CN113917537A - Earth surface consistency residual static correction calculation method - Google Patents
Earth surface consistency residual static correction calculation method Download PDFInfo
- Publication number
- CN113917537A CN113917537A CN202010666200.6A CN202010666200A CN113917537A CN 113917537 A CN113917537 A CN 113917537A CN 202010666200 A CN202010666200 A CN 202010666200A CN 113917537 A CN113917537 A CN 113917537A
- Authority
- CN
- China
- Prior art keywords
- rdd
- static correction
- residual static
- superposed
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012937 correction Methods 0.000 title claims abstract description 93
- 230000003068 static effect Effects 0.000 title claims abstract description 90
- 238000004364 calculation method Methods 0.000 title claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 50
- 238000000034 method Methods 0.000 claims abstract description 46
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 238000009499 grossing Methods 0.000 claims description 11
- 230000009467 reduction Effects 0.000 claims description 6
- 239000000523 sample Substances 0.000 claims description 5
- 238000000638 solvent extraction Methods 0.000 claims description 3
- 208000022417 sinus histiocytosis with massive lymphadenopathy Diseases 0.000 description 66
- 230000008569 process Effects 0.000 description 17
- 238000006243 chemical reaction Methods 0.000 description 11
- 238000012360 testing method Methods 0.000 description 8
- 238000011161 development Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000013500 data storage Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. for interpretation or for event detection
- G01V1/36—Effecting static or dynamic corrections on records, e.g. correcting spread; Correlating seismic signals; Eliminating effects of unwanted energy
- G01V1/362—Effecting static or dynamic corrections; Stacking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V2210/00—Details of seismic processing or analysis
- G01V2210/50—Corrections or adjustments related to wave propagation
- G01V2210/53—Statics correction, e.g. weathering layer or transformation to a datum
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Remote Sensing (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Acoustics & Sound (AREA)
- Environmental & Geological Engineering (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention provides a method for calculating residual static correction of earth surface consistency, which comprises the following steps: step 1: carrying out data preprocessing, dividing all CMP gather data into a plurality of subset data, and storing; step 2: under the Spark cluster environment, the subset data is used as RDD input, and the RDD of the model road is obtained through calculation; and step 3: performing cross correlation on the model trace and the CMP trace set to obtain a residual static correction value; and 4, step 4: repeating the step 2-3 until the iteration times reach the maximum iteration times; at this time, the final residual static correction amount is output, and the calculation is completed. The earth surface consistency residual static correction calculation method can improve the traditional processing mode of residual static correction block series, and improves the supporting capacity and the processing efficiency of the algorithm on mass data.
Description
Technical Field
The invention relates to the technical field of oilfield development, in particular to a method for calculating residual static correction of surface consistency.
Background
With the progress of seismic exploration technology, especially the development of high-density seismic technology in recent years, the volume of seismic data is larger and larger, and the TB data (TB is TeraByte, the size is 2)40A byte. TB-level data generally refers to data sizes in excess of 240One byte, but less than 250A byte of data. GB used in Table 1, GigaByte, is 2 in size30A byte. 1TB is 210(i.e., 1024) GB) has become a norm, and seismic data processing techniques face the challenge of large data. In the traditional mode, a single-node serial scheme is mostly adopted in a conventional seismic data processing method, the storage use is examined when large-scale data are processed, and the production requirement is difficult to meet in the operation period.
In order to deal with massive data, a commonly adopted scheme is to run a seismic processing algorithm in parallel in a multi-node computer cluster. Traditionally, parallel strategies such as MPI multi-process and the like are mostly adopted in the field of seismic exploration, and the parallel strategies are widely applied to computation-intensive algorithms such as migration, forward modeling and the like, but the MPI-based algorithm is very complex to develop, the optimization and the promotion of the processing efficiency are very difficult, and the MPI-based algorithm is difficult to be widely applied to the parallel development of numerous seismic processing algorithms.
In recent years, big data processing technologies and platforms have been rapidly developed under data and technology driving. Hadoop is currently the most popular open source big data platform. The Hadoop realizes the support of mass data storage and calculation through a distributed file system (HDFS) and a distributed parallel computing framework (MapReduce). Hadoop has the advantages of high reliability, high expansibility, high efficiency, high fault tolerance and the like, so that Hadoop is applied and developed in the field of geophysical. Researchers use a MapReduce framework to realize some signal processing and migration algorithms, and efficiency improvement of mass seismic data processing is achieved. However, MapReduce also has some disadvantages, such as simple framework abstraction, and the distribution and synchronization of data heavily depend on HDFS, so that the efficiency obtained when processing algorithms such as iteration is not ideal.
