CN118210810A - Compression and decompression method, device, electronic equipment and medium for logging while drilling data - Google Patents
Compression and decompression method, device, electronic equipment and medium for logging while drilling data Download PDFInfo
- Publication number
- CN118210810A CN118210810A CN202410640615.4A CN202410640615A CN118210810A CN 118210810 A CN118210810 A CN 118210810A CN 202410640615 A CN202410640615 A CN 202410640615A CN 118210810 A CN118210810 A CN 118210810A
- Authority
- CN
- China
- Prior art keywords
- logging
- data
- drilling
- drilling data
- chart
- 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.)
- Granted
Links
- 238000005553 drilling Methods 0.000 title claims abstract description 221
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000007906 compression Methods 0.000 title claims abstract description 43
- 230000006835 compression Effects 0.000 title claims abstract description 41
- 230000006837 decompression Effects 0.000 title claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 60
- 238000013139 quantization Methods 0.000 claims abstract description 49
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000011176 pooling Methods 0.000 claims description 33
- 239000013598 vector Substances 0.000 claims description 30
- 230000006870 function Effects 0.000 claims description 24
- 230000004913 activation Effects 0.000 claims description 17
- 238000013507 mapping Methods 0.000 claims description 16
- 230000009466 transformation Effects 0.000 claims description 13
- 230000002441 reversible effect Effects 0.000 claims description 6
- 230000000717 retained effect Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000013144 data compression Methods 0.000 abstract description 6
- 238000003062 neural network model Methods 0.000 abstract description 5
- 230000002829 reductive effect Effects 0.000 abstract description 5
- 230000008569 process Effects 0.000 description 9
- 230000007246 mechanism Effects 0.000 description 5
- 239000000306 component Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000011551 log transformation method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2282—Tablespace storage structures; Management thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2228—Indexing structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/23—Updating
- G06F16/2365—Ensuring data consistency and integrity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Security & Cryptography (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a compression and decompression method, a device, electronic equipment and a medium of logging while drilling data, belonging to the technical field of data compression, wherein the method comprises the following steps: carrying out nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart; compressing the logging-while-drilling data chart based on a target convolution self-encoder with complete training to obtain a logging-while-drilling floating point number; decompressing the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart; performing inverse quantization processing on the reconstructed data chart to obtain the decompressed logging while drilling data; compressing the logging-while-drilling data chart through a target convolution self-encoder, and extracting features of the logging-while-drilling data chart by utilizing a convolution neural network model, so that the compression ratio of the logging-while-drilling data is greatly improved; because the target decoder and the target convolutional self-encoder are in reciprocal relationship, the reconstruction error rate of the logging while drilling data can be reduced.
Description
Technical Field
The present invention relates to the field of data compression technologies, and in particular, to a method, an apparatus, an electronic device, and a medium for compressing and decompressing logging while drilling data.
Background
In the exploration and development process of oil and gas fields, well logging is required after well drilling, i.e. instruments are put into the well by cables so as to know the oil and gas conditions of stratum. But in some cases, such as: highly deviated wells, even horizontal wells, with a slope exceeding 65 degrees are difficult to place with wireline tools, so Logging While Drilling (LWD) is currently used to produce so as to know various parameters in the formation in real time and adjust the drill bit trajectory in real time.
Currently, LWD data is typically transmitted uphole (at the surface) by mud pulse telemetry. Such techniques are typically limited to data transmission rates (bandwidths) on the order of only a few bits per second. Because LWD imaging sensors typically generate data at a much higher rate than is transmitted to the surface, borehole images are typically processed from data stored in memory only after the tool is removed from the borehole. That is, it is difficult to process data of the LWD imaging sensor in real time. In the prior art, in order to perform real-time processing on logging while drilling data, a KL-AEO compression algorithm is provided for data compression, but the achievable compression ratio is low, the requirement of real-time transmission cannot be met, the reconstruction error rate is high, and the data cannot be accurately restored.
Therefore, in the prior art, in the compression process of logging while drilling data, the problems of low compression ratio and high reconstruction error rate exist.
Disclosure of Invention
In view of the foregoing, it is necessary to provide a method, an apparatus, an electronic device and a medium for compressing and decompressing logging while drilling data, so as to solve the problems of low compression ratio and high reconstruction error rate in the compression process of logging while drilling data in the prior art.
In order to solve the above problems, the present invention provides a method for compressing and decompressing logging while drilling data, comprising:
Carrying out nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart;
compressing the logging-while-drilling data chart based on a target convolution self-encoder with complete training to obtain a logging-while-drilling floating point number;
decompressing the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
and performing inverse quantization processing on the reconstructed data graph to obtain the decompressed logging while drilling data.
