A method for velocity inversion of underground medium layers using reflection wave information

Abstract

The present invention provides a method for performing underground medium layer velocity inversion using reflection wave information, comprising the following steps: obtaining surface seismic records of the underground medium layer to be measured; generating a velocity spectrum; picking up the two-way travel time of the reflection signal; performing image enhancement processing on the velocity spectrum using binarization; picking up the root mean square velocity and two-way travel time from the velocity spectrum after image enhancement processing; and solving for the layer velocity using a generalized inversion method based on the picked two-way travel time, root mean square velocity, and two-way travel time of the reflection signal. The present invention provides a stable and efficient layer velocity inversion method, which improves the error tolerance and operational convenience of the inversion algorithm based on image enhancement technology and a new root mean square velocity picking strategy, reduces the impact of the root mean square velocity picking error on the inverted underground medium layer velocity model, and can effectively utilize the reflection wave information generated by the underground horizontal reflection interface to efficiently and stably invert the underground medium layer velocity model of the target to be measured.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering geological geophysical exploration, and in particular to a method for performing velocity inversion of underground medium layers using reflected wave information. Background Art

[0002] In the interpretation of traditional reflection seismic data, it is often necessary to provide a layer velocity model of the underground medium. In areas without well data or drilling data, the accuracy of time-to-depth conversion has always been a difficult problem that needs to be solved urgently. The strategy for obtaining layer velocity is currently mainly based on the stacking velocity spectrum and the Dix formula. This strategy needs to meet two main requirements: (1) the underground medium is a horizontal layered medium; (2) the accurate root mean square velocity can be picked up from the stacking velocity spectrum. In fact, for sedimentary basins and plains, the small scale (such as tens of meters) above the bedrock surface can be regarded as a horizontally layered sedimentary layer. Therefore, the calculation of the layer velocity distribution in sedimentary basins and plains is mainly limited by Problem 2.

[0003] The Dix formula uses the reflected wave information generated by the underground layered medium interface. If the V at the underground reflection interface can be picked up more accurately, rms (rms velocity of the reflecting interface) and T rms (two-way travel time of the reflection interface), the layer velocity of each layer above the reflection interface can be directly estimated recursively. Using the root mean square velocity to invert the layer velocity is a typical ill-posed problem in mathematics. A small error in the root mean square velocity may cause the layer velocity estimate to diverge. In traditional seismic data processing, V rms and T rms There are usually large errors in the picking of , so the layer velocity results directly estimated using the Dix formula are often not good. In order to obtain a stable and reliable layer velocity result, Koren and Buland proposed a constrained Dix inversion method and a Bayesian Dix inversion method respectively. The essence of this method is to add certain geological constraints in the calculation process to obtain a stable and geologically reasonable layer velocity model, but how to obtain the prior geological constraints is another difficulty. In order to improve the stability of layer velocity inversion, Yang Wencai also used the two-way travel time information T of the reflection signal as input data to jointly invert the layer velocity and reflection surface depth, but V rms 、T rms The picking of T is more complicated, and when the root mean square velocity V rms When the picking error is large, the interval velocity inversion result is also prone to large errors. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides an efficient and stable layer velocity inversion method. The method is based on the reflection wave seismic records collected on the surface, and the root mean square velocity of the underground interface reflection wave information is picked up through seismic data processing and image enhancement technology. The layer velocity model of the underground medium is stably solved by the inversion method. rms and round-trip travel time T rms Even if there are errors, relatively accurate results can still be obtained.

[0005] The present invention adopts the following technical solutions:

[0006] A method for performing underground medium layer velocity inversion using reflection wave information comprises the following steps:

[0007] Step 1: Obtain surface seismic records of the underground medium layer to be measured;

[0008] Step 2: Create a speed spectrum;

[0009] Step 3: Pick up the two-way travel time T of the reflected signal;

[0010] Step 4: Use binarization to perform image enhancement on the velocity spectrum;

[0011] Step 5: Use the RMS velocity picking strategy to pick the RMS velocity V from the velocity spectrum G after image enhancement. rms and round-trip travel time T rms ;

