Content of the invention
For the deficiencies in the prior art, an object of the present invention is to solve present in above-mentioned prior art
One or more problems.For example, an object of the present invention is to provide one kind can be better achieved to surface seismic data
Carry out the method compensating with recovering of real amplitude.
To achieve these goals, the invention provides a kind of surface seismic data amplitude compensation based on vsp preliminary wave
Method.The method comprising the steps of: a, the down going wave isolating in vsp data, and pick up in the down going wave of vsp data
The first arrival time of per pass vsp data, and obtain the corresponding amplitude of first arrival time of per pass vsp data, form a data sequence;
B, adopt data sequence described in exponential curve fitting, and obtain the index x of described exponential curve using nonlinear inversion method;
C, the amplitude of each sampling point of surface seismic data seismic channel is multiplied by respectively the x power of the corresponding time value of described each sampling point, with
Time function gain is carried out to surface seismic data and realizes amplitude compensation.
Alternatively, step a includes: a1, reading zero-offset vsp data;A2, separately described zero-offset vsp data
Upgoing wave and down going wave;A3, per pass is obtained according to the through wave velocity of the corresponding well depth of per pass vsp data and given lower shape ripple
The vsp data corresponding time;A4, corresponding for described per pass vsp data time moved down after the scheduled time with described per pass
Window when constituting between the vsp data corresponding time;A5, per pass vsp data when window in count peak swing;A6, ask for every
Per pass vsp data is extracted through described step a2 the first arrival time of road vsp data, and the down going wave of the vsp data after processing
The amplitude a corresponding to first arrival timei, wherein, certain amplitude corresponding to one vsp data first arrival time and this road vsp data
First arrival time adjacent and be located at this road vsp data first arrival time before the amplitude corresponding to two time points and with this road
Vsp data first arrival time is adjacent and is located at the amplitude sum corresponding to two time points after this road vsp data first arrival time
More than described peak swing divided by correction factor gained value;Wherein i is the road serial number of vsp data, and 1≤i≤n and i is just
Integer, n is the maximum road number of zero-offset vsp data.
Alternatively, step a2 includes: described zero-offset vsp data is made with two-dimensional Fourier transform with by Space Time domain
Data transforms in frequency wavenumber domain, makees Filtering Processing to the figure after conversion, so that up wave attenuation, then will convert again
Data afterwards makees two-dimensional Fourier transform to switch back to the data of frequency wavenumber domain in Space Time domain.
Alternatively, step b includes:
B1, randomly generate m value x in the span of index x using random functionj, 1≤j≤m, and j is just whole
Number, obtains exponent data group { x1,x2,…,xm, wherein, the span of described index x is [1,3];
B2, carry out first time iteration to ask for optimality index x, comprise the steps:
B21, from exponent data group { x produced by described step b11,x2,…,xmArbitrarily choose value xjAnd according to
Equation below (1) is calculated corresponding calculated amplitude value a of every one vsp data first arrival timei':
(fi)x=ai' (1)
B22, set up object function, and according to the calculated a of described step b21i' and from vsp data pickup every
The corresponding amplitude a of road vsp data first arrival timeiCalculate ej:
In formula (1) and (2), fiFor the first arrival time of the i-th road vsp data, i is the road serial number of vsp data, 1≤i
≤ n and i are positive integer, and n is the maximum road number of zero-offset vsp data;
B23, according to described step b21 and b22, be calculated m x produced by step b1jM, corresponding ground ej, and look for
Go out m ejIn minima eminCorresponding index xbestWith m ejIn maximum emaxCorresponding index xworst;
B24, as described m ejIn minima eminDuring less than predictive error precision ε, by index xbestAs optimality index
X, otherwise, continues iteration next time;
B3, carry out the τ time iteration, τ >=2 and be positive integer, to ask for optimality index x, comprise the following steps:
B31, m value x that described step b1 is randomly generatedjIt is calculated new index number according to equation below (3)
According to x':
In formula (3),-0.5≤aj≤1.5;
B32, x' is substituted in described formula (1) and (2), if calculated e-value is less than described m ejIn maximum
Value, then will substitute described exponent data group { x with x'1,x2,…,xmIn xworst, form new exponent data group;
B33, on the basis of described new exponent data group, recalculate according to described step b21~b23 and obtain minima
emin', maximum emax', index xbest' and index xworst′;
If b34 is emin' < ε, by index xbest' as optimality index x, otherwise, continue iteration next time.
Compared with prior art, the present invention asks for referring to by vsp first arrival automatic Picking and using the method for non-linear inversion
Number x, and time function gain compensation is carried out to surface seismic data according to index x, it is capable of the deep layer of effective compensation geological data
Energy, it is achieved that surface seismic data is carried out with compensation and the recovery of real amplitude, is favorably improved the prediction essence of oil-bearing reservoir
Degree.
Specific embodiment
Hereinafter, the ground based on vsp preliminary wave according to the present invention will be described in detail with reference to exemplary embodiment
Geological data vibration amplitude compensation method.
