CN103135133A - Method and device of vector noise reduction of multi-component seismic data - Google Patents

Abstract

The invention provides a method and a device of vector noise reduction of multi-component seismic data. The method comprises the following steps: collecting the multi-component seismic data; carrying out wavelet decomposition on the multi-component seismic data to obtain multi-scale seismic signals; confirming a vector module value and an angle cosine which correspond to each scale seismic signal; filtering stochastic noise and sharp pulses in the seismic signals according to a vector threshold value method; carrying out surface wave suppress and effective wave compensation on the seismic signals which filter the stochastic noise and the sharp pulses; carrying out wavelet reconstruction according to the seismic signals after the surface wave suppress and the effective wave compensation and the angle cosine to obtain the multi-component seismic data after a treatment of the vector noise reduction. Due to the fact that the wavelet decomposition is carried out on the multi-component seismic data, recognition accuracy of the effective signals and noise in simple-component data processing can be improved, and an effect that the surface wave is removed and the effective wave is compensated is achieved accurately and completely.

Description

A kind of vector noise-reduction method and equipment of multi-component seismic data

Technical field

The present invention about the geophysical exploration technology field, particularly about the treatment technology of seismic data, is a kind of vector noise-reduction method and equipment of multi-component seismic data concretely.

Background technology

Carry a large amount of formation informations in seismic data, can be used to study rock property, the fluid properties on stratum, and then understand the subsurface geology situation.But, in seismic data except the finite bandwidth of seismic wavelet, due to the impact of correlation noise (comprising ground roll, multiple reflection and power frequency interference etc.) and random noise (neighbourhood noise, measuring error and the earth ground unrest etc.), be difficult to clearly reflect the information of stratum condition.Therefore, in the processing procedure of seismic data, how effectively eliminating various noises is one and crucial problem.

The random noise that is aliasing in useful signal is one of factor that affects the seismic data processing accuracy, and there is the contradiction between guard signal local feature and inhibition noise in traditional methods such as linear filtering that in prior art, the filtering random signal is used.From mathematics, above-mentioned denoising method is much all based on Fourier transform, and traditional Fourier transform no matter how the time changes, all the Partial Transformation that frequency is identical is to same position, so Fourier transform only has Statistical Effect.The wavelet transformation of being used widely at present can successively become signal decomposition high fdrequency component and low frequency component, and because useful signal and noise have visibly different propagation characteristic after multiple dimensioned decomposition, different qualities for both have carries out noise reduction process in conjunction with methods such as threshold methods to original signal and can access good effect.Yet the deficiency that this kind method exists is, although wavelet transformation can provide, the time-frequency division of signal is processed, and constructing or choose corresponding Optimum wavelet base for different seismic signals is more complicated and difficulty.

In addition, in the real work that seismic data is processed, lie in the ground roll in the low frequency seismic components, be reduction seismic section resolution, cause the lineups skew also finally to affect the critical elements of seismic data interpretation quality.Therefore, the ground roll compacting is the important topic during seismic data is processed always.

Ground roll is as the secondary wave that a quefrency is lower, energy is stronger of bulk wave constructive interference under certain condition and stack generation, that a kind of rule that in the prestack record, energy is very strong is disturbed, its wave field generally is fan-shaped distribution at shortcut in record, the characteristics that energy is strong, frequency is low, apparent velocity is low are arranged.Be embodied in following space-time characteristic:

1. ground roll is mainly propagated along the interphase of medium, and its energy is decay rapidly with the increase of the degree of depth, but at dielectric surface, the decay of ground roll is very slow, and propagation distance is very far away.

2. the frequency of ground roll with the increase relative ground of the degree of depth and propagation distance by high frequency to low frequency transform, can observe obvious surface wave dispersion phenomenon from seismic section.

3. ground roll is the solution of a kind of new model of wave equation, and its velocity of propagation is determined by the parameter of medium fully, and is all lower than compressional wave, shear wave.Again because ground roll is propagated along the earth's surface, therefore its lineups show as one group of near linear on seismic section, with the hyperbolic curve of reflection wave, notable difference are arranged, and because apparent velocity is slow, can intersect with the hyperbolic curve of reflection wave, and serious aliasing occurs in the subregion.