[0012] The root mean square velocity picking strategy is as follows: based on the velocity spectrum G after image enhancement, identify the distribution area where the element in the velocity spectrum G is 1, set a reasonable time sampling interval, and manually pick the root mean square velocity V by human-computer interaction. rms ; When manually picking up the root mean square speed V rms When the element in the velocity spectrum G is 1, the root mean square velocity V near the left and right boundaries needs to be selected at intervals. rms Value, the left and right boundaries are both V rms (L) vector and T rms (L) vector, L is the time sampling length;

[0013] Step 5: Calculate the two-way travel time T and root mean square velocity V based on the picked-up reflected signal rms and round-trip travel time T rms , using the generalized inversion method to solve the layer velocity; the step five is implemented through the following sub-steps:

[0014] (1) According to T rms The relationship between the elements in T and the time-distance curve formula of the reflected wave are listed as follows:

[0015]

[0016] Among them, T rms,1 、T rms,2 、T rms,3 、T rms,L Indicates T rms (L) The round trip travel time of the first, second, third, and Lth sampling points of the vector; V rms,1 、V rms,2 、V rms,3 、V rms,L Indicates V rms (L) is the root mean square velocity of the first, second, third, and Lth sampling points of the vector; K is the number of underground medium layers, C i is the layer velocity to be solved, i=1,2,3,……,K, namely C1, C2, C K represents the layer velocity of the first layer medium, the second layer medium, and the Kth layer medium, and the time sampling length L is greater than the number of underground medium layers K; T0 = 0, T1 represents the two-way travel time from the surface to the interface between the first layer medium and the second layer medium, T2 represents the two-way travel time from the surface to the interface between the second layer medium and the third layer medium, T K-1 T represents the two-way travel time from the surface to the interface between the K-1th layer medium and the Kth layer medium; rms，1 <T rms,2 <T1<T rms,3 ;

[0017] (2) Since the above formula is about The linear equations of equation (1) are written in the form of matrix equations:

[0018] b=Aα (2)

[0019] (3) The generalized inversion method for solving overdetermined equations is used to estimate the solution of equation (2). The least squares solution α is:

[0020] α=(A T A) -1 A T b (3)

[0021] Among them, A T represents the transpose of A, (A T A) -1 Indicates A T The inverse of A;

[0022] (4) Estimate α based on the least squares solution and obtain the layer velocity and depth of the underground medium.

[0023] Furthermore, in step 2, the velocity spectrum is prepared based on the time-distance curve formula of the seismic reflection wave:

[0024]

[0025] Where t(x) represents the two-way travel time of the reflected signal in the seismic record at x offset; t0 represents the two-way travel time of the reflected signal in the seismic record at zero offset, x represents the offset, and v represents the scanning speed;

[0026] In the process of velocity spectrum preparation, similarity coefficient is used as the criterion for velocity analysis;

[0027] The similarity coefficient is defined as:

[0028]

[0029] Where, x i is the offset of the ith channel, N is the number of channels, λ is the width of the time window, u(t i +j,x i ) represents the amplitude of the seismic data; S represents the value of the similarity coefficient, S = 1 means that the similarity coefficient reaches the maximum value, corresponding to the optimal dynamic correction speed, while in other cases, the similarity coefficient S < 1.

[0030] Furthermore, the step 4 is implemented through the following sub-steps:

[0031] (1) The mathematical average value d of the velocity spectrum D is calculated by the following formula: m :

[0032]

[0033] Where D represents the velocity spectrum of the seismic record, which is an M×N two-dimensional matrix;

[0034] (2) The element value in the statistical velocity spectrum D is greater than the average value d m The number of is defined as L;

[0035] (3) Find the statistical average value d of the velocity spectrum D s ,Right now:

[0036]

[0037] Where, α is the image enhancement control factor;

[0038] (4) The velocity spectrum D is greater than or equal to d s The element value of is assigned to 1, which is less than d s The element value of is set to 0, and the velocity spectrum G after image enhancement is obtained:

[0039]

[0040] Where G is an M×N two-dimensional matrix.

[0041] The beneficial effects of the present invention are:

[0042] (1) The present invention adopts a new RMS velocity picking strategy, which reduces the RMS velocity picking deviation and is beneficial to the stability of subsequent layer velocity inversion.

