominant signal for sampling; TLS-ESPRIT analysis will be performed on the new dominant signal to calculate the frequency, attenuation factor, amplitude and phase of each oscillation mode. The present invention can not only better retain the original characteristics of the signal in the case of noise interference, but also improve the signal-to-noise ratio; moreover, the identification is accurate and the precision is high, and the characteristics of the low-frequency oscillation signal can be accurately and comprehensively reflected, so it can The design of early warning and damping controller has a good application prospect.

Description

Low Frequency Oscillation Mode Identification of Power System Based on FastICA and TLS‑ESPRIT method technical field The present invention relates to the field of electric power technology system, relate in particular to the electric power system based on FastICA and TLS-ESPRIT Low frequency oscillation mode identification method. Background technique [0002] With the continuous expansion of the scale of interconnected power grids, the risk of low frequency oscillations is greatly increased. When the system is in some special When there is a weakly damped or negatively damped oscillation mode in a fixed operation mode, severe oscillation will lead to power system cracking and even danger. and stable operation of the entire grid. Therefore, it is found and timely and accurate to extract the low-frequency oscillation mode and obtain the characteristic parameters of the electric power. The safe and stable operation of the system is of great significance. Over the y...

AI Technical Summary

Problems solved by technology

[0004] Wavelet transform has the problem of difficult wavelet base selection; although Prony analysis is a widely used low-frequency oscillation identification method in today's academic circles, this method is very sensitive to noise and cannot obtain ideal results when the signal-to-noise ratio is lower than 50dB; In actual application, the HHT method itself will have modal aliasing, endpoint effects and false components; although the ESPRIT algorithm can accurately identify the oscillation mode of the system, its performance will also vary in the case of colored noise and low signal-to-noise ratio. decrease

Images