In one possible implementation, the nonlinear quantization processing is performed on the logging while drilling data to obtain a logging while drilling data chart, including:
intercepting and grouping logging while drilling data according to a preset time window size to obtain a plurality of logging while drilling data vectors;
Sequentially carrying out logarithmic transformation, linear quantization and truncation on the data in each logging-while-drilling data vector to obtain a plurality of corresponding quantized logging-while-drilling data vectors;
a log while drilling data chart is constructed from the quantized log while drilling data vector.
In one possible implementation, the log transformation, linear quantization and truncation are sequentially performed on the data in each logging-while-drilling data vector to obtain a corresponding plurality of quantized logging-while-drilling data vectors, including:
Adding the data in the logging while drilling data vector with a preset basic value, and carrying out logarithmic transformation to obtain logarithmic data;
According to a preset maximum value and a preset minimum value, combining a linear mapping formula to perform linear processing on the logarithmic data to obtain linear quantized data;
Only an integer portion of the linear quantized data is retained, truncated data is obtained, and a quantized logging-while-drilling data vector is determined from the truncated data.
In one possible implementation, a target convolutional self-encoder includes a convolutional layer, a pooling layer, a first activation function, and a self-attention module; compressing the logging while drilling data chart based on a target convolution self-encoder with complete training to obtain a logging while drilling floating point number, comprising:
extracting local features of the logging while drilling data chart through the convolution layer;
Correcting the local characteristics through a self-attention module, and capturing the long-distance dependence of the logging-while-drilling data chart;
Combining the local features through a pooling layer to obtain pooling features;
Mapping the logging while drilling floating point number according to the pooling characteristic and the long-distance dependence relation through a first activation function.
In one possible implementation, the target decoder includes a deconvolution layer, a pooling layer, and a second activation function; decompression processing is carried out on the logging while drilling floating point number based on a complete training target decoder, and a reconstructed data chart is obtained, wherein the method comprises the following steps:
Up-sampling the logging while drilling floating point number through the reverse pooling layer to obtain reverse pooling logging while drilling data;
Performing image restoration processing on the reverse-pooling logging-while-drilling data through a deconvolution layer to obtain a deconvolution logging-while-drilling image;
A reconstructed data map is mapped from the deconvoluted logging while drilling image by a second activation function.
In one possible implementation, the inverse quantization processing is performed on the reconstructed data chart to obtain the pressure-relieving logging-while-drilling data, including:
and carrying out inverse mapping, inverse digital transformation and rounding processing on the data in the reconstructed data chart in sequence to obtain the pressure-relieving logging-while-drilling data.
In order to solve the above problems, the present invention further provides a compression and decompression device for logging while drilling data, including:
the nonlinear quantization processing module is used for carrying out nonlinear quantization processing on the logging while drilling data to obtain a logging while drilling data chart;
The compression module is used for compressing the logging while drilling data chart based on a target convolution self-encoder with complete training to obtain the logging while drilling floating point number;
The decompression module is used for performing decompression processing on the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
And the inverse quantization processing module is used for performing inverse quantization processing on the reconstructed data chart to obtain the decompressed logging while drilling data.
In one possible implementation, the apparatus includes: an encoder and a decoder, which can be separated and connected by a signal;
the encoder comprises a nonlinear quantization processing module and a compression module;
the decoder comprises a decompression module and an inverse quantization processing module.
In order to solve the above-mentioned problems, the present invention also provides an electronic device including a memory and a processor, wherein,
A memory for storing a program;
A processor coupled to the memory for executing programs stored in the memory to implement steps in the method of compression and decompression of logging while drilling data as described above.
To solve the above problems, the present invention also provides a computer readable storage medium storing a computer readable program or instructions which, when executed by a processor, enable the implementation of the steps in the method of compression and decompression of logging while drilling data as described above.
The beneficial effects of adopting the embodiment are as follows: the invention provides a compression and decompression method of logging while drilling data, which enables the logging while drilling data to have the characteristics of images while keeping the characteristics of the data by carrying out nonlinear quantization processing on the logging while drilling data; compressing the logging while drilling data chart through a target convolution self-encoder, and extracting features of the logging while drilling data chart by utilizing a convolution neural network model to obtain a logging while drilling floating point number as a data result, so that the compression ratio of the logging while drilling data is greatly improved; in addition, because the target decoder and the target convolution self-encoder are in a reciprocal relationship, the consistency of the decompressed logging-while-drilling data and the original logging-while-drilling data can be ensured, namely, the reconstruction error rate of the logging-while-drilling data is reduced.