[0043] (2) The present invention adds an image enhancement step between the process of producing the velocity spectrum and performing the layer velocity inversion, which improves the error tolerance and operation convenience of the inversion algorithm, reduces the impact of the root mean square velocity picking error on the layer velocity model, and is beneficial to the stability of the subsequent layer velocity inversion.

[0044] (3) The present invention improves the error tolerance and operational convenience of the inversion algorithm based on image enhancement technology and a new root mean square velocity picking strategy, reduces the impact of the root mean square velocity picking error on the layer velocity model, and can better utilize the reflection wave information generated by the underground horizontal reflection interface to efficiently and stably invert the layer velocity model of the underground medium. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The underground medium velocity model designed in a specific embodiment 1 of the present invention;

[0046] Figure 2 In a specific embodiment 1 of the present invention, Figure 1 The generated forward modeling earthquake records;

[0047] Figure 3 In a specific embodiment 1 of the present invention, Figure 2 The generated seismic record velocity spectrum;

[0048] Figure 4 The root mean square velocity picking result after the velocity spectrum is enhanced in a specific embodiment 1 of the present invention;

[0049] Figure 5 A comparison diagram of the interval velocity obtained by the final inversion and the actual interval velocity in a specific embodiment 1 of the present invention;

[0050] Figure 6 The underground medium velocity model designed in a specific embodiment 2 of the present invention;

[0051] Figure 7 In a specific embodiment 2 of the present invention, Figure 6 The generated forward modeling earthquake records;

[0052] Figure 8In a specific embodiment 2 of the present invention, Figure 7 The generated seismic record velocity spectrum;

[0053] Figure 9 The root mean square velocity picking result after the velocity spectrum is enhanced in a specific embodiment 2 of the present invention;

[0054] Figure 10 A comparison diagram of the interval velocity obtained by the final inversion and the actual interval velocity in a specific embodiment 2 of the present invention;

[0055] Figure 11 The shot gather seismic record collected in a specific embodiment 3 of the present invention;

[0056] Figure 12 is an effective reflection wave signal intercepted from a shot gather seismic record in a specific embodiment 3 of the present invention;

[0057] Figure 13 is a velocity spectrum calculated using the effective reflected wave signal in a specific embodiment 3 of the present invention;

[0058] Figure 14 The root mean square velocity picking result after the velocity spectrum is enhanced in a specific embodiment 3 of the present invention;

[0059] Figure 15 This is the interval velocity model obtained by the final inversion in a specific embodiment 3 of the present invention. DETAILED DESCRIPTION

[0060] The present invention will be described in detail below based on the accompanying drawings and preferred embodiments. The purpose and effects of the present invention will become more apparent. The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0061] The method for performing underground medium layer velocity inversion using reflection wave information includes the following steps:

[0062] Step 1: Obtain surface seismic records of the underground medium layer to be measured.

[0063] Step 2: Create velocity spectrum. Conventional velocity spectrum creation is based on the time-distance curve formula of seismic reflection waves:

[0064]

[0065] Where t(x) represents the two-way travel time of the reflected signal in the seismic record at x offset; t0 represents the two-way travel time of the reflected signal in the seismic record at zero offset, x represents the offset, and v represents the scanning speed.

[0066] In the process of velocity spectrum preparation, similarity coefficient is used as the criterion for velocity analysis.

[0067] The similarity coefficient is defined as:

[0068]

[0069] in, x i is the offset of the ith channel, N is the number of channels, λ is the width of the time window, u(t i +j,x i ) represents the amplitude of the seismic data. S represents the value of the similarity coefficient. When S = 1, the similarity coefficient reaches its maximum value, corresponding to the optimal NMO speed. In other cases, the similarity coefficient S is less than 1.

[0070] Step 3: Extract the two-way travel time (T) of the reflection signal. This is a one-dimensional vector that can be directly extracted from the velocity spectrum. If the time resolution of the velocity spectrum is low, it can be extracted from the seismic record after preprocessing (including denoising and regularization). This means extracting the two-way travel time (TWT) of the reflection signal at zero offset.