4. the regular variation with the degree of depth and propagation distance of the apparent velocity of ground roll changes, and is reflected in seismologic record, and its lineups are not to be the cluster straight line as thinking in the past, but the cluster curve.

Above feature due to ground roll, in the ground roll pressing process, the traditional conventional surface wave pressing method that adopts has high-pass filtering method, Karhunen-Loeve converter technique and τ-p converter technique, but still there are following technological difficulties: the denoising instrument that (1) is conventional, typically as threshold method, low-pass filtering, all be based on the compacting Design of Problems of high frequency details noise, be not suitable for the ground roll compacting problem of lower frequency region; (2) because the low frequency component of ground roll and significant wave is seriously obscured, adopt conventional surface wave pressing method, the as easy as rolling off a log loss that causes useful signal; (3) from mathematics, the low frequency component of ground roll and significant wave has certain correlativity, may show as same frequency, same energy level under local conditions, therefore, in the ground roll pressing process, is difficult to ground roll is accurately separated with the low frequency component of significant wave.

In addition, multi-component seismic data is more complicated with respect to the simple component data, and seismic event type and energy included in different components are strong and weak different, face more challenges in the process of noise compacting.Therefore, the vector noise reduction technology of multi-component seismic data is the technical barrier that this area needs to be resolved hurrily.

Summary of the invention

the problems referred to above that exist in order to overcome prior art, the invention provides a kind of vector noise-reduction method and equipment of multi-component seismic data, by multi-component seismic data is carried out wavelet decomposition, filtering random noise and spike pulse, Surface Wave Elimination, significant wave is compensated, step of going forward side by side is carried out wavelet reconstruction, adopt the vector noise reduction of seismic wave field, the complementation of temporal signatures and frequency domain character between each seismic components, can improve the simple component data process in to the degree of accuracy of useful signal and noise, reach more accurate, more completely remove ground roll and the effect that compensates significant wave.

One of purpose of the present invention is, a kind of vector noise-reduction method of multi-component seismic data is provided, and comprising: gather multi-component seismic data; Described multi-component seismic data is carried out wavelet decomposition, obtain the seismic signal of a plurality of yardsticks; Determine corresponding Vector Mode value and the angle cosine of seismic signal of each yardstick; According to random noise and the spike pulse in the described seismic signal of Vector Threshold method filtering; Seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation; Carry out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtain the multi-component seismic data after the vector noise reduction process.

One of purpose of the present invention is, a kind of vector noise reducing apparatus of multi-component seismic data is provided, and comprising: harvester is used for gathering multi-component seismic data; The wavelet decomposition device is used for described multi-component seismic data is carried out wavelet decomposition, obtains the seismic signal of a plurality of yardsticks; The Vector Mode value is determined device, is used for determining the Vector Mode value corresponding to seismic signal of each yardstick; Angle cosine is determined device, is used for determining the angle cosine corresponding to seismic signal of each yardstick; Filtering device is used for random noise and spike pulse according to the described seismic signal of Vector Threshold method filtering; The compacting compensation system is used for the seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation; The wavelet reconstruction device is used for carrying out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtains the multi-component seismic data after the vector noise reduction process.

beneficial effect of the present invention is, by multi-component seismic data is carried out wavelet decomposition, filtering random noise and spike pulse, Surface Wave Elimination, significant wave is compensated, step of going forward side by side is carried out wavelet reconstruction, obtain the seismic data after the vector noise reduction process, adopt the vector noise reduction of seismic wave field, can realize the complementation of temporal signatures and frequency domain character between each seismic components, and then improve the degree of accuracy that the simple component data is distinguished useful signal and noise in processing, finally reach more accurate, more completely remove ground roll and the effect that compensates significant wave, further improved the accuracy of reservoir prediction.

For above and other purpose of the present invention, feature and advantage can be become apparent, preferred embodiment cited below particularly, and coordinate appended graphicly, be described in detail below.

Description of drawings

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, the below will do to introduce simply to the accompanying drawing of required use in embodiment or description of the Prior Art, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.