The presence of Spark compensates for these deficiencies. Spark is a high-speed and general big data calculation processing engine, supports various calculation modes such as batch processing and interactive processing, and provides superior characteristics of high expansibility, high storage efficiency and memory calculation for processing massive and diverse data. In the aspect of development of a parallel program for a multi-node cluster, compared with an MPI (message passing interface) parallel strategy, Spark provides a simpler and more efficient interface, and complex operation details such as control of cluster nodes, data distribution, message transmission among nodes, reading and writing of temporary files and the like are set to be transparent to developers, so that algorithm development based on Spark only needs to pay attention to the algorithm, namely an input mode, a processing flow, type conversion, intermediate data caching and the like of data, and development convenience is higher.
Spark is the evolution and improvement of Hadoop, and the parallel computing performance is greatly improved. Spark proposes a Distributed abstract memory data structure, namely an elastic Distributed data set (RDD), to implement memory-based computation on a large cluster, so as to solve the shortcomings of Hadoop. RDDs are a set of read-only, partitionable, distributed data sets that have fault tolerance mechanisms and can be operated in parallel. RDD is only an abstraction of a dataset and does not contain a real data volume. The RDD has the characteristic of automatically switching the storage of the memory and the data storage of the magnetic disk, and can cache the data to the external storage such as the magnetic disk and the like when the memory is insufficient in the calculation process.
The RDD is a core data structure of Spark, and the scheduling order of Spark is formed through the dependency relationship of the RDD. The dependency relationship between the RDDs is constructed by RDD operators. RDD provides two types of operators: transformation and Action. The Transformation operator may obtain a new RDD, such as generating a new RDD from a data source, generating a new RDD from an RDD, etc., and the map, the flatMap, the groupByKey, the join, etc. used later are all the operators. The Action operator obtains results of other data types, such as reduce operator. The Transformation operator does not trigger the execution of the calculation, but when the Action is submitted, the generation of a job (job) is triggered for calculation. Each job is a collection of a series of RDD Transformation operations.
The residual static correction of the earth surface consistency is an important link in the seismic data processing flow. The surface consistency residual static correction algorithm is based on surface consistency hypothesis conditions. The assumption of surface consistency means that the static correction values of the shot point and the receiving point are only related to the positions of the shot point and the receiving point, and are not related to the propagation path of the wave. This assumption applied to the residual static correction algorithm can be expressed as: and the shot point residual static correction values of all the tracks of the common shot point gather are consistent, and the receiving point residual static correction values of all the tracks of the common receiving point gather are consistent in the same way. The basic mathematical model can be expressed as:
wherein Δ Ti,jThe total residual static correction value of a certain seismic channel is obtained, the shot point of the channel is located at the position i, and the receiving point is located at the position j; siAnd GjRespectively obtaining the shot point residual static correction value and the receiving point residual static correction value of the channel; mkThe residual static correction value related to the common central point is called a construction item, and the position k of the common central point can be obtained from the position i of the shot point and the position j of the receiving point; rkIs the correlation coefficient of the residual dynamic correction amount, hi,jThe track offset is the track residual motion correction value calculated from the track offset and the reference offset.
Each residual static correction term in the formula (1) can be conveniently solved by adopting a maximum superposition energy method. The objective function of the algorithm is as follows:
wherein, F (T) is a super gather formed by a shot gather or a receiving point gather, G (T) is a super gather formed by a corresponding model trace or an overlapped trace, P (delta T) is energy corrected based on residual static time difference delta T, and T is a travel time variable. The formula is further developed as follows,
the former term of the formula is a constant, and the latter term is the cross-correlation between the shot gather or the receiving point gather and the model trace, which is the key to the implementation of the method. And obtaining the residual static correction value of the shot point or the receiving point by picking up the maximum energy of the cross correlation.
