A method for recovering weak reflected signals by eliminating noise interference

Abstract

The present invention relates to a method for recovering weak reflection signals that eliminates noise interference, comprising the following steps: obtaining seismic data to be denoised, performing Fourier transform on the seismic data along the time direction, and transforming the seismic data into the frequency space domain; using an autoregressive model to obtain a forward prediction operator and a backward prediction operator for each frequency slice to obtain a bilateral prediction operator; then constructing a prediction error operator, and obtaining the prediction error operator for each frequency to obtain the prediction error operator for the entire two-dimensional data volume; establishing an inversion target equation, inverting each frequency slice to obtain a valid signal, and performing an inverse Fourier transform on the signal to obtain the denoised seismic data. The present invention can better preserve the characteristic information of the signal while denoising, recover the weak reflection signal, and improve the signal-to-noise ratio of the seismic data.

Description

Technical Field

[0001] The present invention belongs to the technical field of seismic data processing, interpretation and inversion in oil and gas geophysical exploration and development, and in particular relates to a weak reflection signal recovery method for eliminating noise interference. Background Art

[0002] Seismic data with a high signal-to-noise ratio (SNR) is the prerequisite and foundation for high-resolution seismic processing, interpretation, parameter inversion, and attribute analysis. Therefore, improving the SNR of seismic data is a crucial step in seismic data processing. As exploration targets become increasingly complex, denoising is becoming increasingly demanding in order to fully utilize the useful information in seismic data and minimize the impact on valid signals, especially weak reflections, during processing. Therefore, to meet the requirements of high-resolution seismic exploration and quantitative interpretation of seismic data, noise removal must be performed without compromising valid signals, thereby improving data processing quality.

[0003] Currently, the main noise reduction techniques used in practical data processing include predictive filtering, fk filtering, wavelet transform, and median filtering. Predictive filtering is the most commonly used random noise suppression method in industry, leveraging the spatial predictability of the effective signal. Predictive filtering methods first estimate a predictive filter operator in the tx or fx domain, then filter the original seismic record using this operator to obtain a noise-attenuated seismic record. However, these methods employ two contradictory basic assumptions regarding noise in their implementation. First, they assume that the seismic record consists of the sum of the effective signal and noise, known as the additive noise model. However, during predictive filtering, they also assume that the random noise is derived from the convolution of the prediction error operator with the seismic data, known as the source noise model. This inconsistent assumption reduces the theoretical rigor of the method, weakening its denoising and amplitude preservation capabilities. Specifically, while suppressing noise, it also loses the effective signal, resulting in low fidelity of the denoised profile and the potential for structural artifacts.

[0004] Therefore, based on these problems, it is of great practical significance to provide a weak reflection signal recovery method that can better retain the characteristic information of the signal while denoising, restore the weak reflection signal, and improve the signal-to-noise ratio of seismic data to eliminate noise interference. Summary of the Invention

[0005] The invention proposes a weak reflection signal recovery method for eliminating noise interference. By taking a prediction filter operator as the reflection feature of an effective signal and introducing it into an inversion system, the seismic signal is directly inverted from the seismic record, thus avoiding the inconsistency of the noise model.

[0006] The present invention solves the technical problem by adopting the following technical solutions:

[0007] A method for recovering weak reflected signals by eliminating noise interference comprises the following steps:

[0008] Obtain the seismic data to be denoised s(t,x), perform Fourier transform on the seismic data along the time direction, and transform the seismic data into the frequency space domain S(f,x);

[0009] The autoregressive model is used to find the forward prediction operator and the backward prediction operator for each frequency slice to obtain the bilateral prediction operator A, and then the prediction error operator h is constructed. f , for each frequency, the prediction error operator is obtained, and the prediction error operator of the entire two-dimensional data volume can be obtained;

[0010]

[0011]

[0012] Among them, m is the length of the unilateral prediction operator, p m is the coefficient of the prediction operator length m position;

[0013] Establish the inversion target equation C:

[0014]

[0015] Among them, S(f i ,x) is the frequency of the noisy data f i The frequency component of The denoised data frequency is f i The frequency component of is the convolution matrix of the prediction error operator; λ is the regularization parameter;

[0016] Since the objective function is a quadratic function, Taking the derivative and setting it to zero, we get:

[0017]

[0018] Where I is the identity matrix of dimension N,

[0019] Invert each frequency slice to obtain the effective signal right Perform inverse Fourier transform to obtain denoised seismic data

[0020] Furthermore, the solution of formula D adopts the conjugate gradient method.

[0021] The advantages and positive effects of the present invention are:

[0022] The present invention estimates a prediction error operator based on the predictability of linear events in the spatial direction of the fx domain; uses the prediction error operator as a reflection structure constraint for the effective signal, establishes an inversion objective function, and directly inverts the effective signal from the original seismic record; the present invention removes noise based on the idea of ​​inversion, alleviating the problem that traditional filtering-based denoising methods damage the effective signal, and can maintain the integrity of the effective wave during the processing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that these drawings are designed for illustrative purposes only and are not intended to limit the scope of the present invention. Furthermore, unless otherwise specified, these drawings are intended only to conceptually illustrate the structures described herein and are not necessarily drawn to scale.