Drawings
FIG. 1 is a flow chart of an embodiment of a method for compressing and decompressing logging while drilling data according to the present invention;
FIG. 2 is a graph of results of one embodiment of a chart of logging while drilling data provided by the present invention;
FIG. 3 is a graph of results obtained for one embodiment of the present invention for floating point number while drilling;
FIG. 4 is a graph of results of one embodiment of a graph of obtained reconstructed data provided by the present invention;
FIG. 5 is a block diagram illustrating an embodiment of a compression and decompression apparatus for logging while drilling data according to the present invention;
fig. 6 is a block diagram of an embodiment of an electronic device according to the present invention.
Detailed Description
The following detailed description of preferred embodiments of the invention is made in connection with the accompanying drawings, which form a part hereof, and together with the description of the embodiments of the invention, are used to explain the principles of the invention and are not intended to limit the scope of the invention.
In order to solve the above problems, the present invention provides a method, an apparatus, an electronic device and a medium for compressing and decompressing logging while drilling data, which are described in detail below.
Fig. 1 is a flow chart of an embodiment of a method for compressing and decompressing logging while drilling data according to the present invention, as shown in fig. 1, the method for compressing and decompressing logging while drilling data includes:
S101: carrying out nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart;
s102: compressing the logging-while-drilling data chart based on a target convolution self-encoder with complete training to obtain a logging-while-drilling floating point number;
S103: decompressing the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
S104: and performing inverse quantization processing on the reconstructed data graph to obtain the decompressed logging while drilling data.
It should be noted that the logging while drilling data exists in a data form, in a specific embodiment, the depth of the well is taken as a reference point, all data corresponding to the depth in the well is obtained, and the data of the logging while drilling instrument at different depths are collected and collated to obtain a corresponding data table. It is obvious that other parameters may be collected and counted when the measurement targets are different, and the specific parameter type is not limited herein.
In the embodiment, by carrying out nonlinear quantization processing on the logging while drilling data, the logging while drilling data can have the characteristics of images while the characteristics of the data are reserved; compressing the logging while drilling data chart through a target convolution self-encoder, and extracting features of the logging while drilling data chart by utilizing a convolution neural network model to obtain a logging while drilling floating point number as a data result, so that the compression ratio of the logging while drilling data is greatly improved; in addition, because the target decoder and the target convolution self-encoder are in a reciprocal relationship, the consistency of the decompressed logging-while-drilling data and the original logging-while-drilling data can be ensured, namely, the reconstruction error rate of the logging-while-drilling data is reduced.
It should be noted that, the reciprocal relationship between the target decoder and the target convolutional encoder means: for data compressed by the target convolutional self-encoder, the target decoder can generate a corresponding decoding rule according to the encoding rule, so as to decode the corresponding data, wherein the decoded data has consistency with the data before the compression, and ideally, the decoded data and the data are completely consistent.
In a preferred embodiment, in S101, in order to perform nonlinear quantization processing on logging while drilling data, a logging while drilling data chart is obtained, as shown in fig. 2, fig. 2 is a result chart of an embodiment of obtaining a logging while drilling data chart provided by the present invention, including:
s201: intercepting and grouping logging while drilling data according to a preset time window size to obtain a plurality of logging while drilling data vectors;
s202: sequentially carrying out logarithmic transformation, linear quantization and truncation on the data in each logging-while-drilling data vector to obtain a plurality of corresponding quantized logging-while-drilling data vectors;
S203: a log while drilling data chart is constructed from the quantized log while drilling data vector.
In this embodiment, by performing vectorization processing on logging while drilling data, discrete data can be changed into directional and sized vectors, so as to determine position information of each data on an image; then, carrying out logarithmic transformation, linear quantization and truncation processing on the logging-while-drilling data vector to realize standardized processing on the logging-while-drilling data, and unifying data quantization standards to avoid the problem of data disorder; finally, constructing a logging while drilling data chart according to the quantized logging while drilling data vector, and converting discrete logging while drilling data into the logging while drilling data chart with image characteristics.
In a specific embodiment, the preset time window size is generally 3 to 5 time units, and the time window size can be adaptively adjusted according to actual needs. By intercepting and grouping the logging while drilling data according to the set time window size, the characteristics of the time sequence in the time dimension can be reserved so as to facilitate the subsequent formation of the logging while drilling data in the form of vectors.
In S302, in order to obtain a quantized logging-while-drilling data vector, firstly, adding data in the logging-while-drilling data vector to a preset base value, and performing logarithmic transformation to obtain log-taking data; then, according to a preset maximum value and a preset minimum value, combining a linear mapping formula to perform linear processing on the logarithmic data to obtain linear quantized data; finally, only the integer portion of the linear quantized data is retained, truncated data is obtained, and a quantized logging-while-drilling data vector is determined from the truncated data.
