Detection method of multi-dense spectrum inter-harmonics based on adaptive TLS-ESPRIT and improved all-phase

It is difficult to extract parameters containing multiple dense spectrum inter-harmonics. To address this issue, this study proposes a multi-dense spectrum inter-harmonic detection method, which is based on adaptive TLS-ESPRIT method and improved all-phase spectrum analysis. Firstly, the Hankel matrix of the multi-dense spectrum inter-harmonic input signal is established and its singular value is achieved. The singular value's adjacent growth ratio is used to formulate an adaptive TLS-ESPRIT algorithm. Based on this, the number of frequency components in the input signal with multiple dense spectrum inter-harmonics is determined. Then, the frequency information in the input signal is extracted. The relationship between the peak spectrum line and the frequency offset in the all-phase spectrum analysis method is derived. The amplitude information of the multiple dense spectrum inter-harmonics is calculated. The phase information is obtained by the "phase invariant" feature of all-phase spectrum analysis. An experiment platform for inter-harmonic parameter detection is developed, which is based on the virtual instrument architecture. Simulations under the conditions of single dense spectrum, multiple dense spectrum, inter-harmonic amplitude fluctuation and noise interference are implemented. Results show that the proposed method can accurately detect the frequency, amplitude and phase of inter-harmonics. Compared with traditional detection methods, the proposed method has higher detection accuracy. The actual measurement results meet the requirements of GB/T 24337-2009 inter-harmonic measurement. © 2020, Science Press. All right reserved.