A high-resolution time-frequency analysis method based on adaptive ultra-wavelet transform

By combining short, low bandwidth and long and high bandwidth wavelets in time-frequency analysis, adjusting the order and center frequency of ultra-wavelets, the problem of low time domain resolution of traditional wavelet transform is solved, and higher time-frequency resolution is achieved rate and effective detection of low-frequency shadows and formation characteristics.

CN116660985BActive Publication Date: 2024-12-03CHENGDU NORMAL UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202310045326.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-01-30
Publication Date
2024-12-03
Estimated Expiration
2043-01-30

AI Technical Summary

Technical Problem

In time-frequency analysis, the time domain resolution of traditional continuous wavelet transform is low, making it difficult to effectively detect low-frequency shadows and prominent formation characteristics, and the selection of parent wavelet parameters is difficult, affecting the resolution.

Method used

Adaptive ultra-wavelet set transformation combines short and low bandwidth wavelets with longer and higher bandwidth wavelets. By adjusting the order and center frequency of ultra-wavelets, the time and frequency resolution are improved, and Morlet wavelets are used as the parent wavelet. , and calculate the complex convolution of the input signal through adaptive ultra-wavelet transformation.

Benefits of technology

A higher time-frequency analysis resolution is achieved, the dependence on parent wavelet parameters is reduced, the detection ability of low-frequency shadows and formation characteristics is improved, and the robustness and resolution of the method is enhanced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116660985B_ABST
    Figure CN116660985B_ABST
Patent Text Reader

Abstract

The present invention provides a resolution time-frequency analysis method based on adaptive ultra-wavelet set transform, which relates to the field of exploration technology. The resolution time-frequency analysis method based on adaptive ultra-wavelet set transform includes continuous wavelet transform and adaptive ultra-wavelet set transform. The adaptive ultra-wavelet transform continuous signal x(t) is defined as following formula (1), where CWT x (τ,s) is a function of the translation parameter τ and the scale parameter s, and ψ*(t) is the complex conjugate of ψ(t). Therefore, in order to solve the problem that the traditional continuous wavelet transform is difficult to select suitable mother wavelet parameters, the time and frequency resolution of the time-frequency analysis results are improved by combining a short low-bandwidth wavelet with a longer high-bandwidth wavelet. We only need to optimize the parameter C, because the ultrawavelet can be adaptively obtained by formula (9), and the value range of o is set [1-30]. Therefore, the proposed method has higher resolution and is more powerful than CWT.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of geophysical exploration, in particular to a resolution time-frequency analysis method based on adaptive ultrawavelet set transformation. Background Art

[0002] Time-Frequency Analysis (TFA) maps a one-dimensional signal into a two-dimensional space of time and frequency, and can express the dynamic time-frequency response of non-stationary signals. TFA has been widely used in data processing and interpretation. Fourier Transform (FT), Continuous Wavelet Transform (CWT) and their variants are widely used in time-frequency analysis. A new method based on Short-Time Fourier Transform (STFT) has been proposed and applied to seismic spectrum decomposition, thereby achieving higher time-frequency resolution. However, the time-frequency resolution based on Short-Time Fourier Transform depends on the selected window length. Others have improved the resolution of non-stationary signals by Fourier transforming the inverse continuous wavelet transform to convert the time-scale mapping into the time-frequency mapping. In the past decade, many methods have been applied to time-frequency analysis, such as Empirical Wavelet Transform (EWT), Synchronous Squeezing Wavelet Transform (SSWT), Synchronous Squeezing Generalized S Transform (SSGST), etc. These methods improve the resolution in the frequency domain by squeezing the energy in the frequency domain. However, they have low resolution in the time domain. Scientists in the subject area of ​​geophysics pointed out that high-frequency resolution can help detect low-frequency shadows associated with hydrocarbon reservoirs, while high time resolution can enhance stratigraphic features.