[0071] Step 4: Perform image enhancement processing on the velocity spectrum. Based on the spatiotemporal distribution characteristics of seismic data, the present invention designs an image enhancement processing method that is adapted to the characteristics of seismic data. This method is specifically manifested as data binarization. The image enhancement processing process of the present invention is achieved through the following sub-steps:

[0072] (1) The mathematical average value d of the velocity spectrum D is calculated by the following formula: m :

[0073]

[0074] Where D represents the obtained seismic recording velocity spectrum, which is an M×N two-dimensional matrix.

[0075] (2) The element value in the statistical velocity spectrum D is greater than the average value d m The number of is defined as L.

[0076] (3) Find the statistical average value d of the velocity spectrum D s ,Right now:

[0077]

[0078] Where α is the image enhancement control factor.

[0079] (4) The velocity spectrum D is greater than or equal to d s The element value of is assigned to 1, which is less than d s The element value of is set to 0, and the velocity spectrum G after image enhancement is obtained:

[0080]

[0081] Where G is an M×N two-dimensional matrix.

[0082] Step 5: Based on the velocity spectrum G after image enhancement, pick up the root mean square velocity V rms and round-trip travel time T rms This step is the new root mean square velocity V used in the present invention. rms Picking strategy. Based on the velocity spectrum G after image enhancement, it is easy to identify a distribution area where the element in the velocity spectrum G is 1. By setting a reasonable time sampling interval, the root mean square velocity V is manually picked using human-computer interaction. rms . When manually picking up the RMS speed V rms When the velocity spectrum G is set to 1, the root mean square velocity V near the left and right boundaries of the distribution area needs to be selected at intervals. rms Value, left and right boundaries are simultaneously obtained V rms (L) vector and T rms (L) vector, where L is the time sampling length.

[0083] Under this RMS speed picking strategy, the picked RMS speed V rms The variance is large, but the deviation is small, which is conducive to the stability of subsequent layer velocity inversion. The traditional RMS velocity picking method is to pick a RMS velocity distribution curve in the velocity spectrum D, which is prone to large deviations and thus easily affects the results of layer velocity inversion.

[0084] Step 6: Based on V rms 、T rms and T, and apply the generalized inversion method to solve for the layer velocity.

[0085] It should be noted that T rms is sampled uniformly along the time axis, while V rms There is a pickup error.

[0086] According to T rms The relationship between the elements in T and the time-distance curve formula of the reflected wave can be listed as follows:

[0087]

[0088] Among them, T rms,1 、T rms,2 、T rms,3 、T rms,L Indicates T rms (L) The round trip travel time of the first, second, third, and Lth sampling points of the vector; V rms,1 、Vrms,2 、V rms,3 、V rms,L Indicates V rms (L) is the root mean square velocity of the first, second, third, and Lth sampling points of the vector; K is the number of underground medium layers, C i is the layer velocity to be solved, i=1,2,3,……,K, namely C1, C2, C K represents the layer velocity of the first layer medium, the second layer medium, and the Kth layer medium, and the time sampling length L is greater than the number of underground medium layers K; T0 = 0, T1 represents the two-way travel time from the surface to the interface between the first layer medium and the second layer medium, T2 represents the two-way travel time from the surface to the interface between the second layer medium and the third layer medium, T K-1 T represents the two-way travel time from the surface to the interface between the K-1th layer medium and the Kth layer medium; rms，1 <T rms,2 <T1<T rms,3 .

[0089] Since the above formula is about Therefore, the system of equations (6) can be written as a matrix equation:

[0090] b=Aα (7)

[0091] Since the time sampling length L is generally greater than the number of reflection interfaces K, the generalized linear inversion technique for solving overdetermined equations is used to estimate the solution of Equation (7). The least squares solution α is:

[0092] α=(A T A) -1 A T b (8)

[0093] Among them, A T represents the transpose of A, (A T A) -1 Indicates A T The inverse of A.

[0094] After obtaining α, since α is a parameter that only includes the layer velocity, the layer velocity of the underground medium can be obtained. When the layer velocity V is known rms The depth can be calculated based on the two-way travel time T.

[0095] Compared with the traditional Dix formula solution method, the inversion solution method adopted in the present invention reduces the influence of the root mean square picking error contained in b on the interval velocity model and also makes the interval velocity solution more stable. rmsThe term is obtained based on a new RMS velocity picking strategy. Even when there is a large error in the RMS velocity picking, a relatively accurate velocity model of the underground medium layer can still be obtained.