The implementation flow of the surface consistency residual static correction algorithm is shown in fig. 1. The basic flow is a multiple iteration process, and each iteration can be divided into two main steps. The first step is to obtain a model channel, that is, each gather is overlapped to obtain an initial overlapped channel for the inputted common center point gather, and a static correction value obtained by the last iteration is needed during the overlapping. And then smoothing the adjacent initial superposed tracks to obtain the model track. The second step is to find the remaining static correction amount for the cross-correlation. Namely, each track in the common central point track set is cross-correlated with a model track formed by the track set, and the residual static correction value is obtained through maximum energy pickup.
However, the earth surface consistency residual static correction algorithm is a multi-iteration algorithm, and in each iteration, firstly, various types of intermediate data are generated, including residual static correction values of a model channel, a shot point, a demodulator probe and the like; smoothing of the model channel requires using data of adjacent channels; and thirdly, the input seismic data are used for a plurality of times. The algorithm of the type is difficult to parallelize, and related patents or documents are not found at present to discuss the parallelization method of the algorithm.
At present, in the actual production of seismic processing, a serial mode is adopted. When mass data is processed, manual block division and block division serial modes are mostly adopted. On one hand, the mode has low efficiency and long treatment period; on the other hand, the block processing results in poor boundary processing effect between blocks, and the overall consistency is affected.
In the application No.: CN201811303414.6, chinese patent application, relates to an improved residual static correction method for surface consistency, comprising the following steps: respectively labeling the multiple pieces of data, and performing residual static correction processing on the single piece of data according to respective acquisition grids; recording the heading characters of each shot point and wave detection point for calculating the residual static correction value by using two idle heads respectively; gridding a plurality of pieces of data according to a uniform grid; covering shot and demodulator probe heading characters used for calculating the residual static correction value with the sequence number edited again before; and carrying out the continuous fusion residual static correction processing to realize the continuous fusion residual static correction processing. The patent creates the residual static correction processing step aiming at the continuous data, and does not create the details of the residual static correction algorithm, so that the traditional scheme is adopted when the final continuous fusion residual static correction processing step of the patent is carried out, and the problem of processing the mass data in the patent still exists.
In the application No.: chinese patent application 201610172753.X relates to a residual static correction method and system. The method can comprise the following steps: performing dynamic correction processing on the common midpoint gather data to obtain dynamically corrected common midpoint gather data and a superposition profile; performing first residual static correction processing on the basis of the common midpoint gather data after dynamic correction to obtain common midpoint gather data and a superposition profile after the first residual static correction; generating an external model of the common midpoint gather data based on the common midpoint gather data and the superposition profile after the first residual static correction; and performing second residual static correction processing on the basis of the external model of the common midpoint gather data to obtain the common midpoint gather data and the superposition profile after the second residual static correction. The innovation point of the patent lies in that the processing steps of the residual static correction are optimized, and the specific algorithm and implementation details of the residual static correction are not involved, so that the difficulty in efficiency and effect is still faced when the residual static correction is processed every time when massive data is processed.
Therefore, a new earth surface consistency residual static correction calculation method is invented, and the technical problems are solved.
Disclosure of Invention
The invention aims to provide a method for calculating the residual static correction of the earth surface consistency, which can support parallel processing of an algorithm on mass data and improve the processing efficiency.
The object of the invention can be achieved by the following technical measures: the method for calculating the residual static correction of the earth surface consistency comprises the following steps: step 1: carrying out data preprocessing, dividing all CMP gather data into a plurality of subset data, and storing; step 2: under the Spark cluster environment, the subset data is used as RDD input, and the RDD of the model road is obtained through calculation; and step 3: performing cross correlation on the model trace and the CMP trace set to obtain a residual static correction value; and 4, step 4: repeating the step 2-3 until the iteration times reach the maximum iteration times; at this time, the final residual static correction amount is output, and the calculation is completed.
The object of the invention can also be achieved by the following technical measures:
in step 1, parameters for calculating the residual static correction of the earth surface consistency are set, wherein the parameters comprise maximum iteration times, the number of time windows, the range of each time window and smooth parameters.
The step 2 comprises the following steps:
step 2 a: partitioning of RDD Each CMP gather is marked with a crossline number and a vertical line number;
in the step 2a, applying a map operator to perform superposition processing on each CMP gather in the RDD to obtain a superposed trace RDD; each gather gets an overlay track, which is marked with its crossline number and the wale number.