[0003] To overcome the limitation and improve the resolution, we propose to apply the Adaptive Superlets Transform (ASLT) method to seismic time-frequency analysis, that is, combining short, low-bandwidth wavelets with longer, high-bandwidth wavelets to achieve higher time and frequency resolution. For this motivation, we first briefly introduce the principle of the adaptive super wavelet transform. Then, we use one-dimensional synthetic signals to demonstrate the performance of wavelet transform and the proposed method. Finally, we apply the proposed method to field data to show its potential in detecting low-frequency shadows and highlighting stratigraphic features. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a resolution time-frequency analysis method based on adaptive ultra-wavelet set transform, which combines short, low-bandwidth wavelets with longer, high-bandwidth wavelets to simultaneously improve the time and frequency resolution of the time-frequency analysis results, thereby solving the problem that the traditional continuous wavelet transform (CWT) is affected by the selected Morlet wavelet.

[0005] To achieve the above objectives, the present invention is implemented by the following technical solutions: a resolution time-frequency analysis method based on adaptive ultra-wavelet set transform, including adaptive ultra-wavelet set transform and continuous wavelet transform, wherein the adaptive ultra-wavelet set transform continuous signal x(t) is defined as following formula (1);

[0006]

[0007]

[0008] CWT x (τ, s) is a function of the translation parameter τ and the scale parameter s, ψ*(t) is the complex conjugate of ψ(t), given by equation (2). Equation (2) is also called the mother wavelet. The mother wavelet is stretched and squeezed by changing the parameter s. In addition, the change of s will not only change the center frequency of the mother wavelet, but also change the window length. Therefore, the parameter selection of the mother wavelet plays an important role in continuous wavelet transform and adaptive ultrawavelet set transform. The mother wavelet used is the Morlet wavelet, which is defined as formula (3);

[0009]

[0010] f c is the center frequency of the mother wavelet, f b is the bandwidth parameter and j is the imaginary unit. For the Morlet wavelet, the frequency f is given corresponding to the scale parameter s;

[0011]

[0012] Preferably, the adaptive super wavelet transform can combine and calculate the geometric mean of multiple spectrograms to improve the time-frequency resolution. Based on these mathematical foundations, the adaptive super wavelet transform (ASLT) combines short, low-bandwidth wavelets with longer, high-bandwidth wavelets to improve the time and frequency resolution.

[0013] Preferably, S1. First, we use the period number c of the wavelet instead of f b parameter;

[0014]

[0015] S2. Combining equation (3) and equation (5), we establish a set of Morlet wavelets ψ fc,c , with a fixed center frequency f c and different period numbers c, called ultrawavelets, which are defined as formula (6);

[0016]

[0017] where o is the order of the ultrawavelet, c1, c2, ..., c o is the number of cycles of each wavelet in the superwavelet;

[0018] S3. Secondly, using equation (7), we use wavelet transform to calculate the complex convolution of the input signal x(t), which is similar to the traditional CWT, except that a set of ultrawavelets is used in equation (7);

[0019]

[0020] Here * is the convolution operator;

[0021] S4. Finally, using equation (8), we obtain the response of a single wavelet in the ensemble;

[0022]

[0023] Here R[SL fc,o ] represents the result of super wavelet transform (SLT).

[0024] Preferably, S3 further includes the following steps:

[0025] The adaptive superwavelet (ASL) adjusts its order to the center frequency to compensate for the decreasing bandwidth as the frequency increases. Here, the order of the superwavelet is given by:

[0026]

[0027] Among them, min and max are the minimum and maximum orders of the ultrawavelet, respectively, and f min and f max are the minimum and maximum center frequencies respectively, and [] is the nearest integer operator.

[0028] Preferably, x1(t), x2(t) and x3(t) are defined as:

[0029] x1(t)=cos[2πt(10+t / 6.2) / 256],0≤t≤256,

[0030] x2(t)=cos[2πt(91-t / 6.2) / 256], 0≤t≤256,

[0031] x3(t)=cos(0.8πt), 50≤t≤70, 180≤t≤200.

[0032] Preferably, the synthesis and FIELD DATA examples, the synthetic signal x(t) is the sum of three components x1(t), x2(t) and x3(t), which are defined as;

[0033]

[0034] The present invention provides a resolution time-frequency analysis method based on adaptive ultra-wavelet set transform. It has the following beneficial effects:

[0035] The present invention improves both the time and frequency resolution of the time-frequency analysis results by combining a short, low-bandwidth wavelet with a longer, high-bandwidth wavelet. We only need to optimize the parameter C, because the ultrawavelet can be adaptively obtained by equation (9), and the value range of o is set to [1-30]. Therefore, the proposed method has higher resolution and is more powerful than CWT. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the time resolution of the present invention;