In a specific embodiment, the preset base value is preferably 1, and because the data in the logging-while-drilling data vector is non-negative, the log transformation can be performed by adding 1 to the data in the logging-while-drilling data vector, so that the occurrence of negative numbers in log data can be avoided.
In particular, log data are takenThe method comprises the following steps:
Wherein, Is data in a logging while drilling data vector,/>Is a preset basic value.
Obviously, in other embodiments, the preset basic value may be selected to be other values according to actual needs, which is not limited herein.
In a specific embodiment, since the distribution of logging while drilling data is certain, the preset maximum value preferentially selects an integer between 10 and 12, the preset minimum value preferentially selects an integer between 0 and 2, and in order to convert log-taking data into a range between 0 and 255, a linear mapping formula is set as follows:
Wherein, For linear quantized data,/>For a preset maximum value,/>A preset minimum value.
In other embodiments, when the value of the logarithmic data needs to be controlled in other ranges, the method can be implemented by adaptively adjusting parameters in the linear mapping formula, for example: when the preset basic value is in the range of 0-100, the linear mapping formula is modified as follows: And so on, no further description is given.
In a specific embodiment, the logging while drilling data with time sequence data property is processed through nonlinear quantization to remove redundant information, extract important features and save computing resources, so that efficient compression and storage of complex time sequence data are realized.
As a preferred embodiment, in S102, the target convolutional self-encoder includes a convolutional layer, a pooling layer, a first activation function and a self-attention module, in order to perform compression processing on a logging-while-drilling data chart based on the target convolutional self-encoder with complete training to obtain a logging-while-drilling floating point number, and implement data compression, as shown in fig. 3, fig. 3 is a result diagram of an embodiment of obtaining the logging-while-drilling floating point number provided by the present invention, including:
S301: extracting local features of the logging while drilling data chart through the convolution layer;
S302: correcting the local characteristics through a self-attention module, and capturing the long-distance dependence of the logging-while-drilling data chart;
S303: combining the local features through a pooling layer to obtain pooling features;
s304: mapping the logging while drilling floating point number according to the pooling characteristic and the long-distance dependence relation through a first activation function.
In this embodiment, the target convolutional self-encoder is essentially a neural network model capable of processing an image and performing deep learning prediction, local feature extraction is performed through a convolutional layer, and features are combined by a pooling layer, so that representative features of an image layer are obtained; and finally, mapping the logging while drilling floating point number according to the pooled characteristic and the long-distance dependency relationship by a first activation function, creatively extracting the logging while drilling floating point number as the characteristic of the logging while drilling data chart, and greatly improving the compression ratio of the logging while drilling data due to less memory occupied by the logging while drilling floating point number.
In a specific embodiment, the target convolution self-encoder can extract features in time and space dimensions simultaneously, semantic information is contained more abundantly, so that the logging while drilling floating point number comprises information in multiple dimensions, and information loss in the data compression process is reduced.
In a particular embodiment, features of logging-while-drilling data are extracted in the temporal and spatial dimensions, respectively, using a self-attention based convolution operation. The encoder portion of the convolutional self-encoder consists of a convolutional layer, a pooling layer, and an activation function. The convolution layer is a core component in the convolution self-encoder, and features extraction is performed on input data through convolution operation. The encoder is used for compressing input data and extracting key features, capturing local features of logging-while-drilling data in space through convolution operation, and reducing the number of parameters through pooling operation. In addition, by introducing a self-attention mechanism into the convolution self-encoder model, the extracted features are modified, and modeling and learning of the relevance between the features are realized.
The convolution operation utilizes a convolution kernel to slide and calculate feature mapping on the input data, and effectively captures local features of the input data in a manner of sharing weights and local connections. Each convolution kernel may learn different features, thereby improving the model's ability to characterize the input data.
The self-attention module is then embedded in the encoder structure so that the features are adjusted by the self-attention mechanism after passing through the convolutional layer. The attention mechanism may help the model better capture long range dependencies in the input data at the encoder stage, thereby improving the global consistency of the feature representation. By calculating the attention weight, the model can automatically learn the importance among all positions in the input sequence, is beneficial to highlighting key features and inhibiting irrelevant information, and improves the feature extraction efficiency.
The calculation formula of the self-attention weight is as follows:
Wherein Q is a query matrix, K is a key matrix, V is a value matrix, dk is a dimension of a key, and a softmax function is used for normalization to obtain attention weight. Output is the Output of the self-attention mechanism.
The pooling layer follows the convolution layer and reduces the size and number of feature maps by downsampling while preserving the main features of the logging while drilling data in spatial locations. Common pooling operations include maximum pooling and average pooling, which can help the model be insensitive to positional changes, reduce overfitting, and reduce model complexity, and speed training.
