Reverse time migration imaging method and device based on lossless compression and medium

Abstract

The invention provides a reverse time migration imaging method and device based on lossless compression and a medium, and belongs to the field of seismic exploration data processing. The method comprises the following steps: determining a check point configuration strategy, and generating a check point list; performing forward delay on the shot point wave field to obtain an inspection point wave field; compressing and storing the obtained check point wave field; the receiving point wave field is subjected to reverse continuation to reach a corresponding check point, and the shot point wave field of the check point is read; the shot point wave field of the read check point is decompressed, and shot point wave fields of two time slices of the check point are obtained through calculation; and calculating to obtain a reverse time migration imaging field. According to the method, the IO bandwidth is improved while the storage requirement of the hard disk is reduced, so that the calculation efficiency of overall reverse time migration is improved.

Description

Technical Field

[0001] This invention belongs to the field of seismic exploration data processing, specifically relating to a method, apparatus, and medium for reverse time migration imaging based on lossless compression. Background Technology

[0002] Pre-stack reverse time migration is a two-way wave migration imaging method that uses the cross-correlation between the forward and reverse propagating wave fields to obtain the position of the reflecting interface. Reverse time migration can use complex wave field information such as gyratory waves and rhombic waves for imaging. Compared with other migration methods, it has the advantages of high accuracy, no approximation processing, and no limitations on tilt angle or drastic velocity changes, thus achieving better imaging results.

[0003] However, reverse time migration suffers from low computational efficiency, large storage requirements, and heavy reliance on the initial velocity model. In recent years, with in-depth research and improved computing power, pre-stack reverse time migration and its corresponding full waveform inversion have made significant progress. New wavefield reconstruction methods and storage strategies have greatly reduced the requirements for computer hardware, making high-resolution imaging of 3D seismic data feasible.

[0004] The checkpoint-based reconstruction method sets a checkpoint at regular intervals during the propagation process and stores the wavefield of two time slices at that point. Starting from the checkpoint, the wavefield at any future time can be obtained by FDTD calculation based on the stored two time slice wavefields.

[0005] The two time-slice wavefields of the checkpoints need to be stored on the hard drive. The actual calculation involves I / O operations, which are time-consuming due to the large size of the time-slice data. Furthermore, saving multiple checkpoint wavefields for each offset calculation consumes a significant amount of hard drive storage space. Therefore, optimizing the storage size of the checkpoint wavefields and accelerating the wavefield read / write bandwidth are urgent problems that need to be solved. Summary of the Invention

[0006] The purpose of this invention is to solve the problems existing in the prior art and provide a method, device and medium for reverse time offset imaging based on lossless compression, which reduces hard disk storage requirements and increases IO bandwidth, thereby improving the overall computational efficiency of reverse time offset.

[0007] This invention is achieved through the following technical solution:

[0008] A first aspect of the present invention provides a reverse time migration imaging method based on lossless compression, comprising:

[0009] Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list;

[0010] Step 2: Forwardly extend the wave field at the shot point to obtain the wave field at the checkpoint;

[0011] Step 3: Compress and store the acquired checkpoint wavefield;

[0012] Step 4: Receive the point wave field and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint.

[0013] Step 5: Decompress the shot wave field of the checkpoint and calculate the shot wave field of the two time slices of the checkpoint.

[0014] Step 6: Calculate the reverse time migration imaging field.

[0015] A further improvement of the present invention is that:

[0016] Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list. Specific operations include:

[0017] The total duration of the finite difference calculation, sampling interval, aperture size, and mesh size are determined based on the inverse time offset calculation. Combined with the available memory size, the checkpoint configuration strategy is determined, that is, a checkpoint is set every N steps, and a checkpoint list is generated.

[0018] A further improvement of the present invention is that:

[0019] Step 2: Forwardly delay the shot point wavefield to obtain the checkpoint wavefield. Specific operations include:

[0020] When calculating the reverse time migration, we start from the shot point position and use the finite difference algorithm to extend the wave field outward along the positive time axis. Every N steps of extension, we reach a check point and obtain the shot point wave field of the two time slices before and after the check point.