Step 2b comprises:
(2b.1) selecting a plurality of adjacent superposed channels for each superposed channel in the RDD according to the set smooth parameters, and dividing the adjacent superposed channels and the superposed channels into a group; thus obtaining an overlapped channel group RDD;
(2b.2) applying a map operator to carry out smooth calculation on the superposed track group RDD to obtain a model track RDD; the model trace is marked with the label of the superimposed trace before smoothing.
In step 2b.1, the specific grouping method is as follows: for a certain superposed track, assuming that the mark is k, selecting the superposed track nearby the certain superposed track according to the smooth parameter, and marking as a set { k }; for each superposed channel j in the { k }, judging whether the smooth calculation of the superposed channel j needs to use the superposed channel k according to the smooth parameters; if necessary, copying one part of the superposed channel k, and adding a mark j to the copied superposed channel, namely adding a superposed channel which is marked as two-stage marks, wherein j and k are respectively marked;
all the superposed tracks are subjected to the grouping step to obtain a new RDD, and all the superposed tracks are two-level marks; and (4) dividing the tracks with the same first-level identification into the same group by using a groupByKey operator, namely completing grouping.
The step 3 comprises the following steps:
(3.1) matching the model track RDD with the CMP gather RDD, utilizing join operator processing to combine the model track and the CMP gather with the same identification into a new data set, and marking by using the identification to obtain a matched data set RDD containing the model track data and the CMP gather data;
(3.2) applying a map operator, performing cross correlation on each track in the CMP track set in the matched data set and the model track according to a time window, and converting the RDD of the matched data set into a cross correlation result set RDD;
(3.3) applying the shot residual static correction value to the CMP gather;
and (3.4) repeating the steps (3.2) - (3.4) for the CMP gather to which the shot point residual static correction value is applied, and calculating to obtain the demodulator probe residual static correction value.
In step 3.2, further processing to obtain a shot point residual static correction value, wherein the specific processing steps are as follows:
firstly, a reduceByKey operator is applied to carry out stipulation on cross-correlation results, the cross-correlation results with the same identification are merged and are still marked by the identification; at the moment, the cross-correlation result set RDD is converted into a reduction result RDD;
and secondly, applying a map operator to pick up energy of the reduction result to obtain the shot point residual static correction value.
According to the method for calculating the residual static correction of the earth surface consistency, algorithm details and implementation optimization are performed on the residual static correction algorithm of the earth surface consistency by adopting a Spark parallelization technology, so that the overall processing of the residual static correction algorithm of the earth surface consistency on mass data can be supported, and the processing efficiency of the algorithm is improved. Based on the method and the device, the traditional processing mode of residual static correction block serialization can be improved, and the supporting capability and the processing efficiency of the algorithm on mass data are improved. Compared with the prior art, the method for calculating the residual static correction of the earth surface consistency has the following advantages:
firstly, the method comprises the following steps: the invention supports the integral calculation of TB-level mass data, and avoids the boundary problem caused by block calculation
Secondly, the method comprises the following steps: the invention realizes the parallelization processing of the earth surface consistency algorithm and greatly improves the efficiency.
Drawings
FIG. 1 is a flow chart of an implementation of a method of calculating residual static correction for surface consistency according to the present invention;
FIG. 2 is a flowchart illustrating a process for calculating the RDD of the model trace from the RDD of the CMP gather in accordance with an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a process of cross-correlating model trace RDD and CMP gather RDD to obtain a shot point residual static correction value according to an embodiment of the present invention;
FIG. 4 is a histogram of actual data test results in an embodiment of the present invention.
Detailed Description
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
The earth surface consistency residual static correction calculation method comprises the following steps:
step 1: and (4) data preprocessing, namely cutting all CMP gather data into a plurality of subset data and storing the subset data in a distributed file system. And setting parameters for calculating the residual static correction of the earth surface consistency, including maximum iteration times, the number of time windows, the range of each time window, smoothing parameters and the like.
Step 2: in the Spark cluster environment, the subset data is used as the RDD input, and the RDD of the model trace is calculated. The specific process is as follows:
(1) partitioning of RDD Each CMP gather is marked with a crossline number and a vertical line number. And applying a map operator to perform superposition processing on each CMP gather in the RDD to obtain the RDD of the superposed trace. Each gather may result in an overlay track, which is marked with its crossline number and the wale number. As shown in the first conversion of the leftmost RDD in fig. 2.