The ReLU (RECTIFIED LINEAR Unit) function is selected as the activation function, and has simple calculation form and nonlinear characteristic, so that the gradient vanishing problem can be effectively solved, and the network is easier to optimize. The ReLU function sets all negative input values to zero, retains positive number parts, introduces nonlinearity, and increases the expression capacity of the model.
The formula for the ReLU function is as follows:
Where x is the input data.
Potential feature representations, i.e., compressed logging-while-drilling floating point numbers, are obtained through a self-attention-based convolutional automatic encoder.
In addition, because the logging while drilling floating point number is convenient to transmit, the data exchange on the well and underground can be well realized.
In one embodiment, the logging-while-drilling floating point number is transmitted uphole over a channel using a 32-bit binary code.
Further, after the downhole logging while drilling data is processed by the convolutional self-encoder and transmitted to the well, in order to decompress the logging while drilling floating point number and timely obtain downhole data information, in S103, the target decoder includes a deconvolution layer, a pooling layer and a second activation function, and in order to decompress the logging while drilling floating point number based on the well-trained target decoder, a reconstructed data chart is obtained, as shown in fig. 4, and fig. 4 is a result chart of an embodiment of the obtained reconstructed data chart provided by the present invention, where the method includes:
s401: up-sampling the logging while drilling floating point number through the reverse pooling layer to obtain reverse pooling logging while drilling data;
S402: performing image restoration processing on the reverse-pooling logging-while-drilling data through a deconvolution layer to obtain a deconvolution logging-while-drilling image;
S403: a reconstructed data map is mapped from the deconvoluted logging while drilling image by a second activation function.
Obviously, contrary to the data compression process of the target convolution self-encoder, in order to perform information expansion on the logging-while-drilling floating point number, the data reduction is performed on the logging-while-drilling floating point number through the deconvolution layer, the anti-pooling layer and the second activation function in sequence, so that image information which can be represented by the logging-while-drilling floating point number, namely a reconstruction data chart, is expanded.
It should be noted that, the target decoder is similar to the target convolutional self-encoder, and is also a neural network model for deep learning, and can rapidly and accurately inverse-solve the reconstructed data chart according to the characteristic information.
Further, in S104, corresponding to S101, since the nonlinear quantization process is performed before the code learning, in order to obtain the most original data for the staff to perform the data processing, the inverse quantization process is also required to perform the inverse quantization process on the reconstructed data chart, so that the pressure-relief logging while drilling data that can be directly used finally can be obtained.
Specifically, contrary to the flow of nonlinear quantization processing, inverse mapping, inverse digital transformation and rounding processing are sequentially performed on the data in the reconstructed data chart to obtain the pressure-relieving logging-while-drilling data.
In the process of performing the inverse quantization, it is necessary to inverse-quantize the data subjected to the nonlinear quantization. First, the data is converted back into the original range by inverse mapping based on quantized data, minimum and maximum values (the same maximum value as set downhole and uphole, and generally not modified, e.g., downhole or uphole, and requiring mutual signaling to remain uniform), this step being to recover the pre-quantized data range. Then, the data restored to the original range is subjected to logarithmic inverse transformation, the logarithmic transformation data is restored to the original data through an exponential function, and meanwhile, an offset value is subtracted (generally, 0.8-1.0 is taken). Finally, rounding the recovered data and reserving a decimal place to obtain a final dequantization result, and finally obtaining the dequantization-processed pressure-relieving logging-while-drilling data.
In one embodiment, the bias value is preferably 0.8.
In the way, the extraction and adjustment of the time sequence data potential characteristics by combining the target convolution self-encoder and the target decoder are introduced into a self-attention mechanism to optimize the extracted potential characteristics, so that the error of the compressed data reconstruction is reduced. In addition, by mapping continuous data to discrete space, the method reduces the calculation cost and improves the compression efficiency, and effectively solves the problems of larger reconstruction error and non-ideal compression rate in the prior art.
In order to solve the above-mentioned problems, the present invention further provides a device for compressing and decompressing logging-while-drilling data, as shown in fig. 5, fig. 5 is a block diagram of an embodiment of the device for compressing and decompressing logging-while-drilling data, where the device 500 for compressing and decompressing logging-while-drilling data includes:
the nonlinear quantization processing module 501 is configured to perform nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart;
The compression module 502 is configured to perform compression processing on the logging while drilling data chart based on a target convolution self-encoder with complete training, so as to obtain a logging while drilling floating point number;
The decompression module 503 is configured to decompress the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
and the inverse quantization processing module 504 is configured to perform inverse quantization processing on the reconstructed data chart to obtain the pressure-relieving logging while drilling data.