[0021] Further improvements of the present invention are in:

[0022] Step 3: Compress and store the acquired checkpoint wavefield. Specific operations include:

[0023] Based on the shot point wavefields of the two time slices before and after the checkpoint obtained in step 2, the gradient values ​​of the two shot point wavefields are calculated as follows:

[0024] Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula:

[0025] Δa = ab;

[0026] The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

[0027] A further improvement of the present invention is that:

[0028] Step 4: Receive the point wavefield and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint. Specific operations include:

[0029] After completing the forward finite difference continuation of the shot point wavefield, the wavefield is reversed using finite difference starting from the receiving point. The seismic wavefield is propagated backward along the time axis. Whenever a checkpoint determined in step 1 is reached, the wavefield continuation at the receiving point is paused, and the shot point wavefield at that checkpoint is read from the hard disk.

[0030] A further improvement of the present invention is that:

[0031] Step 5: Decompress the shot wavefield of the checkpoint and calculate the shot wavefield of the two time slices of the checkpoint. Specific operations include:

[0032] Read the shot point wave field of the checkpoint from step 4 into memory, decompress it using the Kanzi algorithm, and obtain the shot point wave field a and gradient value Δa of the checkpoint.

[0033] The shot point wavefield b = a - Δa is obtained by calculating the gradient value, thus obtaining the shot point wavefield of the two time slices of the checkpoint.

[0034] A further improvement of the present invention is that:

[0035] Step 6: Calculate the reverse time migration imaging field. Specific operations include:

[0036] Using the shot wavefields of the two time slices of the checkpoint obtained in step 5, the finite difference algorithm is used to perform forward extension in the time domain to calculate all shot wavefields between the current checkpoint and the next checkpoint.

[0037] The wavefields of all shot points between the current checkpoint and the next checkpoint are cross-correlated with the wavefield of the receiver point to obtain the reverse time migration imaging field.

[0038] A second aspect of the present invention provides a time-lapse imaging device based on lossless compression, comprising:

[0039] The checkpoint list generation module is used to determine the checkpoint configuration strategy and generate a checkpoint list;

[0040] The forward extension module is used to forward extend the shot point wavefield to obtain the checkpoint wavefield.

[0041] The compressed storage module is used to compress and store the acquired checkpoint wavefield.

[0042] The reverse extension module is used to receive the point wave field in reverse extension, reach the corresponding check point, and read the shot point wave field of the check point.

[0043] The decompression module is used to decompress the shot wave field of the checkpoint and calculate the shot wave field of the two time slices of the checkpoint.

[0044] The calculation module is used to calculate the reverse time-shifted imaging field.

[0045] A further improvement of the present invention is that:

[0046] The compressed storage module performs the following operations:

[0047] Based on the shot point wavefields obtained from the two time slices before and after the checkpoint, the gradient values ​​of the two shot point wavefields are calculated as follows:

[0048] Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula:

[0049] Δa = ab;

[0050] The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

[0051] A third aspect of the present invention provides a computer-readable storage medium storing at least one computer-executable program, which, when executed by the computer, causes the computer to perform the steps in the lossless compression-based reverse time-shift imaging method.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] This invention takes advantage of the small change in wave field values ​​between two time slices at the checkpoint. First, the gradient values ​​of the two wave fields are calculated. Then, the Kanzi floating-point lossless compression algorithm is used to compress the gradients of the wave fields of the two time slices as a whole, thereby reducing the size of the wave field snapshot data. This reduces the hard disk storage requirements while increasing IO bandwidth, thus improving the overall calculation efficiency of the reverse time offset. Attached Figure Description

[0054] Figure 1 This is a flowchart of the reverse time migration imaging method based on lossless compression in an embodiment of the present invention;

[0055] Figure 2 Flowchart of wavefield extension at the shot point;

[0056] Figure 3 This is a flowchart of the wavefield extension imaging process at the receiving point. Detailed Implementation

[0057] The present invention will now be described in further detail with reference to the accompanying drawings:

[0058]

Example 1

[0059] like Figure 1 As shown, this embodiment of the invention provides a reverse time migration imaging method based on lossless compression, which mainly includes the following steps:

[0060] Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list;

[0061] Step 2: Forwardly delay the wavefield at the shot point to obtain a snapshot of the wavefield at the checkpoint;

[0062] Step 3: Compress and store the obtained checkpoint wavefield snapshots;

[0063] Step 4: Receive the point wave field and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint.