[0096] The inter-layer velocity inversion method described in this paper, also known as the geoscope method, effectively utilizes reflection information generated by horizontal reflective interfaces and stably inverts the subsurface inter-layer velocity model. This method incorporates the acquisition of both RMS velocity and two-way travel time information from the reflection signal, and, combined with image enhancement technology, forms a stable and efficient inter-layer velocity inversion process. This method accounts for RMS velocity picking errors, formulates an overdetermined system of equations through denser time sampling, and then uses generalized linear inversion techniques to obtain an estimate of the solution to these equations.

[0097] The following are several specific embodiments of the present invention:

[0098] Example 1:

[0099] In this embodiment, the method for performing underground medium layer velocity inversion using reflected wave information specifically includes the following steps:

[0100] Step 1: Design Figure 1 The layered velocity model shown in Figure 1 increases with depth; the forward modeling method yields the following Figure 2 Surface seismic records for this layered velocity model are shown;

[0101] Step 2: Create a velocity spectrum based on surface earthquake records. The velocity spectrum is as follows: Figure 3 As shown in the figure, the energy of the shallow and deep layers of the velocity spectrum without image enhancement is very unbalanced, making it difficult to extract the accurate root mean square velocity V rms .

[0102] Step 3: Pick up the two-way travel time T of the reflected signal.

[0103] Step 4: Perform image enhancement processing on the velocity spectrum. The processed velocity spectrum is as follows: Figure 4 As shown, a root mean square velocity region with an approximately strip-like distribution can be clearly identified.

[0104] Step 5: Use the root mean square velocity picking strategy to pick the root mean square velocity V from the velocity spectrum after image enhancement processing. rms and round-trip travel time T rms .

[0105] Step 6: Calculate the two-way travel time T and root mean square velocity V based on the picked-up reflected signal rms and round-trip travel time T rms , the generalized inversion method is used to solve the layer velocity. The layer velocity distribution model of the underground medium obtained by inversion is as follows Figure 5As shown, although the root mean square velocity V picked up at this time rms With the exact RMS velocity V rms There is a large error, but the layer velocity inversion result is almost consistent with the designed layer velocity model. This further verifies that the layer velocity inversion method (ground mirror method) proposed in this invention can stably and accurately obtain the layer velocity distribution model of the underground medium.

[0106] Example 2:

[0107] In this embodiment, the method for performing underground medium layer velocity inversion using reflected wave information specifically includes the following steps:

[0108] Step 1: Design Figure 6 The layered velocity model with low-velocity layer or high-velocity layer shown in the figure is obtained by forward modeling method as shown in the figure. Figure 7 Surface seismic records for this layered velocity model are shown;

[0109] Step 2: Create a velocity spectrum based on surface earthquake records. The velocity spectrum is as follows: Figure 8 As shown in the figure, the energy of the shallow and deep layers of the velocity spectrum without image enhancement is very unbalanced. In some areas, almost no signal energy can be seen, making it difficult to pick out the accurate root mean square velocity V rms .

[0110] Step 3: Pick up the two-way travel time T of the reflected signal.

[0111] Step 4: Perform image enhancement processing on the velocity spectrum. The processed velocity spectrum is as follows: Figure 9 As shown, a reasonable RMS velocity picking area can be roughly given.

[0112] Step 5: Use the root mean square velocity picking strategy to pick the root mean square velocity V from the velocity spectrum after image enhancement processing. rms and round-trip travel time T rms .

[0113] Step 6: Calculate the two-way travel time T and root mean square velocity V based on the picked-up reflected signal rms and round-trip travel time T rms , the generalized inversion method is used to solve the layer velocity. The layer velocity distribution model of the underground medium obtained by inversion is as follows Figure 10 As shown, although the root mean square velocity V picked up at this time rms With the exact RMS velocity V rms Although there is a large error, the layer velocity inversion result is almost consistent with the designed layer velocity model. This further shows that the layer velocity inversion method proposed in this invention is still effective for underground medium models containing low-velocity layers or high-velocity layers, and can stably and accurately obtain the layer velocity distribution model of the underground medium.