The process flow diagram of the embodiment one of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 1 provides for the embodiment of the present invention;

The process flow diagram of the embodiment two of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 2 provides for the embodiment of the present invention;

The process flow diagram of the embodiment three of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 3 provides for the embodiment of the present invention;

The structured flowchart of the vector noise reducing apparatus of a kind of multi-component seismic data that Fig. 4 provides for the embodiment of the present invention;

Fig. 5 is the schematic diagram of single track three-component record in embodiment 1 provided by the invention;

Fig. 6 is the seismologic record schematic diagram of mould codomain and angle domain in embodiment 1 provided by the invention;

Fig. 7 is the noise reduction result schematic diagram of single track seismologic record in embodiment 1 provided by the invention;

Fig. 8 is the schematic diagram of pending seismic section x component in embodiment 2 provided by the invention;

Fig. 9 is the schematic diagram of pending seismic section y component in embodiment 2 provided by the invention;

Figure 10 is the schematic diagram of pending seismic section z component in embodiment 2 provided by the invention;

Figure 11 is the schematic diagram of seismic section x component after processing in embodiment 2 provided by the invention;

Figure 12 is the schematic diagram of seismic section y component after processing in embodiment 2 provided by the invention;

Figure 13 is the schematic diagram of seismic section z component after processing in embodiment 2 provided by the invention;

Figure 14 processes the schematic diagram of forward and backward z section local detail in embodiment 2 provided by the invention;

Figure 15 is the schematic diagram of original seismic section z component in embodiment 3 provided by the invention;

Figure 16 is the schematic diagram of simple component noise reduction result in embodiment 3 provided by the invention;

Figure 17 is the schematic diagram of vector noise reduction result in embodiment 3 provided by the invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.

The invention provides a kind of vector noise-reduction method and equipment of multi-component seismic data, by multi-component seismic data is carried out wavelet decomposition, filtering random noise and spike pulse, Surface Wave Elimination, significant wave is compensated, step of going forward side by side is carried out wavelet reconstruction, adopt the vector noise reduction of seismic wave field, the complementation of temporal signatures and frequency domain character between each seismic components, can improve the simple component data process in to the degree of accuracy of useful signal and noise, reach more accurate, more completely remove ground roll and the effect that compensates significant wave.

The vector noise-reduction method of a kind of multi-component seismic data provided by the invention specifically comprises:

S1: gather multi-component seismic data.Multi-component seismic data is more complicated with respect to the simple component data, and seismic event type and energy included in different components are strong and weak different, face more challenges in the process of noise compacting.Therefore, method provided by the invention can be applicable to the vector noise reduction process of multi-component seismic data.

S2: described multi-component seismic data is carried out wavelet decomposition, obtain the seismic signal of a plurality of yardsticks;

S3: corresponding Vector Mode value and the angle cosine of seismic signal of determining each yardstick.In concrete embodiment, can seismic event type contained according to different components in multi-component seismic data and the difference of energy intensity, utilize the mutual relationship between a plurality of components, calculate each Vector Mode value and angle cosine of seismologic record constantly.

Wherein, the Vector Mode value that the seismic signal of described each yardstick is corresponding can obtain by following formula:

$p\left(t\right)=\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, and p (t) is the Vector Mode value of multi-component seismic data.

The angle cosine that the seismic signal of described each yardstick is corresponding obtains by following formula:

$a_x\left(t\right)=\frac{x\left(t\right)}{\sqrt{{x\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_y\left(t\right)=\frac{y\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_z\left(t\right)=\frac{z\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, a _x(t), a _y(t), a _z(t) be the angle cosine of three different components.

S4: according to random noise and the spike pulse in the described seismic signal of Vector Threshold method filtering.For the vector characteristic of multi-component seismic data, consider to be positioned near the kinematics continuity of the particle of seismic receiver, can promote on the basis of wavelet thresholding method, obtain the wavelet thresholding method of vector wave field.

