A time-frequency analysis method and system

Abstract

The invention discloses a time-frequency analysis method and system. The method includes the steps of: performing the first signal decomposition on the input signal; performing the second signal decomposition on the signal obtained by the first signal decomposition to obtain three signals with equal lengths; and performing the second signal decomposition on the three signals obtained by the second signal decomposition Windowing is performed separately to obtain three windowed signals; the three windowed signals are respectively subjected to symmetric discrete Fourier transform to obtain three frequency domain signals; the imaginary parts of the three frequency domain signals are respectively summed to obtain The real part of the three frequency domain signals is summed; the instantaneous frequency of the signal is calculated according to the summation result of the imaginary part; the DC component, origin phase and amplitude of the signal are calculated according to the calculated instantaneous frequency and the result of the real part summation. The invention realizes the calculation of instantaneous frequency, DC component, instantaneous amplitude and instantaneous phase, is suitable for stationary signals and non-stationary signals, has high calculation accuracy, has almost no requirement on signal length, has strong DC interference resistance and short calculation time.

Description

technical field [0001] The invention belongs to the technical field of signal analysis, and more specifically relates to a time-frequency analysis method and system. Background technique [0002] Signal is the carrier of information, and people's understanding of everything in the world and communication are based on the understanding of information. Obtaining information from a signal relies on signal analysis. Usually the actual signals are time-varying signals, and scientists often use non-stationary signals to refer to time-varying signals. Time-frequency analysis is widely used in the analysis of non-stationary signals, such as gravitational waves, animal language, brain waves and other signals. With the deepening of human understanding of the world, the requirements for time-frequency analysis methods are getting higher and higher. [0003] There are various time-frequency analysis methods in the prior art. For example, linear time-frequency representation, bilinea...

AI Technical Summary

Problems solved by technology

[0004] Problem 1 in the existing time-frequency analysis method: There is a large error in the calculation of frequency parameters for non-stationary signals

The theoretical basis of the time-frequency analysis of non-stationary signals is that any non-stationary signal can be regarded as the end-to-end connection of short-term stationary signals. Selecting more samples shakes this theoretical basis, which in turn causes a large gap in the estimation of the frequency parameters of non-stationary signals. error

[0005] There is still problem 2 in the existing time-frequency analysis method: the amplitude and phase change information is lost

Taking the short-time Fourier transform (STFT) as an example, the time-frequency diagram can only approximately reflect the relationship between time and frequency, and loses the amplitude and phase information

Images