[0037] Figure 2 Schematic diagram of the frequency resolution of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Embodiment 1:

[0040] like Figure 1-2 As shown, the embodiment of the present invention provides a resolution time-frequency analysis method based on adaptive ultra-wavelet set transform, including adaptive ultra-wavelet set transform and continuous wavelet transform. The adaptive ultra-wavelet set transform continuous signal x(t) is defined as following formula (1);

[0041]

[0042]

[0043] CWTx (τ,s) is a function of the translation parameter τ and the scale parameter s, ψ*(t) is the complex conjugate of ψ(t), given by equation (2), equation (2) is also called the mother wavelet, the mother wavelet is stretched and squeezed by changing the parameter s, in addition, the change of s will not only change the center frequency of the mother wavelet, but also change the window length. Therefore, the parameter selection of the mother wavelet plays an important role in continuous wavelet transform and adaptive ultrawavelet set transform. The mother wavelet used is the Morlet wavelet, which is defined as formula (3);

[0044]

[0045] f c is the center frequency of the mother wavelet, f b is the bandwidth parameter and j is the imaginary unit. For the Morlet wavelet, the frequency f is given corresponding to the scale parameter s;

[0046]

[0047] The geometric mean of multiple spectrograms can be combined and calculated to increase the time and frequency resolution. Based on these mathematical foundations, the adaptive super wavelet transform (ASLT) combines short, low-bandwidth wavelets with longer, high-bandwidth wavelets to increase the resolution of time and frequency.

[0048] S1. First, we use the period number c of the wavelet instead of f b parameter;

[0049]

[0050] S2. Combining equation (3) and equation (5), we establish a set of Morlet wavelets ψ fc,c , with a fixed center frequency f c and different period numbers c, called ultrawavelets, which are defined as formula (6);

[0051]

[0052] where o is the order of the ultrawavelet, c1, c2, ..., c o is the number of cycles of each wavelet in the superwavelet;

[0053] S3. Secondly, using equation (7), we use wavelet transform to calculate the complex convolution of the input signal x(t), which is similar to the traditional CWT. The difference is that a set of wavelets is used in equation (7);

[0054]

[0055] Here * is the convolution operator;

[0056] S4. Finally, using equation (8), we obtain the response of a single wavelet in the ensemble;

[0057]

[0058] Here R[SL fc,o ] represents the result of super wavelet transform (SLT).

[0059] S3 also includes the following steps:

[0060] The adaptive superwavelet (ASL) adjusts its order to the center frequency to compensate for the decreasing bandwidth as the frequency increases. Here, the order of the superwavelet is given by:

[0061]

[0062] Among them, min and max are the minimum and maximum orders of the ultrawavelet, respectively. min and f max are the minimum and maximum center frequencies respectively, and [] is the nearest integer operator.

[0063] x1(t), x2(t), and x3(t), defined as:

[0064] x1(t)=cos[2πt(10+t / 6.2) / 256],0≤t≤256,

[0065] x2(t)=cos[2πt(91-t / 6.2) / 256], 0≤t≤256,

[0066] x3(t)=cos(0.8πt), 50≤t≤70, 180≤t≤200.

[0067] Synthesis and FIELD DATA Example,The synthetic signal x(t) is the sum of three components x1(t), x2(t) and x3(t), which are defined as;

[0068]

[0069] By combining a short, low-bandwidth wavelet with a longer, high-bandwidth wavelet to simultaneously improve the time and frequency resolution of the time-frequency analysis results, we only need to optimize the parameter c, because the ultrawavelet can be adaptively obtained by equation (9), and the value range of o is set to [1-30]. Therefore, the proposed method has higher resolution and is more powerful than CWT.