[0114] Example 3:

[0115] In this embodiment, the method for performing underground medium layer velocity inversion using reflected wave information specifically includes the following steps:

[0116] Step 1: Use Figure 11 The shot gather seismic records collected at a certain location in the Zijingang Campus of Zhejiang University are shown in the figure. Figure 12 The effective reflected wave signal in the actual seismic data is shown.

[0117] Step 2: Create a velocity spectrum based on the effective reflected wave signal. The velocity spectrum is as follows: Figure 13 As shown in the figure, the energy of the shallow and deep layers of the velocity spectrum without image enhancement is very unbalanced, making it difficult to extract the accurate root mean square velocity V rms .

[0118] Step 3: Pick up the two-way travel time T of the reflected signal.

[0119] Step 4: Perform image enhancement processing on the velocity spectrum. The processed velocity spectrum is as follows: Figure 14 As shown, a reasonable RMS velocity picking area can be roughly given.

[0120] Step 5: Use the root mean square velocity picking strategy to pick the root mean square velocity V from the velocity spectrum after image enhancement processing. rms and round-trip travel time T rms .

[0121] Step 6: Calculate the two-way travel time T and root mean square velocity V based on the picked-up reflected signal rms and round-trip travel time T rms , the generalized inversion method is used to solve the layer velocity. The layer velocity distribution model of the underground medium obtained by inversion is as follows Figure 15 The geotechnical engineering investigation results at a certain location on the Zijingang Campus are shown in Table 1, which are provided by Zhejiang Urban Construction and Investigation Research Institute Co., Ltd.

[0122] The final inversion result is roughly consistent with the geotechnical investigation result of a certain place in Zijingang Campus. The first interface of the layer velocity inversion result of the present invention is approximately located at 47m, which corresponds to the interface between the strongly weathered sandstone and the moderately weathered sandstone in the geotechnical investigation result. Taking into account that the data acquisition position of the geotechnical investigation of Zijingang Campus differs from the data acquisition position of this embodiment by about 400 meters, there will be a slight deviation between the layer velocity inversion result of this embodiment and the geotechnical investigation result. The geotechnical investigation results of Zijingang Campus show that the interface depth between the strongly weathered sandstone and the moderately weathered sandstone is 49.4 meters, which is less than 3 meters different from the interface depth of the layer velocity inversion of the present invention, which is in line with the error expectation. The strongly weathered sandstone layer and the layers above it have a large porosity because they have not been consolidated into rocks or have been affected by strong weathering. Therefore, they are approximately manifested as a low-velocity layer as a whole. The consolidation degree of the moderately weathered sandstone layer is better than that of the upper layer, and the layer velocity is correspondingly larger, so a more obvious underground medium velocity interface can be formed. In addition, the results of layer velocity inversion show that there is a thin low-velocity layer beneath the moderately weathered sandstone layer, and the bedrock surface in this area is approximately at 77m.

[0123] Table 1 Geotechnical engineering investigation results at a location on the Zijingang campus

[0124] Layer medium number Layer medium name Depth of the bottom interface of the layer medium (unit: meter) 1 Miscellaneous fill 2.3m 2 silty clay 4.1m 3 silty clay 17.5m 4 silty clay 23.9m 5 clay 27.5m 6 silty clay 33.3m 7 medium sand 35.6m 8 gravel 45.4m 9 Strongly weathered sandstone 49.4m 10 Moderately weathered sandstone 56.0m

[0125] This paper provides a stable and efficient layer velocity inversion method (the geoscope method). This method combines image enhancement technology with generalized linear inversion techniques for seismic data to develop a method for determining the velocity model of underground media layers based on these inversion and image enhancement techniques. This method effectively utilizes the reflected wave information generated by underground reflective interfaces and accounts for the influence of root mean square velocity picking errors during processing, resulting in improved stability.

[0126] Those skilled in the art will understand that the foregoing descriptions are merely preferred embodiments of the invention and are not intended to limit the invention. Although the invention has been described in detail with reference to the foregoing examples, those skilled in the art will still be able to modify the technical solutions described in the foregoing examples or substitute equivalents for some of the technical features therein. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the invention shall be included within the scope of protection of the invention.