(2) And smoothing the superposed tracks. Further, the step includes the steps of:
and (2.1) selecting a plurality of adjacent superposed channels for each superposed channel in the superposed channel RDD according to the set smoothing parameters, and dividing the adjacent superposed channels and the superposed channels into a group. This results in an overlapping track set RDD. As shown in the second RDD conversion process in fig. 2.
Further, the grouping method specifically comprises the following steps: and for a certain superposed track, assuming that the mark is k, selecting the superposed tracks nearby the certain superposed track according to the smoothing parameters, and marking as a set { k }. And for each superposed track j in the { k }, judging whether the superposed track k is needed to be used for the smooth calculation of the superposed track j according to the smooth parameters. If necessary, the overlap track k is copied by one copy, and a mark j is added to the copied overlap track, namely a new overlap track is added, and the mark is marked as two-level marks which are j and k respectively.
And performing the grouping steps on all the superposed tracks to obtain a new RDD, wherein all the superposed tracks are two-level identifiers (the first-level identifier indicates the grouping of the superposed tracks, and the second-level identifier indicates the original position of the superposed tracks). And (4) dividing the tracks with the same first-level identification into the same group by using a groupByKey operator, namely completing grouping.
And (2.2) applying a map operator to carry out smooth calculation on the superposed track group RDD to obtain the model track RDD. The model trace is marked with the label of the superimposed trace before smoothing. As shown in the third RDD conversion process in fig. 2.
And step 3: and performing cross correlation on the model trace and the CMP trace gather to obtain a residual static correction value. The specific process is as follows:
and (3.1) matching the model trace RDD with the CMP trace gather RDD, utilizing join operator processing to combine the model trace and the CMP trace gather with the same identification into a new data set, and marking by using the identification to obtain a matched data set RDD containing the model trace data and the CMP trace gather data. As shown in the first RDD conversion process in fig. 3.
And (3.2) applying a map operator, cross-correlating each track in the CMP track set in the matched data set with the model track according to time windows, and converting the RDD of the matched data set into a cross-correlation result set RDD. As shown in the second RDD conversion process in fig. 3.
Further processing can obtain the shot point residual static correction value. The specific treatment steps are as follows:
the method comprises the steps of firstly, using a reduceByKey operator to perform stipulation on cross-correlation results, merging the cross-correlation results with the same identification, and still using the identification for marking. At this point, the set of cross-correlation results RDD is converted to the reduction results RDD. As shown in the third RDD conversion process in fig. 3.
And secondly, applying a map operator to pick up energy of the reduction result to obtain the shot point residual static correction value. As shown in the fourth RDD conversion process in fig. 3.
(3.3) applying the shot residual static correction to the CMP gather.
And (3.4) repeating the steps (3.2) - (3.4) for the CMP gather to which the shot point residual static correction value is applied, and calculating to obtain the demodulator probe residual static correction value.
And 4, step 4: and (4) repeating the step (2-3) until the iteration times reach the maximum iteration times. And outputting the final residual static correction value at the moment, and finishing the algorithm.
In an embodiment of the invention, several sets of data with different sizes are selected for testing the efficiency of the computing method and the system of the invention. The test environment is a cluster environment of 60 computing nodes. Each node is configured with 32-core Intel Xeon series CPUs, and the main frequency is 2.5 GHz. The memory of each node is configured to be 128 GB. The test of the present invention covers 6 sets of data of different sizes from 50G to 6TB as shown in table 1. Based on the data in table 1, the number of tracks processed per second and the data amount processed per minute in each set of tests can be calculated, and as shown in the bar chart of fig. 4, it can be found that the residual static correction algorithm based on the earth surface consistency of Spark is good for supporting the processing of TB-level large data, and has quite high processing efficiency, the number of tracks processed per second is above 33000 tracks, and the average data amount processed per minute exceeds 30 GB.
Table 1 test results table for data of different sizes
FIG. 1 is a process for implementing a residual static correction algorithm for surface consistency; in the figure, rectangular boxes represent specific operations in the flow, and oblique parallelogram boxes represent data resulting from the operations. The arrow from the operation box to the data box indicates that the operation gets the data, and the arrow from the data box to the operation box indicates that the data participates in the processing of the operation.
