Archives of Electrical Engineering

Abstract

In order to improve the detection accuracy of harmonics/inter-harmonics in power systems, a new method of harmonic/inter-harmonic detection based on synchrosqueezed transform and the Hilbert operator based on local spectrum maximum is proposed. Firstly, the spectrum of inter-harmonic signals is obtained through short-time Fourier transform, and the local maximum value of the spectrum in the frequency direction is detected. Then, based on the maximum frequency of the spectrum, a new frequency redistribution operator and synchronous extraction operator are constructed. It combines the operators with ridge detection for the decomposition of harmonic/inter-harmonic signals, so as to obtain a series of intrinsic mode function (IMF) components. Finally, the instantaneous amplitude and frequency of the IMF components is obtained by using the Hilbert operator. Meanwhile, according to the instantaneous frequency mutation point in the spectrum, the starting and ending time of transient harmonics/inter-harmonics is located accurately. Based on a low signal-to-noise ratio (SNR), the wavelet packet method (WP), Hilbert Marginal Spectrum method (HMS), synchrosqueezing wavelet transform method (SST), the Hybrid SST method (HSST), enhanced empirical wavelet transform (EEWT) and the proposed method are used to identify the harmonic/inter-harmonic parameters, respectively. The experimental results show that the proposed LMSST method can effectively separate the steady-state and transient harmonic/inter-harmonic signals, and has higher detection accuracy and better noise robustness.