Claims

1. A method for inverting underground medium layer velocity using reflected wave information, characterized in that: The following steps are involved: Step 1: Obtain surface seismic records of the underground medium layer to be measured; Step 2: Create a speed spectrum; Step 3: Pick up the two-way travel time T of the reflected signal; Step 4: Use binarization to perform image enhancement on the velocity spectrum; Step 5: Use the RMS velocity picking strategy to pick the RMS velocity V from the velocity spectrum G after image enhancement. rms and round-trip travel time T rms ; The root mean square velocity picking strategy is as follows: based on the velocity spectrum G after image enhancement, identify the distribution area where the element in the velocity spectrum G is 1, set a reasonable time sampling interval, and manually pick the root mean square velocity V by human-computer interaction. rms ; When manually picking up the root mean square speed V rms When the element in the velocity spectrum G is 1, the root mean square velocity V near the left and right boundaries needs to be selected at intervals. rms Value, left and right boundaries are simultaneously obtained V rms (L) vector and T rms (L) vector, L is the number of time sampling points; Step 6: Calculate the two-way travel time T and root mean square velocity V based on the picked-up reflected signal rms and round-trip travel time T rms , using the generalized inversion method to solve the layer velocity; step six is ​​achieved through the following sub-steps: (1) According to T rms The relationship between the elements in T and the time-distance curve formula of the reflected wave are listed as follows: Among them, T rms,1 、T rms,2 、T rms,3 、T rms,L Indicates T rms (L) The round trip travel time of the first, second, third, and Lth sampling points of the vector; V rms,1 、V rms,2 、V rms,3 、V rms,L Indicates V rms (L) is the root mean square velocity of the first, second, third, and Lth sampling points of the vector; K is the number of underground medium layers, C i is the layer velocity to be solved, i=1,2,3,……,K, namely C1, C2, C K represents the layer velocity of the first layer medium, the second layer medium, and the Kth layer medium, and the time sampling length L is greater than the number of underground medium layers K; T0 = 0, T1 represents the two-way travel time from the surface to the interface between the first layer medium and the second layer medium, T2 represents the two-way travel time from the surface to the interface between the second layer medium and the third layer medium, T K-1 T represents the two-way travel time from the surface to the interface between the K-1th layer medium and the Kth layer medium; rms，1 <T rms,2 <T1<T rms,3 ; (2) Since the above formula is about C i 2 The linear equations of equation (1) are written in the form of matrix equations: b=Aα (2) (3) The generalized inversion method for solving overdetermined equations is used to estimate the solution of equation (2). The least squares solution α is: α=(A T A) -1 A T b (3) Among them, A T represents the transpose of A, (A T A) -1 Indicates A T The inverse of A; (4) Estimate α based on the least squares solution and obtain the layer velocity and depth of the underground medium.

2. The method for performing underground medium layer velocity inversion using reflected wave information according to claim 1, characterized in that: In step 2, the velocity spectrum is prepared based on the time-distance curve formula of the seismic reflection wave: Where t(x) represents the two-way travel time of the reflected signal in the seismic record at x offset; t0 represents the two-way travel time of the reflected signal in the seismic record at zero offset, x represents the offset, and v represents the scanning speed; In the process of velocity spectrum preparation, similarity coefficient is used as the criterion for velocity analysis; The similarity coefficient is defined as: Where, x i is the offset of the ith channel, N is the number of channels, λ is the width of the time window, u(t i +j,x i ) represents the amplitude of the seismic data; S represents the value of the similarity coefficient, S = 1 means that the similarity coefficient reaches the maximum value, corresponding to the optimal dynamic correction speed, while in other cases, the similarity coefficient S < 1.

3. The method for performing underground medium layer velocity inversion using reflected wave information according to claim 1, characterized in that: The fourth step is implemented through the following sub-steps: (1) The mathematical average value d of the velocity spectrum D is calculated by the following formula: m : Where D represents the velocity spectrum of the seismic record, which is an M×N two-dimensional matrix; (2) The element value in the statistical velocity spectrum D is greater than the average value d m The number of is defined as P; (3) Find the statistical average value d of the velocity spectrum D s ,Right now: Where, β is the image enhancement control factor; (4) The velocity spectrum D is greater than or equal to d s The element value of is assigned to 1, which is less than d s The element value of is set to 0, and the velocity spectrum G after image enhancement is obtained: Where G is an M×N two-dimensional matrix.

Images