S5: the seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation.In concrete embodiment, can carry out the time-frequency domain explication de texte of vector earthquake, record the z component with advantage and be as the criterion, determine time-frequency piece to be analyzed, the geological data of choosing local carries out ground roll compacting and significant wave compensation.In this step, adopt the space zero-crossing filter based on many component recordings to carry out, the space zero-crossing filter is expressed as follows:

$p\left(t\right)=\left\{\begin{array}{c}\frac{p\left(t\right)}{S}\mathrm{ift}\&Element;(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2),(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2)\&SubsetEqual;(L_x,L_y,L_z)\mathrm{and}{\mathrm{<figure-callout\; id="1"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_1<\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}<E_2\\ p\left(t\right)\mathrm{ift}\&Element;(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2),(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2)\&NotSubset;(L_x,L_y,L_z)\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_2\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&le;{\mathrm{<figure-callout\; id="1"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_1\end{array}\right.$

Wherein, t1, t2 are the adjacent zero crossing of either direction, f (t1)=0, and f (t2)=0, when t ∈ (t1, t2) | and f (t) |〉0, S is decay multiplying power, effective wave attenuation multiplying power S＜1, the decay multiplying power S of ground roll〉1, L _x, L _y, L _zBe the time period threshold of vector space, E ₁, E ₂Energy threshold for vector space.

Be also in step S5, the space zero-crossing filter can make up for each other's deficiencies and learn from each other based on the zero crossing judgement of three directions.For the judgement of ground roll, take the criterion of " it is not excessive to be would rather be scarce ", the record on any direction satisfies the zero crossing judgement, this segment data is taked to suppress algorithm.For the compensation problem of significant wave, the record on any direction satisfies its zero crossing condition, this segment data is taked backoff algorithm.

S6: carry out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtain the multi-component seismic data after the vector noise reduction process.

Method provided by the invention adopts the vector noise reduction of seismic wave field, the complementation of temporal signatures and frequency domain character between each seismic components, can improve the simple component data process in to the degree of accuracy of useful signal and noise, the effect of reach more accurate, more completely removing ground roll and compensation significant wave, improve the signal to noise ratio (S/N ratio) of multi-component seismic data, further improved the accuracy of reservoir prediction.

The process flow diagram of the embodiment one of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 1 provides for the embodiment of the present invention, as shown in Figure 1, in this embodiment, step S4 specifically comprises:

S104: according to random noise and the spike pulse in the described seismic signal of vector soft-threshold method filtering of wavelet analysis.In embodiment one, the vector soft-threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ _iBe the Vector Mode value threshold of i yardstick,

Vector soft-threshold method for wavelet analysis.

The process flow diagram of the embodiment two of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 2 provides for the embodiment of the present invention, as shown in Figure 2, in embodiment two, step S4 specifically comprises:

S204: according to random noise and the spike pulse in the described seismic signal of vector hard threshold method filtering of wavelet analysis.In this embodiment, the vector hard threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain,

Vector hard threshold method for wavelet analysis.

The process flow diagram of the embodiment three of the vector noise-reduction method of a kind of multi-component seismic data that Fig. 3 provides for the embodiment of the present invention, as shown in Figure 3, in this embodiment three, step S4 specifically comprises:

S3041: according to random noise and the spike pulse in the described seismic signal of vector hard threshold method filtering of wavelet analysis.The vector hard threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

S3042: according to random noise and the spike pulse in the described seismic signal of vector soft-threshold method filtering of wavelet analysis.

The vector soft-threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Also namely in embodiment three, both adopted the vector hard threshold method of wavelet analysis, adopt again the vector soft-threshold method of wavelet analysis, vector hard threshold method and vector soft-threshold method there is no successively smoothly, also can first adopt vector soft-threshold method, adopt again the vector hard threshold method, make multi component seismic records complementary not enough, can make up may occur in simple component threshold value noise reduction because the factors such as phase place variation, machine error cause thinking useful signal by mistake data degradation that noise causes.Process random noise and the spike pulse of many component recordings of filtering high-frequency domain by the high-frequency signal detail section after wavelet transformation being carried out threshold.

The structured flowchart of the vector noise reducing apparatus of a kind of multi-component seismic data that Fig. 4 provides for the embodiment of the present invention, as shown in Figure 4, this equipment specifically comprises:

Harvester 100 is used for gathering multi-component seismic data.Multi-component seismic data is more complicated with respect to the simple component data, and seismic event type and energy included in different components are strong and weak different, face more challenges in the process of noise compacting.Therefore, method provided by the invention can be applicable to the vector noise reduction process of multi-component seismic data.