[0070] Embodiment 2:

[0071] The composite signal and its three components are as follows Figure 1The TFA diagrams obtained by CWT and ASLT are shown in Figure 2 As shown. Figure 2 From (a) and (b), we can clearly see that when f b is fixed, a small f c can improve the temporal resolution. Figure 2 It can be clearly seen from (b) and (c) that when f c is fixed, a larger f b The frequency resolution can be improved. Therefore, the TFA result of CWT is highly dependent on f c and f b The traditional adaptive ultra-wavelet transform is difficult to select the appropriate mother wavelet parameters. The time-frequency distribution diagram obtained by ASLT is as follows: Figure 2 As shown in (d), (e), and (f). Obviously, ASLT has higher resolution in both time and frequency. More importantly, in this method, we only need to optimize the parameter c, because the ultrawavelet can be adaptively obtained by formula (9). In this paper, the value range of o is set. Therefore, the proposed method has higher resolution and stronger robustness, and is more powerful than CWT. This paper introduces an adaptive ultrawavelet transform (ASLT) for time-frequency analysis. We verify the proposed method using synthetic signals and field data as examples and compare it with the results of continuous wavelet transform. The results of synthetic and one-dimensional field measured data show that ASLT can provide higher time and frequency resolution than continuous wavelets. In addition, two-dimensional field data examples show that ASLT can detect low-frequency shadows and highlight stratigraphic features for identifying seismic reservoirs.

[0072] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A resolution time-frequency analysis method based on adaptive ultra-wavelet set transform, including continuous wavelet transform and adaptive ultra-wavelet set transform, characterized in that: The adaptive ultra-wavelet set transform continuous signal x(t) is defined as following formula (1); Where CW Tx(τ,s) is a function of the translation parameter τ and the scale parameter s, ψ*(t) is the complex conjugate of ψ(t), given by (2). Equation (2) is also called the mother wavelet. The mother wavelet is stretched and squeezed by changing the parameter s. In addition, the change of s will not only change the center frequency of the mother wavelet, but also change the window length. Therefore, the parameter selection of the mother wavelet plays an important role in continuous wavelet transform and adaptive ultrawavelet set transform. The mother wavelet used is the Morlet wavelet, which is defined as formula (3); f c is the center frequency of the mother wavelet, f b is the bandwidth parameter and j is the imaginary unit. For the Morlet wavelet, the frequency f is given corresponding to the scale parameter s; S1. First, we use the period number c of the wavelet instead of f b parameter; S2. Combining equation (3) and equation (5), we establish a set of Morlet wavelets ψ fc,c , with a fixed center frequency f c and different period numbers c, called ultrawavelets, which are defined as formula (6); where o is the order of the ultrawavelet, c1, c2, ..., c o is the number of cycles of each wavelet in the superwavelet; S3. Secondly, using equation (7), we use wavelet transform to calculate the complex convolution of the input signal x(t), which is similar to the traditional CWT, except that a set of wavelets is used in equation (7); Here * is the convolution operator; S4. Finally, using equation (8), we obtain the response of a single wavelet in the ensemble; Here R[SL fc,o ] represents the result of super wavelet transform SLT, The S3 further comprises the following steps: The adaptive superwavelet set ASL adjusts its order to the center frequency to compensate for the decreasing bandwidth as the frequency increases. Here, the order of the superwavelet is given by: Among them, min and max are the minimum and maximum orders of the ultrawavelet, respectively, and f min and f max are the minimum and maximum center frequencies respectively, and [] is the nearest integer operator.

2. The resolution time-frequency analysis method based on adaptive ultrawavelet set transform according to claim 1 is characterized in that: The adaptive super wavelet transform (ASLT) can combine and calculate the geometric mean of multiple spectrograms to improve the time and frequency resolution. The adaptive super wavelet transform (ASLT) combines short low-bandwidth wavelets with longer high-bandwidth wavelets to improve the time and frequency resolution.

3. The resolution time-frequency analysis method based on adaptive ultrawavelet set transform according to claim 1 is characterized in that: x1(t), x2(t), and x3(t), defined as: x1(t)=cos[2πt(10+t / 6.2) / 256], 0≤t≤256, x2(t)=cos[2πt(91-t / 6.2) / 256], 0≤t≤256, x3(t)=cos(0.8πt), 50≤t≤70, 180≤t≤200.

Citation Information

Patent Citations

  • Variable-bandwidth OFDM (Orthogonal Frequency Division Multiplexing) transmitter and receiver

    CN103023850A

  • Method and device applied to laboratory seismic wave prediction

    CN111427079A

  • Space charge noise suppression method and device based on time-frequency analysis

    CN113705347A

  • Farm inspection robot

    CN210502923U

  • Methods and apparatus for cooperative noise attenuation in data sets related to the same underground formation

    US20150346369A1