[0024] FIG1( a ) is a noise-free synthetic seismic record provided by an embodiment of the present invention;

[0025] FIG1( b ) is a synthetic seismic record containing noise provided by an embodiment of the present invention;

[0026] FIG1( c ) is a seismic record after noise removal using conventional fx deconvolution according to an embodiment of the present invention;

[0027] FIG1( d ) is a seismic record after noise removal according to the method of the present invention provided in an embodiment of the present invention;

[0028] FIG2( a ) shows the noise removed by conventional fx deconvolution according to an embodiment of the present invention;

[0029] FIG2( b ) shows the noise removed by the method of the present invention provided in an embodiment of the present invention;

[0030] FIG3( a ) is actual seismic data provided by an embodiment of the present invention;

[0031] FIG3( b ) shows the seismic data after noise removal using the method of the present invention according to an embodiment of the present invention; DETAILED DESCRIPTION

[0032] First of all, it should be noted that the specific structure, characteristics and advantages of the present invention will be specifically described below in an exemplary manner. However, all descriptions are only used for illustration and should not be understood as limiting the present invention. In addition, any single technical feature described or implied in the embodiments mentioned herein, or any single technical feature displayed or implied in the drawings, can still be combined or deleted between these technical features (or their equivalents) to obtain more other embodiments of the present invention that may not be directly mentioned herein. In addition, in order to simplify the drawings, the same or similar technical features may be marked in only one place in the same drawing.

[0033] In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, or are the orientation or position relationship in which the product of the invention is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they should not be understood as limiting the present invention.

[0034] It should be noted that, unless there is any conflict, the embodiments and features in the embodiments of this application can be combined with each other.

[0035] This embodiment provides a method for recovering weak reflected signals by eliminating noise interference, including the following steps:

[0036] Obtain the seismic data to be denoised s(t,x), perform Fourier transform on the seismic data along the time direction, and transform the seismic data into the frequency space domain S(f,x);

[0037] The autoregressive model is used to find the forward prediction operator and the backward prediction operator for each frequency slice to obtain the bilateral prediction operator A, and then the prediction error operator h is constructed. f , for each frequency, the prediction error operator is obtained, and the prediction error operator of the entire two-dimensional data volume can be obtained;

[0038]

[0039]

[0040] Among them, m is the length of the unilateral prediction operator, p m is the coefficient of the prediction operator length m position;

[0041] Establish the inversion target equation C:

[0042]

[0043] Among them, here is the error term, which represents the error between the noisy data and the denoised data, S(f i ,x) is the frequency of the noisy data f i The frequency component of The denoised data frequency is f i The frequency component of is the spatial constraint term, is the convolution matrix of the prediction error operator; λ is the regularization parameter, which adjusts the weight of the constraint term in the objective function.

[0044] Since the objective function is a quadratic function, Taking the derivative and setting it to zero, we get:

[0045]

[0046] Where I is the identity matrix of dimension N, and the solution of formula D can be solved by the conjugate gradient method;

[0047] Invert each frequency slice to obtain the effective signal right Perform inverse Fourier transform to obtain denoised seismic data

[0048] As an example, in this embodiment, the present method is described in detail using seismic data from a block in the eastern oil field. The original seismic profile is shown in FIG3(a), and its time sampling rate is 2 ms. It can be seen that the signal-to-noise ratio of the data is low and it contains a lot of noise. The method of this embodiment is used to process it, wherein FIG1(a) is a synthetic seismic record without noise; FIG1(b) is a synthetic seismic record with noise. After processing, FIG1(c) and FIG1(d) are the results of denoising the model in FIG1(b) using the conventional fx deconvolution method and the method of this embodiment, respectively. FIG2(a) and FIG2(b) are the noise obtained after processing using the conventional fx deconvolution method and the method of this embodiment, respectively. By comparison, it can be seen that since the method of this embodiment uses the prediction error factor as the spatial constraint of the effective signal and directly inverts the effective signal from the data, the effective signal is retained to the maximum extent while removing the noise, and the weak reflection signal is well restored; while the conventional fx deconvolution method damages the effective signal in the original record, making the continuity of the weak signal poor; as shown in Figure 3(b), after being processed by the method of this embodiment, the noise in the original seismic data is more thoroughly eliminated, the effective signal is highlighted, and the reflection wave phase axis becomes clear and continuous. In addition, this method does not damage the effective signal while eliminating the noise, and well preserves the weak reflection phase axis in the original seismic record.

[0049] The above embodiments describe the present invention in detail, but the contents described are only preferred embodiments of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for recovering weak reflected signals by eliminating noise interference, characterized in that: The steps include: Obtain the seismic data to be denoised s(t,x), perform Fourier transform on the seismic data along the time direction, and transform the seismic data into the frequency space domain S(f,x); The autoregressive model is used to find the forward prediction operator and the backward prediction operator for each frequency slice to obtain the bilateral prediction operator A, and then the prediction error operator h is constructed. f , for each frequency, the prediction error operator is obtained, and the prediction error operator of the entire two-dimensional data volume can be obtained; Among them, m is the length of the unilateral prediction operator, p m is the coefficient of the prediction operator length m position; Establish the inversion target equation C: Among them, S(f i ,x) is the frequency of the noisy data f i The frequency component of The denoised data frequency is f i The frequency component of is the convolution matrix of the prediction error operator; λ is the regularization parameter; Since the objective function is a quadratic function, Taking the derivative and setting it to zero, we get: Where I is the identity matrix of dimension N, Invert each frequency slice to obtain the effective signal right Perform inverse Fourier transform to obtain denoised seismic data 2. The method for recovering weak reflected signals by eliminating noise interference according to claim 1, wherein: The solution of Equation D is the conjugate gradient method.

Images