Wavelet decomposition device 200 is used for described multi-component seismic data is carried out wavelet decomposition, obtains the seismic signal of a plurality of yardsticks.

The Vector Mode value is determined device 300, is used for determining the Vector Mode value corresponding to seismic signal of each yardstick.In concrete embodiment, can seismic event type contained according to different components in multi-component seismic data and the difference of energy intensity, utilize the mutual relationship between a plurality of components, calculate each Vector Mode value of seismologic record constantly.The Vector Mode value that the seismic signal of described each yardstick is corresponding can obtain by following formula:

$p\left(t\right)=\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, and p (t) is the Vector Mode value of multi-component seismic data.

Angle cosine is determined device 400, is used for determining the angle cosine corresponding to seismic signal of each yardstick.In concrete embodiment, can seismic event type contained according to different components in multi-component seismic data and the difference of energy intensity, utilize the mutual relationship between a plurality of components, calculate each angle cosine of seismologic record constantly.The angle cosine that the seismic signal of described each yardstick is corresponding obtains by following formula:

$a_x\left(t\right)=\frac{x\left(t\right)}{\sqrt{{x\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_y\left(t\right)=\frac{y\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_z\left(t\right)=\frac{z\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, a _x(t), a _y(t), a _z(t) be the angle cosine of three different components.

Filtering device 500 is used for random noise and spike pulse according to the described seismic signal of Vector Threshold method filtering.For the vector characteristic of multi-component seismic data, consider to be positioned near the kinematics continuity of the particle of seismic receiver, can promote on the basis of wavelet thresholding method, obtain the wavelet thresholding method of vector wave field.

In equipment provided by the invention, filtering device 500 has following three kinds of embodiments:

1, in embodiment one, filtering device 500 is according to random noise and spike pulse in the described seismic signal of vector soft-threshold method filtering of wavelet analysis.In this embodiment, the vector soft-threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ _iBe the Vector Mode value threshold of i yardstick,

Vector soft-threshold method for wavelet analysis.

2, in embodiment two, filtering device 500 is according to random noise and spike pulse in the described seismic signal of vector hard threshold method filtering of wavelet analysis.In this embodiment, the vector hard threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain, Vector hard threshold method for wavelet analysis.

3, in embodiment three, filtering device 500 is at first according to random noise and spike pulse in the described seismic signal of vector hard threshold method filtering of wavelet analysis.The vector hard threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

Last according to random noise and spike pulse in the described seismic signal of vector soft-threshold method filtering of wavelet analysis.The vector soft-threshold method of wavelet analysis is undertaken by following formula:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Also namely in embodiment three, both adopted the vector hard threshold method of wavelet analysis, adopt again the vector soft-threshold method of wavelet analysis, vector hard threshold method and vector soft-threshold method there is no successively smoothly, also can first adopt vector soft-threshold method, adopt again the vector hard threshold method, make multi component seismic records complementary not enough, can make up may occur in simple component threshold value noise reduction because the factors such as phase place variation, machine error cause thinking useful signal by mistake data degradation that noise causes.Process random noise and the spike pulse of many component recordings of filtering high-frequency domain by the high-frequency signal detail section after wavelet transformation being carried out threshold.

Compacting compensation system 600 is used for the seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation.In concrete embodiment, can carry out the time-frequency domain explication de texte of vector earthquake, record the z component with advantage and be as the criterion, determine time-frequency piece to be analyzed, the geological data of choosing local carries out ground roll compacting and significant wave compensation.In this step, adopt the space zero-crossing filter based on many component recordings to carry out, the space zero-crossing filter is expressed as follows:

$p\left(t\right)=\left\{\begin{array}{c}\frac{p\left(t\right)}{S}\mathrm{ift}\&Element;(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2),(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2)\&SubsetEqual;(L_x,L_y,L_z)\mathrm{and}{\mathrm{<figure-callout\; id="1"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_1<\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}<E_2\\ p\left(t\right)\mathrm{ift}\&Element;(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2),(\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2)\&NotSubset;(L_x,L_y,L_z)\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_2\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&le;{\mathrm{<figure-callout\; id="1"\; label="E"\; filenames="HDA00002777750500051.png,HDA00002777750500052.png"\; state="\{\{state\}\}">E</figure-callout>}}_1\end{array}\right.$

Wherein, t1, t2 are the adjacent zero crossing of either direction, f (t1)=0, and f (t2)=0, when t ∈ (t1, t2) | and f (t) |〉0, S is decay multiplying power, effective wave attenuation multiplying power S＜1, the decay multiplying power S of ground roll〉1, L _x, L _y, L _zBe the time period threshold of vector space, E ₁, E ₂Energy threshold for vector space.