It should be noted that, in a specific embodiment, to better adapt to the requirement of downhole data processing, the compression and decompression device 500 of logging while drilling data specifically includes two parts, namely an encoder and a decoder, which can be separated and connected by a signal;
wherein the encoder comprises a nonlinear quantization processing module 501 and a compression module 502;
the decoder comprises a decompression module 503 and an inverse quantization processing module 504.
As shown in fig. 6, the present invention further provides an electronic device 600 accordingly. The electronic device 600 comprises a processor 601, a memory 602 and a display 603. Fig. 6 shows only a portion of the components of the electronic device 600, but it should be understood that not all of the illustrated components are required to be implemented and that more or fewer components may be implemented instead.
Processor 601, in some embodiments, may be a central processing unit (Central Processing Unit, CPU), microprocessor, or other data processing chip that runs program code or process data stored in memory 602, such as the compression and decompression methods of logging while drilling data of the present invention.
In some embodiments, the processor 601 may be a single server or a group of servers. The server farm may be centralized or distributed. In some embodiments, the processor 601 may be local or remote. In some embodiments, the processor 601 may be implemented in a cloud platform. In an embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multiple cloud, or the like, or any combination thereof.
The memory 602 may be an internal storage unit of the electronic device 600 in some embodiments, such as a hard disk or memory of the electronic device 600. The memory 602 may also be an external storage device of the electronic device 600 in other embodiments, such as a plug-in hard disk provided on the electronic device 600, a smart memory card (SMART MEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD), or the like.
Further, the memory 602 may also include both internal storage units and external storage devices of the electronic device 600. The memory 602 is used for storing application software and various types of data for installing the electronic device 600.
The display 603 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like in some embodiments. The display 603 is used for displaying information of the electronic device 600 and for displaying a visualized user interface. The components 601-603 of the electronic device 600 communicate with each other via a system bus.
In one embodiment, when the processor 601 executes the compression and decompression program of logging while drilling data in the memory 602, the following steps may be implemented:
Carrying out nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart;
compressing the logging-while-drilling data chart based on a target convolution self-encoder with complete training to obtain a logging-while-drilling floating point number;
decompressing the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
and performing inverse quantization processing on the reconstructed data graph to obtain the decompressed logging while drilling data.
It should be understood that: processor 601, when executing the compression and decompression programs for logging while drilling data in memory 602, may perform other functions in addition to the above, as described above with particular reference to corresponding method embodiments.
Further, the type of the electronic device 600 is not particularly limited, and the electronic device 600 may be a portable electronic device such as a mobile phone, a tablet computer, a personal digital assistant (personaldigital assistant, PDA), a wearable device, a laptop (laptop), etc. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices that carry IOS, android, microsoft or other operating systems. The portable electronic device described above may also be other portable electronic devices, such as a laptop computer (laptop) or the like having a touch-sensitive surface, e.g. a touch panel. It should also be appreciated that in other embodiments of the invention, the electronic device 600 may not be a portable electronic device, but rather a desktop computer having a touch-sensitive surface (e.g., a touch panel).
Correspondingly, the embodiment of the invention also provides a computer readable storage medium, and the computer readable storage medium is used for storing a computer readable program or instruction, and when the program or instruction is executed by a processor, the steps or functions in the compression and decompression methods of logging while drilling data provided by the embodiments of the method can be realized.
Those skilled in the art will appreciate that all or part of the flow of the methods of the embodiments described above may be accomplished by way of a computer program stored in a computer readable storage medium to instruct related hardware (e.g., a processor, a controller, etc.). The computer readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory.
The compression and decompression methods, devices, electronic equipment and media of logging while drilling data provided by the invention are described in detail, and specific examples are applied to illustrate the principles and embodiments of the invention, and the description of the examples is only used to help understand the method and core ideas of the invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in light of the ideas of the present invention, the present description should not be construed as limiting the present invention.
Claims (10)
1. A method for compressing and decompressing logging-while-drilling data, comprising:
Carrying out nonlinear quantization processing on logging while drilling data to obtain a logging while drilling data chart;
Compressing the logging while drilling data chart based on a target convolution self-encoder with complete training to obtain a logging while drilling floating point number;
Decompressing the logging while drilling floating point number based on a complete training target decoder to obtain a reconstructed data chart;
And performing inverse quantization processing on the reconstructed data graph to obtain the pressure-relieving logging-while-drilling data.
2. The method for compressing and decompressing logging while drilling data according to claim 1, wherein the performing nonlinear quantization on the logging while drilling data to obtain a logging while drilling data graph comprises:
intercepting and grouping logging while drilling data according to a preset time window size to obtain a plurality of logging while drilling data vectors;
Sequentially carrying out logarithmic transformation, linear quantization and truncation on the data in each logging-while-drilling data vector to obtain a plurality of corresponding quantized logging-while-drilling data vectors;
and constructing a logging-while-drilling data chart according to the quantized logging-while-drilling data vector.