[0064] Step 5: Decompress the shot wave field of the checkpoint and calculate the shot wave field of the two time slices of the checkpoint.

[0065] Step 6: Calculate the reverse time migration imaging field.

[0066] This invention takes advantage of the small change in the wave field values ​​of the shot points in the two time slices of the checkpoint. First, the gradient values ​​of the two wave fields are calculated. Then, the Kanzi floating-point lossless compression algorithm is used to compress the gradient of the wave field of the shot points in the two time slices as a whole, thereby reducing the size of the wave field snapshot data. While reducing the hard disk storage requirements, it increases the IO bandwidth, thereby improving the overall calculation efficiency of the reverse time offset.

[0067]

Example 2

[0068] Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list. Specific operations include:

[0069] Based on the inverse time offset calculation, the total duration of the finite difference calculation, sampling interval, aperture size, mesh size, etc. are determined. Combined with the available memory size, the checkpoint configuration strategy is determined, that is, a checkpoint is set every N steps, and a checkpoint list is generated.

[0070] It should be understood that user parameters are calculated based on the reverse time offset, including the time domain T, sampling interval, propagation time, aperture size in the XYZ spatial direction, grid size, etc. First, the required space size A, such as hard disk and memory, is determined by the time domain T and the aperture size in the XYZ spatial direction. Then, the number of data blocks B that can be stored under the current hardware configuration is calculated by combining the hard disk size and memory size of the actual cluster hardware configuration. The number of copies of data that need to be stored, C, is calculated. Here, C = A / B, and C is the configuration of the checkpoint.

[0071]

Example 3

[0072] Step 2: Forwardly extend the wave field at the shot point to obtain the wave field at the checkpoint;

[0073] The specific operations include:

[0074] When calculating the reverse time migration, we start from the shot point position (the time corresponding to the bubble point position is t=0) and use the finite difference algorithm to extend the wave field outward along the positive time axis. Every N steps of extension, we reach a check point and obtain the shot point wave field of the two time slices before and after the check point.

[0075]

Example 4

[0076] Step 3: Compress and store the acquired checkpoint wavefield;

[0077] The specific operations include:

[0078] Based on the shot point wavefields of the two time slices before and after the checkpoint obtained in step 2, the gradient values ​​of the two shot point wavefields are calculated as follows:

[0079] Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula:

[0080] Δa = ab;

[0081] The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

[0082] The Kanzi algorithm used in this embodiment for lossless floating-point compression is an existing method and will not be described in detail here.

[0083] like Figure 2 As shown, repeat steps 2 and 3 until the shot point wavefields of all checkpoints in the checkpoint list generated in step 1 are obtained and then compressed and stored.

[0084]

Example 5

[0085] Step 4: Receive the point wavefield and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint. Specific operations include:

[0086] After completing the forward finite difference continuation of the shot point wavefield, the wavefield is reversed using finite difference starting from the receiver point (time t_ = tmax corresponding to the receiver point location). The seismic wavefield is propagated backward along the time axis. Whenever a checkpoint determined in step 1 is reached, the receiver point wavefield continuation is paused, and the shot point wavefield of that checkpoint is read from the hard disk.

[0087]

Example 6

[0088] Step 5: Decompress the shot wavefield of the checkpoint and calculate the shot wavefield of the two time slices of the checkpoint. Specific operations include:

[0089] Read the shot point wave field of the checkpoint from step 4 into memory, decompress it using the Kanzi algorithm, and obtain the shot point wave field a and gradient value Δa of the checkpoint.

[0090] The shot point wavefield b = a - Δa is obtained by calculating the gradient value, thus obtaining the shot point wavefield of the two time slices of the checkpoint.

[0091]

Example 7

[0092] Step 6: Calculate the reverse time migration imaging field. Specific operations include:

[0093] Using the shot wavefields of the two time slices of the checkpoint obtained in step 5, the finite difference algorithm is used to perform forward extension in the time domain to calculate all shot wavefields between the current checkpoint and the next checkpoint.

[0094] The wavefields of all shot points between the current checkpoint and the next checkpoint are cross-correlated with the wavefield of the receiver point to obtain the reverse time migration imaging field.