Vsp (vertical seismic profiling, i.e. VSP) earthquake record (or referred to as vsp ground
Shake data, vsp seismic data, vsp data) observed pattern of downhole receiving is excited for ground, that is, shot point is arranged on ground and examines
Wave point is arranged in well, and its descending direct wave is affected by noise less, can more accurately reflect underground compared with surface seismic record
Medium proposes to extract earth-attenuation index using the through down going wave of vsp to amplitude attenuation, therefore inventor, and is referred to using this
Several surface seismic data that the same area is collected carry out time function gain compensation.Here, down going wave and upgoing wave be by
The direction traveling to receiving point according to ripple to be changed point, wherein, is down going wave, below receiving point above receiving point
Call traveling wave.
Surface seismic data (or referred to as surface seismic record, surface-seismic data) refers to that ground excites the sight of ground receiver
Survey mode, that is, on the ground and geophone station is also provided in ground for shot point setting.
Specifically, a kind of surface seismic data vibration amplitude compensation method based on vsp preliminary wave, the method are inventors herein proposed
Isolate the down going wave in vsp data first, then in the down going wave of vsp data automatic Picking per pass vsp data first arrival (i.e.
First arrival time), and obtain the corresponding amplitude of first arrival time of per pass vsp data take out formation one DS, adopt one
This data of bar exponential curve fitting, thus obtain the index x determining, further according to this index x, to the surface seismic data collecting
Carry out time function gain compensation, thus realizing surface seismic data is carried out with compensation and the recovery of real amplitude.Wherein, preliminary wave
Refer to first arrival time: when seismic wave wavefront arrives first at certain observation station, the particle of this medium start to occur vibration when
Carve, the referred to as first arrival time of ripple, and the ripple recorded, referred to as preliminary wave that is to say, that in earthquake record record the
The ripple of one arrival is referred to as preliminary wave.
The present invention is achieved through the following technical solutions:
A, the down going wave isolated first in vsp data, then automatic Picking per pass vsp number in the down going wave of vsp data
According to first arrival time, and obtain per pass vsp data the corresponding amplitude of first arrival time formed a data sequence;
B, adopt data sequence described in an exponential curve fitting, and described index is obtained using nonlinear inversion method
The index x of curve;
C, the various kinds point of surface seismic record seismic channel is multiplied by the x power of the corresponding time value of each sampling point, with over the ground
Face geological data carries out time function gain compensation.
In one exemplary embodiment, the geological data vibration amplitude compensation method based on vsp preliminary wave according to the present invention
Comprise the following steps:
(1) read in zero-offset vsp data v (i, t).
Wherein, be zero-offset vsp data road serial number, i={ 1,2......, n }, n be zero-offset vsp data
Maximum road number;T is the time, and t={ 1,2......, t }, t are the zero-offset vsp Data Data dominant record time.Below
In, with regard to the implication all same of i, repeat no more.
(2) upgoing wave and the down going wave of zero-offset vsp data are separated.
This step can separate the upgoing wave of zero-offset vsp data and the side of down going wave using commonly used in the prior art
Method, for example, longitudinal stack, multiple tracks pie slice, f-k filtering, medium filtering, best of breed filtering, least squares filtering etc..
In the present embodiment, filtered using frequency wavenumber domain.That is, by two dimension is done to zero-offset vsp data s (i, t)
The data in Space Time domain is transformed in frequency wavenumber domain by Fourier transformation, and at this moment in positive half-plane, upgoing wave exists down going wave
In negative half-plane;Filtering Processing is made to the figure after conversion, the data in negative half-plane is multiplied by decimal (such as 0.00001) and makes
Traveling wave decays, and the down going wave of positive half-plane is unaffected;Data after conversion is returned to Space Time domain in the two-dimentional Fourier transform of work
Result sdw(i, t), upgoing wave has been decayed, and down going wave strengthens.
(3) read every track data corresponding well depth h from vsp data trace headeri, give the through wave velocity of down going wave simultaneously
V, and it is corresponding to obtain per pass vsp data according to the through wave velocity of the corresponding well depth of per pass vsp data and given lower shape ripple
Time, that is,
(4) the time t being obtained according to the every track data of step (3)i, move down time tl, window (t when obtainingi,ti+tl).This
Place, tlValue to comprise the first arrival in all roads according to the when window being so that formation as far as possible, but be difficult excessive, typically take tl=
100ms.
(5) in the when window (t of per pass vsp datai,ti+tl) interior statistics peak swing, obtain maxampi.
(6) ask for the first arrival time of per pass vsp data, specifically:
Order takes window (t during the i-th roadi,ti+tl) in a time point tp, by this time point tpAmplitude and with its before and after
It is adjacent to two time point tp-2、tp-1、tp+1、tp+2Corresponding amplitude is added summation and obtains amplitude sub, when this amplitude
Sub is more than maxampiDivided by the value of correction factor sf gained, when being considered as the first arrival that this time point is corresponding vsp track data
Between.