Fig. 2 is a process flow of obtaining the model trace RDD by calculating the CMP trace gather RDD in an embodiment of the present invention, which corresponds to step 2 in the summary of the invention. The large rectangular boxes in the figure represent the data RDD, and each small rectangular box in the large rectangular boxes represents one piece of partition data of the RDD, and the figure only illustrates three partitions by way of example. The middle of the two large rectangular boxes represents the conversion of RDD, English marks the operator used for conversion at the lower part, and Chinese marks the corresponding algorithm process at the upper part. The arrows identify the use of the data during the conversion process.
Fig. 3 is a process flow of cross-correlation between the model trace RDD and the CMP gather RDD to obtain the remaining shot point static correction value in an embodiment of the present invention, which corresponds to (3.1) - (3.3) in step 3 in the summary of the invention. The legend is synonymous with fig. 2.
FIG. 4 is a histogram of actual data test results in one embodiment of the present invention, corresponding to the test results data in Table 1.
Claims (8)
1. The method for calculating the residual static correction of the earth surface consistency is characterized by comprising the following steps of:
step 1: carrying out data preprocessing, dividing all CMP gather data into a plurality of subset data, and storing;
step 2: under the Spark cluster environment, the subset data is used as RDD input, and the RDD of the model road is obtained through calculation;
and step 3: performing cross correlation on the model trace and the CMP trace set to obtain a residual static correction value;
and 4, step 4: repeating the step 2-3 until the iteration times reach the maximum iteration times; at this time, the final residual static correction amount is output, and the calculation is completed.
2. The method according to claim 1, wherein in step 1, parameters for performing the surface consistency residual static correction calculation are set, including maximum iteration number, number of time windows, range of each time window, and smoothing parameters.
3. The method of calculating residual static correction of surface consistency according to claim 1, wherein step 2 comprises:
step 2 a: partitioning of RDD Each CMP gather is marked with a crossline number and a vertical line number;
and step 2 b: and smoothing the superposed tracks.
4. The method of calculating residual static correction for surface consistency according to claim 3, wherein in step 2a, a map operator is applied to perform superposition processing on each CMP gather in the RDD to obtain a superposed trace RDD; each gather gets an overlay track, which is marked with its crossline number and the wale number.
5. The method of calculating residual static correction of surface consistency according to claim 3, wherein step 2b comprises:
(2b.1) selecting a plurality of adjacent superposed channels for each superposed channel in the RDD according to the set smooth parameters, and dividing the adjacent superposed channels and the superposed channels into a group; thus obtaining an overlapped channel group RDD;
(2b.2) applying a map operator to carry out smooth calculation on the superposed track group RDD to obtain a model track RDD; the model trace is marked with the label of the superimposed trace before smoothing.
6. The method for calculating the residual static correction of the earth surface consistency according to the claim 5, characterized in that in the step 2b.1, the grouping method is as follows: for a certain superposed track, assuming that the mark is k, selecting the superposed track nearby the certain superposed track according to the smooth parameter, and marking as a set { k }; for each superposed channel j in the { k }, judging whether the smooth calculation of the superposed channel j needs to use the superposed channel k according to the smooth parameters; if necessary, copying one part of the superposed channel k, and adding a mark j to the copied superposed channel, namely adding a superposed channel which is marked as two-stage marks, wherein j and k are respectively marked;
all the superposed tracks are subjected to the grouping step to obtain a new RDD, and all the superposed tracks are two-level marks; and (4) dividing the tracks with the same first-level identification into the same group by using a groupByKey operator, namely completing grouping.
7. The method of calculating residual static correction of surface consistency according to claim 1, wherein step 3 comprises:
(3.1) matching the model track RDD with the CMP gather RDD, utilizing join operator processing to combine the model track and the CMP gather with the same identification into a new data set, and marking by using the identification to obtain a matched data set RDD containing the model track data and the CMP gather data;
(3.2) applying a map operator, performing cross correlation on each track in the CMP track set in the matched data set and the model track according to a time window, and converting the RDD of the matched data set into a cross correlation result set RDD;
(3.3) applying the shot residual static correction value to the CMP gather;
and (3.4) repeating the steps (3.2) - (3.4) for the CMP gather to which the shot point residual static correction value is applied, and calculating to obtain the demodulator probe residual static correction value.