[0095] like Figure 3 As shown, repeat steps 4 to 6 until all checkpoints in the checkpoint list generated in step 1 have been traversed.

[0096]

Example 8

[0097] This invention provides a lossless compression-based reverse time migration imaging method, comprising:

[0098] The checkpoint list generation module is used to determine the checkpoint configuration strategy and generate a checkpoint list. Specifically, it performs the following operations:

[0099] Based on the inverse time offset calculation, the total duration of the finite difference calculation, sampling interval, aperture size, mesh size, etc. are determined. Combined with the available memory size, the checkpoint configuration strategy is determined, that is, a checkpoint is set every N steps, and a checkpoint list is generated.

[0100] The forward extension module is used for forward extension of the shot point wavefield to obtain the checkpoint wavefield. Specifically, it performs the following operations:

[0101] When calculating the reverse time migration, we start from the shot point position (the time corresponding to the bubble point position is t=0) and use the finite difference algorithm to extend the wave field outward along the positive time axis. Every N steps of extension, we reach a check point and obtain the shot point wave field of the two time slices before and after the check point.

[0102] The compression storage module is used to compress and store the acquired checkpoint wavefield. Specifically, it performs the following operations:

[0103] Based on the shot point wavefields obtained from the two time slices before and after the checkpoint, the gradient values ​​of the two shot point wavefields are calculated as follows:

[0104] Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula:

[0105] Δa = ab;

[0106] The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

[0107] The Kanzi algorithm used in this embodiment for lossless floating-point compression is an existing method and will not be described in detail here.

[0108] The reverse continuation module is used to receive the point wavefield and perform reverse continuation, reaching the corresponding checkpoint and reading the shot point wavefield at that checkpoint. Specifically, it performs the following operations:

[0109] After completing the forward finite difference continuation of the shot point wavefield, the wavefield is reversed using finite difference starting from the receiver point (time t_ = tmax corresponding to the receiver point location). The seismic wavefield is propagated backward along the time axis. Whenever a checkpoint determined in step 1 is reached, the receiver point wavefield continuation is paused, and the shot point wavefield of that checkpoint is read from the hard disk.

[0110] The decompression module is used to decompress the shot wavefield of the read checkpoint and calculate the shot wavefield of the two time slices of the checkpoint. Specifically, it performs the following operations:

[0111] The shot point wavefield of the checkpoint is read into memory and decompressed using the Kanzi algorithm to obtain the shot point wavefield a and gradient value Δa at the checkpoint.

[0112] The shot point wavefield b = a - Δa is obtained by calculating the gradient value, thus obtaining the shot point wavefield of the two time slices of the checkpoint.

[0113] The calculation module is used to calculate the reverse time migration imaging field, and specifically performs the following operations:

[0114] By using the shot wavefields of two time slices of the checkpoint obtained by calculation, and employing the finite difference algorithm to perform forward extension in the time domain, all shot wavefields between the current checkpoint and the next checkpoint are calculated.

[0115] The wavefields of all shot points between the current checkpoint and the next checkpoint are cross-correlated with the wavefield of the receiver point to obtain the reverse time migration imaging field.

[0116] This invention takes advantage of the small change in wave field values ​​between two time slices at the checkpoint. First, the gradient values ​​of the two wave fields are calculated. Then, the Kanzi floating-point lossless compression algorithm is used to compress the gradients of the wave fields of the two time slices as a whole, thereby reducing the size of the wave field snapshot data. This reduces the hard disk storage requirements while increasing IO bandwidth, thus improving the overall calculation efficiency of the reverse time offset.

[0117]

Example 9

[0118] This invention provides a computer-readable storage medium storing at least one computer-executable program, which, when executed by the computer, causes the computer to perform the steps in the lossless compression-based reverse time-shift imaging method.

[0119] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0120] The above technical solution is only one embodiment of the present invention. For those skilled in the art, based on the principles disclosed in the present invention, it is easy to make various types of improvements or modifications, and not limited to the technical solutions described in the specific embodiments of the present invention. Therefore, the foregoing description is only a preferred option and is not restrictive.