That is: sub=sdw(i,tp-2)+sdw(i,tp-1)+sdw(i,tp)+sdw(i,tp+1)+sdw(i,tp+2), if sub >
(maxampi/ sf), then tpIt is exactly the first arrival time of this road vsp data, wherein, sf refers to a corrected parameter, can pass through
Set a sf value and observe each road first arrival that automatic Picking obtains and manual observation to the difference of the actual first arrival in each road to sentence
Whether rationally disconnected this correction factor sf value, if the unreasonable value adjusting sf again, for example, first takes sf=1, in the manner described above
Obtain the first arrival time of per pass vsp data, and the difference of the first arrival time of artificial judgment pickup and actual first arrival time, if difference
Larger, again choose sf value and pick up again, the first arrival until picking up is consistent that is to say, that being directed to substantially with actual first arrival
Data with different, can adjust the precision of sf value adjustment pickup, sf=10 in this example.Set in the present embodiment and automatically pick up
The method taking the vsp data first arrival of all roads, but the invention is not restricted to this, per pass can also be picked up by the method for manual observation
The first arrival time of vsp data, but the automatic Picking mode according to the present embodiment, decrease the workload of artificial pickup, save
Time.
Wherein, sdw(i,tp-2) represent the t of the i-th road vsp datap-2Moment corresponding amplitude, sdw(i,tp+1)、sdw(i,
tp)、sdw(i,tp+2) implication in the same manner.
(7) corresponding first arrival time f of per pass is obtained according to above (1)~(6) stepi, will be from the vsp after step (2) is processed
Data sdwCorresponding taking-up per pass first arrival time corresponding amplitude a in (i, t)i.
(8) set up following relational expression (1), ask for index x:
(fi)x=ai' (1)
Wherein, fiFor the first arrival time of the i-th road vsp data, for example, f1Represent first arrival time t of the 1st road vsp datap, f2
Represent first arrival time t of the 2nd road vsp datap.
(9) in the span [1,3] of given index x, that is, x, in the range of 1~3, randomly generates m with random function
Value xj, 1≤j≤m, and j is positive integer, obtains exponent data group { x1,x2,…,xm, and carry out according to step (10)~(15)
First time iteration, τ=1.
(10) index x is from exponent data group { x produced by step (9)1,x2,…,xmArbitrarily choose value xjAnd root
It is calculated x according to above formula (1) and be taken as xjWhen corresponding calculated amplitude value a of every one vsp data first arrival timei′;
(11) set up object function, and according to the calculated a of described step (10)i' and step (7) from vsp data
The corresponding amplitude a of every track data first arrival time of pickupiCalculate x and be taken as xjWhen ej:
(12) according to described step (10) and (11), it is calculated m x produced by (9)jM, corresponding ground ej, and look for
Go out m ejIn minima eminAnd its corresponding index xbest, m ejIn maximum emaxAnd it is corresponding
Index xworst
(13) give error precision ε=0.001.
(14) work as emin> ε, then repeat below step (15)~(19);Otherwise, by minima eminCorresponding
Index xbestAs optimum index x, and carry out step (20).
(15) τ=τ+1, carries out next iteration.
(16) according to equation below (3) to { x1,x2,…,xmCarrying out recombinating generates new data x':
In formula (3),-0.5≤ai≤1.5.
(17) this new data x' is substituted into formula (1) and (2), if the calculated e-value of fruit is less than xworstCorresponding e-value
(i.e. emax), then will substitute described exponent data group { x with x'1,x2,…,xmIn xworst, form new exponent data group.
(18) on the basis of the new exponent data group of gained, according to step (8), (10), (11) and (12), that is, again according to
Formula (1) and (2), recalculate minima e obtaining emin', the maximum e of emax' and the corresponding index x of its differencebest′、
Index xworst′;
(19) if emin' < ε, by index xbest' as optimality index x, and carry out step (20);Otherwise, continue next time
Iteration, that is, repeat above step (15)~(18).
(20) according to index x, surface seismic data is carried out with the recovery of real amplitude, the amplitude of each sampling point of seismic channel is taken advantage of
With the x power of its time value it may be assumed that
a0=attx(4)
Wherein, a0It is the amplitude after compensating, atRefer to the amplitude of a sampling point of seismic channel, t refers to atThe corresponding time
Value (moment in other words).That is, above (1st) be provided to obtain index x, per pass surface seismic data to (19) step
It is all with identical index x, formula (4) is that to surface seismic data seismic channel carries out amplitude compensation, other earthquakes
Road also compensates according to this formula (4).
According to the present invention, by vsp first arrival automatic Picking, and quantitative obtain index x, without empirically or experimentally coming
Given index x, it is to avoid lose time and surface seismic data compensation effect bad problem, realizes surface seismic data is entered
The compensation of row real amplitude and recovery.Apply the present invention in the amplitude compensation process of surface-seismic data, being capable of effective compensation
The deep energy of surface seismic data, realizes the true amplitude recovery of surface seismic data, is so favorably improved oil-bearing reservoir
Precision of prediction;This technical side's convenient to operate, the speed of service are fast simultaneously, disclosure satisfy that the demand of actual production.Using the present invention
Further positive role can be played during seismic data treatment, application prospect is good.
Although above by describing the present invention with reference to exemplary embodiment, those skilled in the art should be clear
Chu, in the case of the spirit and scope being limited without departing from claim, can be carried out respectively to the exemplary embodiment of the present invention
Plant modifications and changes.