Also namely in compacting compensation system 600, the space zero-crossing filter can make up for each other's deficiencies and learn from each other based on the judgement of the zero crossing of three directions.For the judgement of ground roll, take the criterion of " it is not excessive to be would rather be scarce ", the record on any direction satisfies the zero crossing judgement, this segment data is taked to suppress algorithm.For the compensation problem of significant wave, the record on any direction satisfies its zero crossing condition, this segment data is taked backoff algorithm.

Wavelet reconstruction device 700 is used for carrying out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtains the multi-component seismic data after the vector noise reduction process.

Equipment provided by the invention adopts the vector noise reduction of seismic wave field, the complementation of temporal signatures and frequency domain character between each seismic components, can improve the simple component data process in to the degree of accuracy of useful signal and noise, the effect of reach more accurate, more completely removing ground roll and compensation significant wave, improve the signal to noise ratio (S/N ratio) of multi-component seismic data, further improved the accuracy of reservoir prediction.

Below in conjunction with specific embodiment, introduce in detail vector noise-reduction method and the equipment of a kind of multi-component seismic data provided by the invention.

Embodiment 1: the three-component record vector noise reduction process of single track seismic event

Vector noise-reduction method and the equipment and technology of using a kind of multi-component seismic data provided by the invention carry out the vector noise reduction process to three-component single track seismologic record, Fig. 5 is initial untreated three-component one-channel record, with its seismologic record that changes to mould codomain and angle domain as shown in Figure 6, vector noise reduction flow process (as Fig. 1) through multi-component seismic data, Attenuating Random Noise and ground roll, and after significant wave was compensated, the one-channel record after being processed as shown in Figure 7.

Embodiment 2: the prestack 3-component earthquake records the vector noise reduction process

Vector noise-reduction method and the equipment of using a kind of multi-component seismic data provided by the invention carry out the vector noise reduction process to three-component pre-stack seismic section.Fig. 8, Fig. 9, Figure 10 are the initial noisy three-component record that contains, and through the vector noise reduction process (as Fig. 1) of multi-component seismic data, the 3-component earthquake section after being processed is as Figure 11, Figure 12, shown in Figure 13.More as can be known, all can be observed the low frequency ground roll from the three-component record section and substantially obtained effective compacting, the ground roll clearance is more than 95%.And, through the compensation deals of useful signal, by can be observed obvious reflection line-ups on Figure 13 vertical section.And, as shown in Figure 14, even for the situation of ground roll and the serious aliasing of significant wave, adopt the present invention not only to remove preferably ground roll, also make significant wave obtain enough compensation, reconstruct a series of reflection line-upss on the section after rejecting ground roll.

Embodiment 3: vector Method of Noise and the contrast of simple component Method of Noise treatment effect

To same z component seismic section (as shown in figure 15), adopt respectively vector Method of Noise and simple component Method of Noise to process, can get Different Results (as Figure 16 and shown in Figure 17).Contrast is found, although the ground roll of two kinds of methods removal ratio is close, adopts the vector Method of Noise can better compensate the useful signal loss, improves the resolution of seismic data.This be due to, adopt the vector noise reduction of seismic wave field, can utilize the complementation of temporal signatures and frequency domain character between each seismic components, improve the degree of accuracy that the simple component data is distinguished useful signal and noise in processing, reach more accurate, more completely remove ground roll and the effect that compensates significant wave.

in sum, useful achievement of the present invention is: vector noise-reduction method and equipment that a kind of multi-component seismic data is provided, by multi-component seismic data is carried out wavelet decomposition, filtering random noise and spike pulse, Surface Wave Elimination, significant wave is compensated, step of going forward side by side is carried out wavelet reconstruction, obtain the seismic data after the vector noise reduction process, adopt the vector noise reduction of seismic wave field, can realize the complementation of temporal signatures and frequency domain character between each seismic components, and then improve the degree of accuracy that the simple component data is distinguished useful signal and noise in processing, finally reach more accurate, more completely remove ground roll and the effect that compensates significant wave, further improved the accuracy of reservoir prediction.

