Time-frequency characterization method of solar oscillation signal

Abstract

The invention discloses a time-frequency characterization method for a solar oscillation signal, and belongs to the technical field of astronomy and signal processing. According to the method, the oscillation signal is normalized by using the mean value and the standard deviation, so that the influence of a direct-current component in the oscillation signal on signal power spectrum representationcan be removed; meanwhile, the time-frequency spectrum of the oscillation signal is represented by combining ConceFT, so that the problems of broadening, inaccurate frequency positioning and noise sensitivity of the frequency spectrum represented by CWT during time-frequency spectrum representation can be solved; yet further, the global power spectrum and least square method fitting are continuously utilized to estimate the noise in the time-frequency representation obtained by ConceFT for estimation; and the power spectrum at each moment in the time-frequency characterization is denoised according to the mode of dividing the number of the sampling points by 95% of the noise confidence level, so that the accuracy and precision of the time-frequency characterization of the oscillation signal can be further improved.

Description

technical field [0001] The invention relates to a time-frequency characterization method of a solar oscillation signal, belonging to the technical fields of astronomy and signal processing. Background technique [0002] Scientists have discovered from long-term observations and studies on the sun that structural features on the sun, such as sunspots, coronal loops, flares, chromospheric fibers, and needles, have quasi-periodic oscillations. The oscillation frequency of these structural features is an important clue for studying the structure of the solar magnetic field, the process of solar energy transmission, and the formation and evolution of the solar atmosphere. Therefore, accurate time-frequency characterization of solar oscillation signals has important physical and scientific significance. [0003] The time-frequency characterization method is to map the one-dimensional time-domain signal to a two-dimensional time-frequency plane to reflect the time-frequency charac...

AI Technical Summary

Problems solved by technology

However, although the wavelet transform provides the frequency spectrum distribution in time-frequency space, the represented spectrum is affected by the mother wavelet, which leads to spectrum broadening, etc., resulting in inaccurate positioning

At the same time, due to the limited ability of continuous wavelet transform to suppress signal noise, the accuracy of characterizing the signal spectrum is further affected.

Images