Method and device for predicting crude oil fluid parameters based on geochemical chromatography and electronic equipment

Through image processing and optical character recognition technology, the geochemical well recording chromatogram is processed, combined with neural network modeling, and the problem of normalization processing of geochemical chromatograms is solved, and the quantitative prediction of crude oil density and viscosity is achieved, which improves the improvement of The accuracy and universality of data analysis are achieved.

CN118914435APending Publication Date: 2024-11-08SHANGHAI BRANCH CHINA OILFIELD SERVICES
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410963787.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The normalization of geothermal evaporation hydrocarbon gas chromatograms is difficult, which leads to different spectrum analysis cycles and scales of different manufacturers, making it difficult to dig out effective information, affecting the quantitative prediction of crude oil fluid parameters.

Method used

Image processing and optical character recognition technology are used to vectorize the geochemical well recording chromatogram, remove irrelevant factors, identify real coordinates, calculate curve area, construct input sequences, and use neural network modeling to predict crude oil density and viscosity.

Benefits of technology

Quantitative prediction of crude oil density and viscosity is achieved, data analysis difficulties under different standards are overcome, and data universality and accuracy are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118914435A_ABST
    Figure CN118914435A_ABST
Patent Text Reader

Abstract

The invention relates to a method and device for predicting crude oil fluid parameters based on geochemical chromatography and electronic equipment, in particular to the technical field of geochemical logging, and the method comprises the steps that image processing and character recognition are conducted on a geochemical logging chromatogram, and actual coordinates of a curve are obtained; performing curve area calculation according to the actual coordinates of the obtained curve; and constructing different input sequences according to the obtained curve areas, preprocessing the input sequences, constructing a fluid parameter prediction model by using the multi-layer sensor, and then operating the fluid parameter prediction model to obtain predicted crude oil fluid parameters. According to the method for predicting the crude oil fluid parameters based on the geochemical logging chromatogram, quantitative prediction of the viscosity and the density of the crude oil can be achieved by designing a specific treatment process on the geochemical logging chromatogram.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of geochemical logging technology, and specifically to a method, device, and electronic equipment for predicting crude oil fluid parameters based on geochemical chromatography, and in particular to a method for predicting crude oil fluid parameters based on geochemical logging chromatograms, and further to a digital analysis method, device, and electronic equipment for predicting crude oil viscosity and density in underground reservoirs using thermal evaporation hydrocarbon gas chromatography based on image processing, optical character recognition technology, and neural network modeling. Background Art

[0002] At present, in the technical application of geochemical logging, thermal evaporation hydrocarbon gas chromatography is used to identify the characteristics of oil and gas layers, such as determining the type of oil and gas layers, evaluating the maturity of source rocks and calculating source rock TOC. The spectra collected by geochemical thermal evaporation hydrocarbon technology during logging contain a large amount of crude oil information. The peaks on the chromatogram represent specific compounds, and the size of the peaks is usually proportional to the concentration of the compound. However, the spectra are currently mainly interpreted by experience to qualitatively analyze the types of crude oil into three types: light oil, medium oil and heavy oil. Quantitative calculation has not been achieved, and the scope and depth of application are limited. The numerical values ​​of crude oil density and viscosity affect the fluid dynamics of the reservoir, and thus affect the flow, recovery rate and production efficiency of oil and gas. In addition, the prediction of crude oil density and viscosity data during drilling is of great guiding significance for the selection of oil exploration testing tools and development tools.

[0003] For example, CN109001355A discloses a logging oil and gas interpretation and discrimination method based on thermal evaporation hydrocarbon gas chromatography spectrum comparison, which establishes a thermal evaporation hydrocarbon gas chromatography standard spectrum database for each preset layer in the completed wells in the drilling area; extracts characteristic spectrum data of the unfinished layers to be interpreted in the drilling area; calculates the similarity distances between the characteristic spectrum data of the layers to be interpreted and each group of thermal evaporation hydrocarbon gas chromatography standard spectrum data of the same layer in the thermal evaporation hydrocarbon gas chromatography standard spectrum database, and the oil quality information and test conclusion information contained in the thermal evaporation hydrocarbon gas chromatography standard spectrum data with the smallest similarity distance in the corresponding layer of the thermal evaporation hydrocarbon gas chromatography standard spectrum database are used as the oil quality conclusion and fluid type conclusion of the corresponding layer to be interpreted.