3. The method for compressing and decompressing logging-while-drilling data according to claim 2, wherein the sequentially performing logarithmic transformation, linear quantization and truncation on the data in each logging-while-drilling data vector to obtain a corresponding plurality of quantized logging-while-drilling data vectors comprises:
adding the data in the logging while drilling data vector with a preset basic value, and carrying out logarithmic transformation to obtain logarithmic data;
According to a preset maximum value and a preset minimum value, combining a linear mapping formula to perform linear processing on the logarithmic data to obtain linear quantized data;
only an integer portion of the linear quantized data is retained, truncated data is obtained, and the quantized logging while drilling data vector is determined from the truncated data.
4. The method of compression and decompression of logging while drilling data of claim 1, wherein the target convolutional self-encoder comprises a convolutional layer, a pooling layer, a first activation function, and a self-attention module; the target convolution self-encoder based on complete training compresses the logging while drilling data chart to obtain the logging while drilling floating point number, which comprises the following steps:
extracting local features of the logging while drilling data chart through the convolution layer;
Correcting the local characteristics through the self-attention module, and capturing long-distance dependence of the logging-while-drilling data chart;
combining the local features through the pooling layer to obtain pooling features;
mapping out the logging while drilling floating point number according to the pooling feature and the long-distance dependence relation through the first activation function.
5. The method of compression and decompression of logging while drilling data according to claim 1, wherein the target decoder comprises a deconvolution layer, a pooling layer, and a second activation function; the training completion-based target decoder performs decompression processing on the logging while drilling floating point number to obtain a reconstructed data chart, and the method comprises the following steps:
Up-sampling the logging while drilling floating point number through the reverse pooling layer to obtain reverse pooling logging while drilling data;
Performing image restoration processing on the reverse-pooling logging-while-drilling data through the deconvolution layer to obtain a deconvolution logging-while-drilling image;
The reconstructed data map is mapped from the deconvolution logging while drilling image by the second activation function.
6. The method for compressing and decompressing logging while drilling data according to claim 1, wherein said performing inverse quantization on the reconstructed data map to obtain decompressed logging while drilling data comprises:
And carrying out inverse mapping, inverse digital transformation and rounding processing on the data in the reconstruction data chart in sequence to obtain the pressure-relieving logging-while-drilling data.
7. A compression and decompression apparatus for logging while drilling data, comprising:
the nonlinear quantization processing module is used for carrying out nonlinear quantization processing on the logging while drilling data to obtain a logging while drilling data chart;
The compression module is used for carrying out compression processing on the logging while drilling data chart based on a target convolution self-encoder with complete training to obtain logging while drilling floating point numbers, wherein the target convolution self-encoder comprises a convolution layer, a pooling layer, a first activation function and a self-attention module;
The decompression module is used for performing decompression processing on the logging while drilling floating point number based on a complete training target decoder to obtain a reconstruction data chart;
And the inverse quantization processing module is used for performing inverse quantization processing on the reconstructed data chart to obtain the pressure-relieving logging while drilling data.
8. The compression and decompression apparatus for logging while drilling data of claim 7, comprising: an encoder and a decoder, the encoder and the decoder being detachable and connected by a signal;
Wherein the encoder comprises the nonlinear quantization processing module and the compression module;
the decoder comprises the decompression module and the inverse quantization processing module.
9. An electronic device comprising a memory and a processor, wherein,
The memory is used for storing programs;
the processor, coupled to the memory, is configured to execute the program stored in the memory to implement the steps in the method of compression and decompression of logging while drilling data as claimed in any one of the preceding claims 1 to 6.
