基于K-ESPRIT的快速宽频测量方法

Fast Wide-frequency Measurement Method Based on Kurtosis-Estimation of Signal Parameters via Rotation Invariance Technique

随着电力系统中电力电子装置渗透率越来越高,电气量的宽频特性越来越显著,并引发了新的涉网稳定问题,对电力系统信号的监测频带范围和响应速度提出了新的要求。因此,提出了一种快速宽频测量方法,实现了宽频信号的快速跟踪与高精度测量。利用空间谱估计的"超分辨率特性"实现了短时间窗的宽频测量。为提高算法在低信噪比条件下的稳定性,提出了基于峭度的信号频率成分数估计方法,保证了测量精度。最后,在硬件上实现该方法,通过仿真测试、硬件测试以及现场测试,验证了所提方法的有效性。

With the increasing penetration of electronic devices in power systems,electrical quantities are increasingly exhibiting wide-frequency characteristics, and thus new stability issues of power grid have arisen. New requirements for monitoring frequency range and response speed of power system signals are put forward. In this paper, a fast wide-frequency measurement method is proposed to realize the fast and high-precision measurement of the wide-frequency signals. The super-resolution characteristics of spatial spectrum estimation are used to achieve the wide-frequency measurement in a short time window. In order to improve the algorithm stability in low signal noise ratio(SNR) conditions, a kurtosis based method for estimating the number of signal frequency components is proposed,which guarantees the measurement accuracy. Finally, the proposed method is implemented on hardware. The results of simulation, hardware and field tests verify the validity of the proposed method.