One of ordinary skill in the art will appreciate that all or part of flow process that realizes in above-described embodiment method, can come the relevant hardware of instruction to complete by computer program, described program can be stored in general computer read/write memory medium, this program can comprise the flow process as the embodiment of above-mentioned each side method when carrying out.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.

Those skilled in the art can also recognize that the various functions that the embodiment of the present invention is listed are to realize depending on the designing requirement of specific application and whole system by hardware or software.Those skilled in the art can be for every kind of specific application, and can make ins all sorts of ways realizes described function, but this realization should not be understood to exceed the scope of embodiment of the present invention protection.

Used specific embodiment in the present invention principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.

Claims (14)

1. the vector noise-reduction method of a multi-component seismic data, is characterized in that, described method comprises:

Gather multi-component seismic data;

Described multi-component seismic data is carried out wavelet decomposition, obtain the seismic signal of a plurality of yardsticks;

Determine corresponding Vector Mode value and the angle cosine of seismic signal of each yardstick;

According to random noise and the spike pulse in the described seismic signal of Vector Threshold method filtering;

Seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation;

Carry out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtain the multi-component seismic data after the vector noise reduction process.

2. method according to claim 1, is characterized in that, the Vector Mode value that the seismic signal of described each yardstick is corresponding obtains by following formula:

$p\left(t\right)=\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, and p (t) is the Vector Mode value of multi-component seismic data.

3. method according to claim 1, is characterized in that, the angle cosine that the seismic signal of described each yardstick is corresponding obtains by following formula:

$a_x\left(t\right)=\frac{x\left(t\right)}{\sqrt{{x\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_y\left(t\right)=\frac{y\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_z\left(t\right)=\frac{z\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, a _x(t), a _y(t), a _z(t) be the angle cosine of three different components.

4. method according to claim 2, is characterized in that, undertaken by following formula according to random noise and spike pulse in the described seismic signal of Vector Threshold method filtering:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ _iBe the Vector Mode value threshold of i yardstick,

Vector soft-threshold method for wavelet analysis.

5. method according to claim 2, is characterized in that, undertaken by following formula according to random noise and spike pulse in the described seismic signal of Vector Threshold method filtering:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain,

Vector hard threshold method for wavelet analysis.

6. method according to claim 2, is characterized in that, undertaken by following formula according to random noise and spike pulse in the described seismic signal of Vector Threshold method filtering:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain, λ _iBe the Vector Mode value threshold of i yardstick,

The vector hard threshold method of wavelet analysis,

Vector soft-threshold method for wavelet analysis.

7. method according to claim 1, it is characterized in that, to the seismic signal after filtering random noise and spike pulse carry out ground roll compacting, the significant wave compensation is undertaken by the space zero-crossing filter, describedly be expressed as follows by the space zero-crossing filter:

$p\left(t\right)=\left\{\begin{array}{c}\frac{p\left(t\right)}{S}\mathrm{ift}\&Element;(t1,t2),(t1,t2)\&SubsetEqual;(L_x,L_y,L_z)\mathrm{and}{E}_1<\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}<E_2\\ p\left(t\right)\mathrm{ift}\&Element;(t1,t2),(t1,t2)\&NotSubset;(L_x,L_y,L_z)\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&GreaterEqual;E_2\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&le;E_1\end{array}\right.$

Wherein, t1, t2 are the adjacent zero crossing of either direction, f (t1)=0, and f (t2)=0, when t ∈ (t1, t2) | and f (t) |〉0, S is decay multiplying power, effective wave attenuation multiplying power S＜1, the decay multiplying power S of ground roll〉1, L _x, L _y, L _zBe the time period threshold of vector space, E ₁, E ₂Energy threshold for vector space.

8. the vector noise reducing apparatus of a multi-component seismic data, is characterized in that, described equipment comprises:

Harvester is used for gathering multi-component seismic data;

The wavelet decomposition device is used for described multi-component seismic data is carried out wavelet decomposition, obtains the seismic signal of a plurality of yardsticks;

The Vector Mode value is determined device, is used for determining the Vector Mode value corresponding to seismic signal of each yardstick;

Angle cosine is determined device, is used for determining the angle cosine corresponding to seismic signal of each yardstick;

Filtering device is used for random noise and spike pulse according to the described seismic signal of Vector Threshold method filtering;

The compacting compensation system is used for the seismic signal after filtering random noise and spike pulse is carried out ground roll compacting, significant wave compensation;

The wavelet reconstruction device is used for carrying out wavelet reconstruction according to seismic signal and described angle cosine after ground roll compacting, significant wave compensation, obtains the multi-component seismic data after the vector noise reduction process.

9. equipment according to claim 8, is characterized in that, described Vector Mode value determines that device is undertaken by following formula:

$p\left(t\right)=\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, and p (t) is the Vector Mode value of multi-component seismic data.

10. equipment according to claim 8, is characterized in that, described angle cosine determines that device is undertaken by following formula:

$a_x\left(t\right)=\frac{x\left(t\right)}{\sqrt{{x\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_y\left(t\right)=\frac{y\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

$a_z\left(t\right)=\frac{z\left(t\right)}{\sqrt{x{\left(t\right)}^2+y{\left(t\right)}^2+z{\left(t\right)}^2}}$

Wherein, x (t), y (t), z (t) are three different component recordings of multi-component seismic data, a _x(t), a _y(t), a _z(t) be the angle cosine of three different components.

11. equipment according to claim 9 is characterized in that, described filtering device is undertaken by following formula:

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ _iBe the Vector Mode value threshold of i yardstick,

Vector soft-threshold method for wavelet analysis.

12. equipment according to claim 9 is characterized in that, described filtering device is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain,

Vector hard threshold method for wavelet analysis.

13. equipment according to claim 9 is characterized in that, described filtering device is undertaken by following formula:

$D_{\mathrm{hard}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<\mathrm{\&lambda;}\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;\mathrm{\&lambda;}\end{array}\right.$

$D_{\mathrm{soft}}^i=\left\{\begin{array}{cc}0& \mathrm{if}\left|D_i\right|<{\mathrm{\&lambda;}}_i\\ D_i& \mathrm{if}\left|D_i\right|\&GreaterEqual;{\mathrm{\&lambda;}}_i\end{array}\right.$

Wherein, p (t) is carried out being D after discretize is processed _i, i is the yardstick sequence number, λ is the fixedly threshold of Vector Mode codomain, λ _iBe the Vector Mode value threshold of i yardstick,

The vector hard threshold method of wavelet analysis,

Vector soft-threshold method for wavelet analysis.

14. equipment according to claim 8 is characterized in that, described compacting compensation system is undertaken by the space zero-crossing filter, describedly is expressed as follows by the space zero-crossing filter:

$p\left(t\right)=\left\{\begin{array}{c}\frac{p\left(t\right)}{S}\mathrm{ift}\&Element;(t1,t2),(t1,t2)\&SubsetEqual;(L_x,L_y,L_z)\mathrm{and}{E}_1<\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}<E_2\\ p\left(t\right)\mathrm{ift}\&Element;(t1,t2),(t1,t2)\&NotSubset;(L_x,L_y,L_z)\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&GreaterEqual;E_2\mathrm{or}\underset{t1}{\overset{t2}{\&Integral;}}p{\left(t\right)}^2\mathrm{dt}\&le;E_1\end{array}\right.$

Wherein, t1, t2 are the adjacent zero crossing of either direction, f (t1)=0, and f (t2)=0, when t ∈ (t1, t2) | and f (t) |〉0, S is decay multiplying power, effective wave attenuation multiplying power S＜1, the decay multiplying power S of ground roll〉1, L _x, L _y, L _zBe the time period threshold of vector space, E ₁, E ₂Energy threshold for vector space.

Images