Claims

1. A reverse time migration imaging method based on lossless compression, characterized in that, include: Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list; Step 2: Forwardly extend the wave field at the shot point to obtain the wave field at the checkpoint; Step 3: Compress and store the acquired checkpoint wavefield; Step 4: Receive the point wave field and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint. Step 5: Decompress the shot wave field of the checkpoint and calculate the shot wave field of the two time slices of the checkpoint. Step 6: Calculate the reverse time migration imaging field.

2. The method according to claim 1, characterized in that, Step 1: Determine the checkpoint configuration strategy and generate a checkpoint list. Specific operations include: The total duration of the finite difference calculation, sampling interval, aperture size, and mesh size are determined based on the inverse time offset calculation. Combined with the available memory size, the checkpoint configuration strategy is determined, that is, a checkpoint is set every N steps, and a checkpoint list is generated.

3. The method according to claim 1, characterized in that, Step 2: Forwardly delay the shot point wavefield to obtain the checkpoint wavefield. Specific operations include: When calculating the reverse time migration, we start from the shot point position and use the finite difference algorithm to extend the wave field outward along the positive time axis. Every N steps of extension, we reach a check point and obtain the shot point wave field of the two time slices before and after the check point.

4. The method according to claim 3, characterized in that, Step 3: Compress and store the acquired checkpoint wavefield. Specific operations include: Based on the shot point wavefields of the two time slices before and after the checkpoint obtained in step 2, the gradient values ​​of the two shot point wavefields are calculated as follows: Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula: Δa = ab; The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

5. The method according to claim 4, characterized in that, Step 4: Receive the point wavefield and perform reverse extension to reach the corresponding checkpoint and read the shot field at that checkpoint. Specific operations include: After completing the forward finite difference continuation of the shot point wavefield, the wavefield is reversed using finite difference starting from the receiving point. The seismic wavefield is propagated backward along the time axis. Whenever a checkpoint determined in step 1 is reached, the wavefield continuation at the receiving point is paused, and the shot point wavefield at that checkpoint is read from the hard disk.

6. The method according to claim 5, characterized in that, Step 5: Decompress the shot wavefield of the checkpoint and calculate the shot wavefield of the two time slices of the checkpoint. Specific operations include: Read the shot point wave field of the checkpoint from step 4 into memory, decompress it using the Kanzi algorithm, and obtain the shot point wave field a and gradient value Δa of the checkpoint. The shot point wavefield b = a - Δa is obtained by calculating the gradient value, thus obtaining the shot point wavefield of the two time slices of the checkpoint.

7. The method according to claim 6, characterized in that, Step 6: Calculate the reverse time migration imaging field. Specific operations include: Using the shot wavefields of the two time slices of the checkpoint obtained in step 5, the finite difference algorithm is used to perform forward extension in the time domain to calculate all shot wavefields between the current checkpoint and the next checkpoint. The wavefields of all shot points between the current checkpoint and the next checkpoint are cross-correlated with the wavefield of the receiver point to obtain the reverse time migration imaging field.

8. A time-shifting imaging device based on lossless compression, characterized in that, include: The checkpoint list generation module is used to determine the checkpoint configuration strategy and generate a checkpoint list; The forward extension module is used to forward extend the shot point wavefield to obtain the checkpoint wavefield. The compression storage module is used to compress and store the acquired checkpoint wavefield. The reverse extension module is used to receive the point wave field in reverse extension, reach the corresponding check point, and read the shot point wave field of the check point. The decompression module is used to decompress the shot wave field of the checkpoint and calculate the shot wave field of the two time slices of the checkpoint. The calculation module is used to calculate the reverse time-shifted imaging field.

9. The apparatus according to claim 8, characterized in that, The compressed storage module specifically performs the following operations: Based on the shot point wavefields obtained from the two time slices before and after the checkpoint, the gradient values ​​of the two shot point wavefields are calculated as follows: Assuming the wavefield at the first shot point is denoted as 'a' and the wavefield at the second shot point is denoted as 'b', the gradient value Δa is calculated using the following formula: Δa = ab; The shot point wave field a and the gradient value Δa are compressed using the Kanzi algorithm to obtain the compressed shot point wave field, which is then stored on the hard disk.

10. A computer-readable storage medium storing at least one computer-executable program, which, when executed by the computer, causes the computer to perform the steps of the lossless compression-based reverse time-shift imaging method as described in any one of claims 1-7.

Images