8. The earth surface consistency residual static correction calculation method according to claim 7, characterized in that in step 3.2, shot point residual static correction values are obtained by further processing, and the specific processing steps are as follows:
firstly, a reduceByKey operator is applied to carry out stipulation on cross-correlation results, the cross-correlation results with the same identification are merged and are still marked by the identification; at the moment, the cross-correlation result set RDD is converted into a reduction result RDD;
and secondly, applying a map operator to pick up energy of the reduction result to obtain the shot point residual static correction value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010666200.6A CN113917537A (en) | 2020-07-10 | 2020-07-10 | Earth surface consistency residual static correction calculation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010666200.6A CN113917537A (en) | 2020-07-10 | 2020-07-10 | Earth surface consistency residual static correction calculation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113917537A true CN113917537A (en) | 2022-01-11 |
Family
ID=79231403
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010666200.6A Pending CN113917537A (en) | 2020-07-10 | 2020-07-10 | Earth surface consistency residual static correction calculation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113917537A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199106A (en) * | 2014-09-11 | 2014-12-10 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | Seismic data residual static correction method based on ant colony algorithm |
| WO2017106127A1 (en) * | 2015-12-18 | 2017-06-22 | Saudi Arabian Oil Company | Automated near surface analysis by surface-consistent refraction methods |
| WO2017209787A1 (en) * | 2016-06-02 | 2017-12-07 | Shell Oil Company | Method of processing a geospatial dataset |
-
2020
- 2020-07-10 CN CN202010666200.6A patent/CN113917537A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104199106A (en) * | 2014-09-11 | 2014-12-10 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | Seismic data residual static correction method based on ant colony algorithm |
| WO2017106127A1 (en) * | 2015-12-18 | 2017-06-22 | Saudi Arabian Oil Company | Automated near surface analysis by surface-consistent refraction methods |
| WO2017209787A1 (en) * | 2016-06-02 | 2017-12-07 | Shell Oil Company | Method of processing a geospatial dataset |
Non-Patent Citations (1)
| Title |
|---|
| 廉西猛等: "基于Spark的地震地表一致性处理算法并行开发", 地球物理学进展, pages 1 - 6 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108459909B (en) | Multi-body separation grid overlapping method and system suitable for parallel processing | |
| US8259104B2 (en) | Domain decomposition by the advancing-partition method for parallel unstructured grid generation | |
| CN102565854B (en) | Mass data GPU (graphics processing unit) wave equation reverse time migration imaging method | |
| Wehr et al. | Parallel kd-tree construction on the gpu with an adaptive split and sort strategy | |
| CN103838852A (en) | Method for rapidly searching for multi-block structured grid docking relations | |
| CN109657197A (en) | A kind of pre-stack depth migration calculation method and system | |
| CN110264392B (en) | Strong connection graph detection method based on multiple GPUs | |
| CN105205052A (en) | Method and device for mining data | |
| CN116755876A (en) | Large model hybrid parallel training acceleration method and system | |
| JP7069897B2 (en) | Learning identification device and learning identification method | |
| CN103675908A (en) | Wave-equation reverse-time migration imaging method for mass-data graphic processing unit | |
| Zhao et al. | Progressive discrete domains for implicit surface reconstruction | |
| CN110187386B (en) | DTW seismic body attribute analysis method for automatically and rapidly identifying geological structure | |
| Senthilkumar et al. | An efficient FP-Growth based association rule mining algorithm using Hadoop MapReduce | |
| CN103235879A (en) | Bi-phase medium parametric inversion method based on niche master-slave parallel genetic algorithm | |
| CN103472481B (en) | A kind of GPU of utilization carries out the method that angular-trace gather is extracted in reverse-time migration | |
| CN106780747B (en) | A kind of method that Fast Segmentation CFD calculates grid | |
| CN113917537A (en) | Earth surface consistency residual static correction calculation method | |
| CN105738949B (en) | A kind of nine bin uniformity method for parallel processing for time-lapse seismic | |
| Wu | Data association rules mining method based on improved apriori algorithm | |
| CN116595302A (en) | Method for realizing irregular matrix SPMV on GPU, electronic equipment and medium | |
| Kelly et al. | Quad-tree construction on the gpu: A hybrid cpu-gpu approach | |
| JP7040168B2 (en) | Learning identification device and learning identification method | |
| CN109960750A (en) | A kind of parallel figure division methods based on label probability of spreading | |
| Cheng | Efficient level set methods for constructing wavefronts in three spatial dimensions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220111 |