High precision instant phase estimation method based on full-phase FFT

Abstract

The invention belongs to the technical field of digital signal processing, and relates to a high-accuracy instantaneous phase estimation method based on the full-phase FFT. The steps of the estimation method comprise directly sampling an input signal x(t) to obtain a discrete data sequence (x(n-N+1)..., x(n), ..., x(n+N-1)), storing 2N-1 data obtained by sampling in the RAM of a DSP, utilizing a convolution window of which the length is 2N-1 in number to perform windowing for the sequence, superimposing the data of which the distance is N sampling intervals, fast performing Fourier transformation to the N data, outputting N plurals which namely correspond as N spectral values Y(k), performing spectral peak searching, finding a spectral Y(k*) with the highest amplitude in the N spectral values Y(k), and performing rating to the imaginary part and the real part of the spectral Y(k*), thereby an instantaneous phase estimation result of a time n can be obtained by performing arctangent calculation to the ratio.

Description

technical field [0001] The invention belongs to the technical field of digital signal processing, and in particular relates to a method for estimating the instantaneous phase of signals in various engineering fields. Background technique [0002] In engineering fields such as communications, instrumentation, electric power, optical applications, and fault diagnosis, there are a large number of high-precision, efficient, and fast estimation problems for the phase of signals. For example, in a communication system using QAM and PSK modulation, only by quickly estimating the initial phase of the carrier in each symbol period at the receiving end can the constellation diagram be correctly restored and correct decoding be realized [1] ; Among all kinds of instruments, the phase meter is a common one. The basic function of the phase meter is to accurately measure and calculate the phase of the signal at each moment [2] ; and in the case of grid-connected closing of the power syst...

AI Technical Summary

Problems solved by technology

[0008] It can be seen that the disadvantages of this method are: (1) a reference sine wave needs to be generated, and the frequency of the reference signal is required to be exactly the same as that of the measured sine signal. If there is a slight deviation, it will bring serious deviations to the measurement results ; (2) The pulse counting signal needs to be specially generated, and the measurement accuracy largely depends on the basic pulse width T 0 , only by reducing T 0 Only in this way can the measurement accuracy be improved, which requires high hardware equipment; (3) only single-frequency signal can be measured and estimated

[0021] However, it can be proved through experiments and theory that all kinds of phase estimation methods based on FFT spectrum correction have the following defects: (1) FFT has spectral leakage phenomenon, when the signal contains multiple frequency components, there will be interspectral interference between various frequency components , which will greatly affect the measurement accuracy of the phase; (2) due to the existence of spectral leakage, these correction methods are especially not suitable for phase estimation occasions with dense spectrum; For the phase, the frequency is estimated first, and then the phase estimation is performed after obtaining the frequency deviation value. Therefore, a direct impact of this is that the error of the frequency estimation will be brought into the phase estimation.

[0022] The present invention proposes phase estimation based on the all-phase FFT method. The all-phase FFT is a new spectrum analysis method proposed by the patent applicant Wang Zhaohua in the patent document [22]. The spectrum analysis includes two aspects: the amplitude spectrum analysis and the phase spectrum analysis. The patent Literature [22] only studied the amplitude spectrum characteristics of all-phase FFT, and did not involve the study of phase spectrum. The present invention proposes a new phase estimation method based on the phase spectrum performance of all-phase FFT

Images