10. A computer readable storage medium storing a computer readable program or instructions which when executed by a processor is capable of carrying out the steps of the method of compression and decompression of logging while drilling data as claimed in any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410640615.4A CN118210810B (en) | 2024-05-22 | 2024-05-22 | Compression and decompression method, device, electronic equipment and medium for logging while drilling data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410640615.4A CN118210810B (en) | 2024-05-22 | 2024-05-22 | Compression and decompression method, device, electronic equipment and medium for logging while drilling data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118210810A true CN118210810A (en) | 2024-06-18 |
| CN118210810B CN118210810B (en) | 2024-08-27 |
Family
ID=91448973
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410640615.4A Active CN118210810B (en) | 2024-05-22 | 2024-05-22 | Compression and decompression method, device, electronic equipment and medium for logging while drilling data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118210810B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2857317A1 (en) * | 2004-04-02 | 2005-10-13 | Schlumberger Canada Limited | Data compression methods and systems |
| US20090192711A1 (en) * | 2008-01-25 | 2009-07-30 | Pathfinder Energy Services, Inc. | Data compression method for use in downhole applications |
| US20140229113A1 (en) * | 2013-02-14 | 2014-08-14 | Schlumberger Technology Corporation | Compression of Multidimensional Sonic Data |
| CN104090307A (en) * | 2014-07-18 | 2014-10-08 | 中国海洋石油总公司 | Method and system for processing nuclear magnetic resonance logging-while-drilling data |
| CN106130562A (en) * | 2016-06-15 | 2016-11-16 | 中国石油集团渤海钻探工程有限公司 | Logging-while-drilling data real-time compression method and decompression method |
| CN111852458A (en) * | 2019-04-10 | 2020-10-30 | 中国石油化工股份有限公司 | Method and system for transmitting logging data by imaging while drilling |
| CN115749734A (en) * | 2022-12-08 | 2023-03-07 | 电子科技大学 | Method for learning geological structure on line by ultrasonic imaging logging instrument while drilling |
-
2024
- 2024-05-22 CN CN202410640615.4A patent/CN118210810B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2857317A1 (en) * | 2004-04-02 | 2005-10-13 | Schlumberger Canada Limited | Data compression methods and systems |
| US20090192711A1 (en) * | 2008-01-25 | 2009-07-30 | Pathfinder Energy Services, Inc. | Data compression method for use in downhole applications |
| US20140229113A1 (en) * | 2013-02-14 | 2014-08-14 | Schlumberger Technology Corporation | Compression of Multidimensional Sonic Data |
| CN104090307A (en) * | 2014-07-18 | 2014-10-08 | 中国海洋石油总公司 | Method and system for processing nuclear magnetic resonance logging-while-drilling data |
| CN106130562A (en) * | 2016-06-15 | 2016-11-16 | 中国石油集团渤海钻探工程有限公司 | Logging-while-drilling data real-time compression method and decompression method |
| CN111852458A (en) * | 2019-04-10 | 2020-10-30 | 中国石油化工股份有限公司 | Method and system for transmitting logging data by imaging while drilling |
| CN115749734A (en) * | 2022-12-08 | 2023-03-07 | 电子科技大学 | Method for learning geological structure on line by ultrasonic imaging logging instrument while drilling |
Non-Patent Citations (1)
| Title |
|---|
| 杨诚: "声波测井井下数据压缩算法研究", 《中国优秀硕士学位论文全文数据库》（电子期刊）, 15 January 2024 (2024-01-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118210810B (en) | 2024-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112001914B (en) | Depth image complement method and device | |
| CN111898696B (en) | Pseudo tag and tag prediction model generation method, device, medium and equipment | |
| CN111696112B (en) | Automatic image cutting method and system, electronic equipment and storage medium | |
| US7058200B2 (en) | Method for the prior monitoring of the detectability of a watermarking signal | |
| JP2004177965A (en) | System and method for coding data | |
| CN109074512A (en) | Image is compressed using neural network | |
| CN110046116B (en) | Tensor filling method, device, equipment and storage medium | |
| CN111832227A (en) | Shale gas saturation determination method, device and equipment based on deep learning | |
| CN112785493A (en) | Model training method, style migration method, device, equipment and storage medium | |
| CN112819199A (en) | Precipitation prediction method, device, equipment and storage medium | |
| CN112598039A (en) | Method for acquiring positive sample in NLP classification field and related equipment | |
| CN112418291A (en) | Distillation method, device, equipment and storage medium applied to BERT model | |
| CN118210810B (en) | Compression and decompression method, device, electronic equipment and medium for logging while drilling data | |
| CN113038134B (en) | Picture processing method, intelligent terminal and storage medium | |
| CN116257488B (en) | Geotechnical engineering investigation big data archiving method, device, electronic equipment and medium | |
| CN116821113A (en) | Time sequence data missing value processing method and device, computer equipment and storage medium | |
| CN116740078A (en) | Image segmentation processing method, device, equipment and medium | |
| CN115601235A (en) | Image super-resolution network training method, device, equipment and storage medium | |
| CN115619950A (en) | Three-dimensional geological modeling method based on deep learning | |
| CN113298265B (en) | Heterogeneous sensor potential correlation learning method based on deep learning | |
| CN115273110A (en) | Text recognition model deployment method, device, equipment and storage medium based on TensorRT | |
| CN118155270B (en) | Model training method, face recognition method and related equipment | |
| CN112749560A (en) | Address text processing method, device and equipment and computer storage medium | |
| Maleika et al. | Lossless compression method for digital terrain model of seabed shape | |
| CN117058668B (en) | Three-dimensional model face reduction evaluation method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |