CN104268404A - Automatic pick-up method of residual depth - Google Patents

Abstract

The invention discloses an automatic pick-up method of residual depth. The automatic pick-up method of the residual depth is characterized by including full-automatic picking based on energy superposed graph, semi-automatic picking based on fixed window fourth-order cumulant estimation and determining of applying time of two pick-up strategies, the full-automatic picking is used in an earlier stage of a travel-time chromatography inversion process, capable of accelerating chromatography iteration speed without manual intervention and suitable for inversion of large scale speed features, the semi-automatic picking is used in a later stage of the travel-time chromatography inversion process, capable of increasing accuracy of residual depth picking with manual intervention and suitable for inversion of fine scale speed features, and the applying time of the two pick-up strategies can be determined through recognition of energy superposed graph features.

Description

A kind of residue degree of depth automatic pick method

Technical field

The present invention relates to the method for in the whilst on tour tomographic inversion system robotization pickup residue degree of depth, for the different pickup strategy of refutation process early stage and late design.

Background technology

Whilst on tour chromatography is the one of chromatography method, is to adopt the propagation of high-frequency approximation theoretical modeling seismic event in underground of ray, utilizes the relation set up ray travel routes and hourage between observation data and subterranean depth and speed.Whilst on tour chromatography comprises multiple process, many wheel iteration are needed to carry out inverting target signature, the pickup of the residue degree of depth is an important step of often taking turns in iteration, it is the bridge between contact spatial information and temporal information, its pickup efficiency and accuracy to overall chromatography efficiency and influential effect great, as shown in Figure 1.

Based in the grid chromatography of ray, pre-stack depth migration can obtain the imaging body of one group of redundancy under different offset distance, and for a reflection spot of underground, if migration velocity is correct, so the position of this reflection spot in different imaging body is the same; On the contrary, if speed is incorrect, this may change in the imaging body position of different offset distance.And the distance between zero-offset position is called the residue degree of depth.The pickup of the residue degree of depth is a cumbersome job, and early stage pickup is main by manually completing, inefficiency, and it is low to pick up precision.Along with improving constantly of chromatography scale and model resolution, the use of angles of azimuth data set in addition, simple artificial pickup is difficult to be competent at.At present, extensively employing pick-up method is the method based on curve, namely to the energy distribution situation of designated analysis point on different offset distance, carries out best-fitting of the curve, according to the matched curve automatic Picking residue degree of depth.Although workload alleviates greatly, but still do not break away from manual pickup, nx bar Inline line is had with one, the imaging model of ny bar Crossline line and nz degree of depth section is example, if adopt the relatively simple para-curve of parameter, still need to pick up nx × ny × nz point, for one for the fairly large and chromatography problem needing successive ignition, pick up workload and still can not despise.In addition, conic fitting is the pickup based on variation tendency, when offset distance step-length is less, the residue depth information error of pickup is larger, be unfavorable for the target signature of inverting fine dimension, the parametric line of high-order can realize the pickup of fine dimension, but parameter of curve itself is determined very difficult.Higher Order Cumulants has the insensitive characteristic of Gaussian noise, is successfully utilized to follow the trail of stratum and tomography lineups, but has not yet to see the relevant report that application Higher Order Cumulants carries out remaining degree of depth pickup.

The invention discloses a kind of method remaining degree of depth automatic Picking, for chromatography iterative process early stage and later stage different characteristics, devise two kinds of different pickup strategies.Pickup based on energy supposition figure is the extension to curve thought, automatically maximum stack energy curve is determined by autoscan curve family, and carry out automatic selection analysis point by extracting energygram feature, the automatic Picking residue degree of depth, be suitable for the early stage of refutation process, to accelerate the convergence of chromatography iterative process, the target signature of inverting large scale; According to energygram feature, the fourth order cumulant that automatically can proceed to the later stage estimates pickup, to extract meticulous residue depth information, the target signature of inverting small scale.

Summary of the invention

The technical problem to be solved in the present invention is: in conjunction with the feature of chromatography inversion iterates process, is respectively early stage and the later stage provides two kinds of different residue degree of depth pickup strategies, picks up efficiency to solve in current method and picks up the difficulty that precision can not take into account simultaneously.

The present invention solves the technical scheme that its technical matters takes: provide a kind of method remaining degree of depth automatic Picking, for early stage and the later stage of whilst on tour tomographic inversion process, devises two kinds of different pickup strategies respectively.It is characterized in that, based on the full-automatic pickup of energy supposition figure, be used in chromatography iterative process in earlier stage; Based on the semi-automatic pickup that stationary window fourth order statistic is estimated, be used in the chromatography iterative process later stage.

Based on the full-automatic pickup of energy supposition figure, it is characterized in that, comprise the following steps:

Steps A, according to common imaging gather under different offset distance, structure energy supposition figure automatically;

Step B, extracts energy supposition figure feature, full-automatic selection chromatographic analysis point, the residue degree of depth under the different offset distance of full-automatic pickup.

The structure of energy supposition figure in steps A, further comprising the steps:

A1, selected matching base curves, and the coverage of assignment curve and curve step-length; Curve ranges refers to the intersection point scope of curve and offset gather farthest, and curve step-length refers to the step-length of intersection point, determines the quantity of curve and the length of energy supposition figure transverse axis;

A2, according to the zero-offset degree of depth and maximum offset degree of depth determination parameter of curve;

A3, according to parameter of curve curve construction, the signal energy of superimposed curves and the not intersection point of common-range gather, and generate energy supposition figure.

In step B, the principal character of energy supposition figure comprises:

B1, each point on energy supposition figure represents stack power corresponding to a matched curve, and the energy strong point of every a line represents the optimum fit curve at this depth location place, i.e. aim curve;

B2, energy supposition figure are divided into two parts in left and right, and energy strong point is positioned on the left of center line and represents that Tomography Velocity is on the low side, are positioned at right side and then represent that speed is higher;

B3, the Weighted distance of energy strong point and center line and, represent the departure degree of Tomography Velocity and target velocity, as the foundation on opportunity proceeding to the semi-automatic pickup estimated based on the fourth order statistic of stationary window from the full-automatic pickup based on energy supposition figure, also can as the foundation judging to stop in chromatography iterative process;

B4, carries out energy ordering to energy strong point, according to energy threshold and the subregion setting of setting, automatically determines chromatographic analysis point.

Based on the semi-automatic pickup that stationary window fourth order statistic is estimated, it is characterized in that, comprise the following steps:

Steps A, the fixing analysis window size of setting, Selecting parameter should avoid steep dip to scan;

Step B, calculates fourth order statistic to each space delay step-length in fixing analysis window;

Step C, the delay that calculating obtains maximal value is the target delay of data track relative to library track, new seed point is determined according to delay, data track becomes library track, follow-up road becomes new data road, the like, until track offset gather or fourth order statistic value to be farthest less than the threshold value preset.

The present invention is compared with prior art advantageously: the pickup strategy of energy supposition figure utilizes finds a stack power maximum curve with bunch curve autoscan, achieve the full-automatic pickup automatically choosing and remain the degree of depth of analysis site, be applicable to chromatography process inverting in early stage large scale velocity characteristic; Meanwhile, by the feature of analysing energy stacking diagram, chromatography status information can be obtained, determine the conversion opportunity picking up strategy, contribute to the flexible control of chromatography process.Based on the pickup of fourth order statistic, can catch the meticulous change of the residue degree of depth, pickup accuracy is high, but needs manually to pick up Seed Points, belongs to semi-automatic pickup, is applicable to chromatography process later stage inverting small scale velocity characteristic.The automatically working of residue degree of depth pickup can improve efficiency and the effect of chromatography work in actual production greatly.

Accompanying drawing explanation

Fig. 1 is the basic procedure of the ray tomography Inversion System based on grid;

Fig. 2 is for adopting para-curve (left side) and hyperbolic curve (right side) autoscan respectively and stack power to appointment chromatography analysis site;

Fig. 3 a is the original common imaging gather of part;

Fig. 3 b is the automatic Picking effect directly adopting energy supposition figure to pick up strategy, and pickup curve has crossing phenomenon, there is ambiguity;

Fig. 3 c is the pickup effect after disambiguation;

Fig. 4 is the schematic diagram estimating pickup based on fourth order statistic;

Fig. 5 adopts para-curve to pick up strategy Analysis on Selecting point and the automatic Picking residue degree of depth automatically as the energy supposition figure of matching base curves;

Fig. 6 adopts hyperbolic curve to pick up strategy Analysis on Selecting point and the automatic Picking residue degree of depth automatically as the energy supposition figure of matching base curves;

Fig. 7 is the energy supposition figure (left side) and automatic Picking effect (right side) that on complicated migrated section, a track data is corresponding;

Actual shifts section carries out residue degree of depth automatic Picking effect to appointment track data to Fig. 8;

Fig. 9 a is the original common imaging gather of part;

Fig. 9 b directly adopts energy supposition figure to pick up the automatic Picking effect of strategy in the chromatography iteration later stage;

Fig. 9 c adopts fourth order statistic to estimate the pickup effect of pickup strategy in the chromatography iteration later stage.

Embodiment

The residue degree of depth has substantial connection with offset distance change with the yardstick of offset distance.When offset distance step-length is less, change finer and smoother, and when yardstick is larger, due to overlaying relation, this change is more level and smooth.Theoretically, this consecutive variations can use curve carry out matching, but when this change slightly shows complicated, the determination of curve itself becomes very difficult, even if choose the less quafric curve of some parameters, also needs extra work to determine the parameter of curve.In order to improve automatization level further, can cluster curve autoscan analysis site be used, superpose signal energy on every bar curve, choose the maximum curve of stack power as optimum fit curve; A nearlyer step, can all depth points of autoscan, structure energy supposition figure, by extracting energy supposition figure feature, the automatic Picking automatically choosing and remain the degree of depth of Realization analysis point.

The structure of energy supposition figure and pick-up method are according to following steps:

1) selected matching base curves, the coverage of assignment curve and curve step-length.Curve ranges refers to the intersection point scope of curve and offset gather farthest, and curve step-length refers to the step-length of intersection point, determines the quantity of curve and the length of energy supposition figure transverse axis.

2) according to the zero-offset residue degree of depth and maximum offset residue depth calculation parameter of curve;

3) parameter of curve curve construction, the signal energy of superimposed curves and the not intersection point of common-range gather, and generate energy supposition figure.

4) optimum fit curve (maximum stack energy) is determined to each depth point;

5) set relative energy threshold value or energy subregion, the depth point higher than threshold value elects analysis site as, and according to the optimum fit curve automatic Picking residue degree of depth.

Matching base curves can be para-curve, hyperbolic curve or other quafric curve.Para-curve is one-parameter curve, can uniquely determine parabolical focus by the zero-offset degree of depth and the maximum offset degree of depth; Compared to para-curve, hyp situation is slightly more complex, because 2 can not be unique determination hyperbolic curve, therefore the length of the real imaginary semi-axis of prior assignment curve is needed, in scanning process, root also needs the symbol changing parameter of curve according to curve relative to the position of the zero-offset degree of depth, and does corresponding translation.Para-curve and hyperbolic curve is adopted to carry out autoscan process as shown in Figure 2.

Energy supposition figure reflects the superposition situation with each offset distance signal energy under different curve in bunch curve, and energy supposition figure has following characteristics:

1) each point on energy supposition figure represents stack power corresponding to a matched curve, and energy strong point represents the optimum fit curve at this depth location place, i.e. aim curve;

2) energy supposition figure is divided into two parts in left and right, and strong point is positioned at left side and represents that Tomography Velocity is on the low side, is positioned at right side and then represents that speed is higher;

3) energy strong point and center line Weighted distance and, represent the departure degree of Tomography Velocity and target velocity, can be used as the foundation on opportunity proceeding to the semi-automatic pickup that the fourth order statistic based on stationary window is estimated from the full-automatic pickup based on energy supposition figure, also can as the foundation judging to stop in chromatography iterative process;

Pickup strategy based on energy supposition figure can realize the full-automatic pickup remaining the degree of depth, but may there is ambiguity situation, and Fig. 3 b describes this situation.Two distances are close to depth location having occurred the phenomenon that matched curve intersects.Usually, two ground coordinatees are identical, degree of depth analysis site relatively, and the residue change in depth on different offset distance is more similar, because they have identical overlying strata, this phenomenon is especially obvious in deep, and occurs inverse variation trend and intersect to there is ambiguity.In order to eliminate this ambiguity, need to carry out a top-down scanning again from the analysis site determined, for the situation of many curve intersections, the curve that the stack power in reservation intersection curve is maximum and corresponding analysis site (Fig. 3 c).

Pickup strategy based on energy supposition figure achieves the full-automation realizing the pickup of the residue degree of depth, accelerates chromatography iteration speed, but to lose tomographic inversion small scale velocity characteristic as cost.Because when offset distance yardstick is larger, due to synergistic effect, the otherness of the residue degree of depth can be reduced, the average case of the residue degree of depth of energy reflection certain deviation in scope of superposition, is suitable for inverting large portion velocity characteristic; And when offset distance yardstick reduces, the otherness brought by little velocity characteristic can show, adopt simple quafric curve to carry out matching pickup and can miss these tiny characteristics.Fig. 9 a is that of chromatography iteration later stage is total to imaging point section, has strong lineups, reflect offset distance from 1500 meters to the distribution situation of the residue degree of depth within the scope of 5000m the degree of depth 1850 meters.Can find out, little offset distance scope (being less than 1500 meters) does not significantly remain depth information, and this phenomenon may be the low-speed region owing to existing above analysis site among a small circle, as gas bag or fault block etc.Fig. 9 b is the automatic Picking based on energy supposition drawing method, and pickup effect is also unsatisfactory, and one is that curve effect is bad, and two is that mistake has picked up the residue degree of depth of little offset distance scope.Therefore, the full-automatic pickup based on energy supposition figure is used in chromatography iterative process in earlier stage, for inverting large scale velocity characteristic, and makes iterative process be tending towards convergence as early as possible; And in the later stage, need to adopt residue degree of depth pick-up method more accurately, and to ensure to pick up precision, inverting small scale target signature.

In the chromatography iteration later stage, adopt the semi-automatic pickup strategy that fourth order statistic is estimated.If s (k) is source signal, d is data delay, w _ik () (i=1,2) are random Gaussian, and independent with s (k), and { x (k) } and { y (k) } is respectively library track and data track, i.e. x (k)=s (k)+w ₁(k), y (k)=s (k-d)+w ₂(k).Character according to Higher Order Cumulants: the Higher Order Cumulants of the seismic signal containing Gaussian noise is equal with the Higher Order Cumulants of useful signal, this character can be effective to resist noise, improves the correctness of Event tracking.The estimation function formula of fourth order cumulant is:

$J\left(d\right)=\frac{\left|{\mathrm{cum}}_4\right\{x(k-d),x(k-d),y\left(k\right),y\left(k\right)\left\}\right|}{\sqrt{\left|{\mathrm{CUM}}_4\right\{x\left(k\right)\left\}\right|\left|{\mathrm{CUM}}_4\right\{y\left(k\right)\left\}\right|}}$

Wherein,

$\begin{array}{c}{\mathrm{cum}}_4\{x(k-d),x(k-d),y\left(k\right),y\left(k\right)\}=\\ \frac{1}{N}\underset{{k=K}_1}{\overset{K_2}{\mathrm{\Σ}}}{x}^2{(k-d)y}^2\left(k\right)-2{\left[\underset{k=K_1}{\overset{K_2}{\mathrm{\Σ}}}x(k-d)y\left(k\right)\right]}^2\\ -\left[\frac{1}{N}\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{x}^2\left(k\right)\right]\left[\frac{1}{N}\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{y}^2\left(k\right)\right]\end{array}$

K ₁＝max(1,d+1),K ₂＝max(K,K+d)，

CUM ₄{x}＝cum ₄{x,x,x,x}.

Although fourth order cumulant theory is usually used in carrying out time delay estimadon, according to the characteristic distributions of the residue degree of depth on different offset distance component, use the method equally very effective.The method pickup estimated with fourth order statistic belongs to semi-automatic technique, need manually to pick up Seed Points (namely picking up the residue degree of depth of start offset distance), then automatic tracing remains strong lineups (adjacent has the residue degree of depth that the offset distance of residue depth information is corresponding), the distribution situation on each offset distance component according to the yardstick of offset distance and the residue degree of depth, analysis site may have the pickup more than 1 time.Pickup step is as follows:

1) the fixing analysis window size of setting, namely data track is relative to the dip scanning scope of library track, and the selection of this parameter should avoid steep dip to scan;

2) in fixing analysis window, fourth order statistic is calculated to each space delay step-length;

3) delay obtaining maximal value is the space delay of data track relative to library track, new seed point is determined according to delay, data track becomes library track, follow-up road becomes new data road, the like, until track offset gather or J (d) value to be farthest less than the threshold value (in order to prevent sky information or to pick up without the residue degree of depth of obvious information) preset.

In sum, following embodiment is provided.The residue degree of depth automatic Picking design sketch of Fig. 5 ~ Figure 9 shows that embodiment of the present invention.Fig. 5 and Fig. 6 is for gross data, para-curve and hyperbolic curve is adopted to pick up the design sketch of strategy Analysis on Selecting point and the automatic Picking residue degree of depth automatically as the energy supposition figure of matching base curves respectively, in figure, grey square frame represents the analysis site automatically chosen, and Grey curves represents the residue degree of depth of automatic Picking; Fig. 7 is the energy supposition figure and automatic Picking effect that on complicated migrated section, a track data is corresponding, and relative energy threshold value is set to 0.5; Fig. 8 is the effect carrying out remaining degree of depth automatic Picking on actual shifts section to one group of skew track data; For adopting energy supposition figure to pick up strategy (Fig. 9 b) and fourth order statistic respectively in the chromatography iteration later stage, Fig. 9 estimates that pickup strategy (Fig. 9 c) remains Contrast on effect of the degree of depth to same analysis site pickup, wherein, Fig. 9 c is using 1500 meters of offset distances as initial seed.

Certainly; the present invention can also have other various embodiments; when not deviating from the present invention's spirit and essence thereof; those of ordinary skill in the art are when making various corresponding change and distortion according to the present invention, but these change accordingly and are out of shape the protection domain that all should belong to the claim appended by the present invention.

Claims (5)

1. remain a method for degree of depth automatic Picking, be used in the system of whilst on tour tomographic inversion, for chromatography iterative process early stage and the feature in later stage, take different pickup strategies, it is characterized in that, based on the full-automatic pickup of energy supposition figure, be used in chromatography iterative process in earlier stage; Based on the semi-automatic pickup that stationary window fourth order statistic is estimated, be used in the chromatography iterative process later stage.

2. a kind of method remaining degree of depth automatic Picking according to claim 1 ,it is characterized in that, the full-automatic pickup based on energy supposition figure comprises the following steps:

Steps A, according to the common imaging gather of different offset distance, structure energy supposition figure automatically;

Step B, extracts energy supposition figure feature, full-automatic selection chromatographic analysis point, the residue degree of depth under the different offset distance of full-automatic pickup.

3. a kind of method remaining degree of depth automatic Picking according to claim 1, is characterized in that, the semi-automatic pickup estimated based on stationary window fourth order statistic comprises the following steps:

Steps A, the fixing analysis window size of setting, Selecting parameter should avoid steep dip to scan;

Step B, calculates fourth order statistic to each space delay step-length in fixing analysis window;

Step C, the delay that calculating obtains maximal value is the target delay of data track relative to library track, new seed point is determined according to delay, data track becomes library track, follow-up road becomes new data road, the like, until track offset gather or fourth order statistic value to be farthest less than the threshold value preset.

4. a kind of method remaining degree of depth automatic Picking according to claim 2, is characterized in that, the structure of energy supposition figure in described steps A is further comprising the steps:

A1, selected matching base curves, and the coverage of assignment curve and curve step-length; Curve ranges refers to the intersection point scope of curve and offset gather farthest, and curve step-length refers to the step-length of intersection point, determines the quantity of curve and the length of energy supposition figure transverse axis;

A2, according to the zero-offset residue degree of depth and maximum offset residue degree of depth determination parameter of curve;

A3, according to parameter of curve curve construction, the signal energy of superimposed curves and the not intersection point of common-range gather, and generate energy supposition figure.

5. a kind of method remaining degree of depth automatic Picking according to claim 2, it is characterized in that, in described step B, energy supposition figure comprises:

B1, each point on energy supposition figure represents stack power corresponding to a matched curve, and the energy strong point of every a line represents the optimum fit curve at this depth location place, i.e. aim curve;

B2, energy supposition figure are divided into two parts in left and right, and energy strong point is positioned on the left of center line and represents that Tomography Velocity is on the low side, are positioned at right side and then represent that speed is higher;

B3, the Weighted distance of energy strong point and center line and, represent the departure degree of Tomography Velocity and target velocity, as the foundation on opportunity proceeding to the semi-automatic pickup estimated based on the fourth order statistic of stationary window from the full-automatic pickup based on energy supposition figure, also can as the foundation judging to stop in chromatography iterative process;

B4, carries out energy ordering to energy strong point, according to energy threshold and the subregion setting of setting, automatically determines chromatographic analysis point.