[0004] However, the current geochemical thermal evaporation hydrocarbon gas chromatogram data are mainly static images. Due to the different acquisition standards of different target areas and different equipment, the normalization processing of geochemical chromatograms has always been very difficult, which is manifested in the different analysis cycles of spectra from different manufacturers and the different scales of displaying spectrum peak heights, making it difficult to mine the effective information contained in the spectra. Summary of the invention

[0005] In view of the problems existing in the prior art, the purpose of the present invention is to provide a method, device and electronic equipment for predicting crude oil fluid parameters based on geochemical chromatography, so as to solve the problem that the normalization processing of geochemical chromatograms is difficult, which is manifested as different analysis cycles of spectra from different manufacturers, different scales for displaying spectrum peak heights, and difficulty in mining the effective information contained in the spectra.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for predicting crude oil fluid parameters based on geochemical logging chromatograms, the method comprising:

[0008] Perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve;

[0009] Calculate the curve area based on the actual coordinates of the obtained curve;

[0010] Different input sequences are constructed according to the obtained curve area. After preprocessing the input sequence, a fluid parameter prediction model is constructed using a multi-layer perceptron. Then, the fluid parameter prediction model is run to obtain the predicted crude oil fluid parameters.

[0011] The method for predicting crude oil fluid parameters based on geochemical logging chromatograms provided by the present invention can realize quantitative prediction of the viscosity and density of crude oil by designing a specific processing process for the geochemical logging chromatograms.

[0012] As a preferred technical solution of the present invention, the image processing includes sequentially performing HSV color space change, irrelevant factor removal, smoothing, edge detection and coordinate axis recognition.

[0013] As a preferred technical solution of the present invention, the HSV color space change includes converting the image into the HSV color space.

[0014] Preferably, the irrelevant factor removal includes applying a threshold operation based on the hue channel range to remove irrelevant parts.

[0015] Preferably, the range of the hue channel H in the threshold application operation is 100<H≤255.

[0016] Preferably, the smoothing process includes smoothing the image using a Gaussian kernel function.

[0017] Preferably, the edge detection includes calculating the gradient amplitude and gradient direction of each pixel in the image, then performing non-maximum suppression and pixel classification to obtain weak edge pixels and strong edge pixels, then connecting the weak edge pixels around the strong edge pixels to form an edge line, and performing Canny edge detection.

[0018] Preferably, the coordinate axis identification includes detection and identification using Hough transform.

[0019] As a preferred technical solution of the present invention, the character recognition includes using character recognition technology to identify the real coordinates of the coordinate axis, obtain the real coordinates of the geochemical logging chromatogram and the values ​​of the pixel coordinates, and obtain the actual coordinates of the curve through scaling conversion.

[0020] As a preferred technical solution of the present invention, the curve area calculation includes determining the peaks and troughs according to filtering rules, and then calculating the peak area, the lower baseline ridge area and the envelope area respectively.

[0021] As a preferred technical solution of the present invention, the filtering rule includes retaining only the points with the largest ordinate as the peak and the points with the smallest ordinate as the trough for adjacent points with the horizontal coordinate.

[0022] Preferably, the peak area includes the area enclosed by the broken line portion of the peak and the line connecting the corresponding two troughs.

[0023] Preferably, the raised area of ​​the lower baseline includes the area enclosed by the trough line and the horizontal axis.

[0024] Preferably, the envelope area includes the area enclosed by all trough lines and the horizontal axis.

[0025] As a preferred technical solution of the present invention, constructing different input sequences according to the obtained curve area includes constructing a density input sequence and constructing a viscosity input sequence.

[0026] Preferably, the structural density input sequence divides the geochemical logging chromatogram into first equal parts, and calculates the ratio of peak areas between different regions to complete the construction of input data.

[0027] Preferably, the first division includes dividing the geochemical logging chromatogram into at least three equal parts.

[0028] Preferably, the structural viscosity input sequence includes dividing the geochemical logging chromatogram into a second equal part, calculating the ratio of the peak area to the lower baseline uplift area between different regions, and then using the ratio of different units between different regions as the input sequence to complete the construction of the input data.

[0029] Preferably, the second division comprises dividing the geochemical logging chromatogram into at least three equal parts.

[0030] As a preferred technical solution of the present invention, the preprocessing includes preprocessing the data using maximum and minimum normalization.

[0031] Preferably, the mean absolute error is used as the loss function in constructing the fluid parameter prediction model using the multi-layer perceptron.

[0032] Preferably, the crude oil fluid parameters include crude oil density and / or crude oil viscosity.

[0033] In a second aspect, the present invention provides a method and device for predicting crude oil fluid parameters based on geochemical logging chromatograms, the device comprising:

[0034] The actual coordinate acquisition module is used to perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve;

[0035] A curve area calculation module, used to calculate the curve area according to the actual coordinates of the obtained curve;

[0036] The crude oil fluid parameter prediction module is used to construct different input sequences according to the obtained curve area, and then use the multi-layer perceptron to build a fluid parameter prediction model after preprocessing the input sequence, and then run the fluid parameter prediction model to obtain the predicted crude oil fluid parameters.

[0037] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0038] at least one processor; and

[0039] a memory communicatively coupled to the at least one processor;

[0040] Wherein, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for predicting crude oil fluid parameters based on geochemical logging chromatograms described in the first aspect.

[0041] Compared with the prior art solutions, the present invention has the following beneficial effects:

[0042] (1) The present invention utilizes image processing technology to complete the vectorization processing of the geochemical chromatogram and remove irrelevant factors therein, and the method is universal.

[0043] (2) The present invention combines optical character recognition (OCR) technology to complete the recognition and conversion of the real coordinates and pixel coordinates of the geochemical chromatogram, which can not only complete the analysis of new data, but also analyze the stock data of different standards.

[0044] (3) Based on the professional knowledge of the oil and gas industry and combined with neural networks, the present invention models the mapping relationship between geochemical chromatograms and viscosity and density, and completes the viscosity and density prediction based on geochemical chromatograms for the first time. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of a method for predicting crude oil fluid parameters based on geochemical logging chromatograms provided by an embodiment of the present invention;

[0046] Figure 2 It is a schematic diagram of a device for predicting crude oil fluid parameters based on geochemical logging chromatograms provided by an embodiment of the present invention;

[0047] Figure 3 is a schematic diagram of an electronic device provided by an embodiment of the present invention;

[0048] Figure 4 is a flow chart of the method for predicting crude oil fluid parameters based on geochemical logging chromatogram in Example 1 of the present invention;

[0049] Figure 5 is the geochemical logging chromatogram used in Example 1 of the present invention;

[0050] Figure 6 It is an image of the geochemical logging chromatogram in Example 1 of the present invention after irrelevant factors are removed;

[0051] Figure 7 is an image of the geochemical logging chromatogram processed by the Canny operator in Example 1 of the present invention;

[0052] Figure 8 It is the coordinate axis positioning result image of the geochemical logging chromatogram in Example 1 of the present invention;

[0053] Fig. 9 This is a preliminary analysis diagram of the peaks and troughs in the geochemical logging chromatogram in Example 1 of the present invention;

[0054] Fig.10 This is a peak and valley optimization detection result diagram in the geochemical logging chromatogram in Example 1 of the present invention;

[0055] Fig.11 It is a schematic diagram of three equal parts of the geochemical logging chromatogram in Example 1 of the present invention.

[0056] The present invention is further described in detail below. However, the following examples are only simplified examples of the present invention and do not represent or limit the scope of protection of the present invention. The scope of protection of the present invention shall be subject to the claims. DETAILED DESCRIPTION

[0057] To better illustrate the present invention and facilitate understanding of the technical solution of the present invention, typical but non-limiting embodiments of the present invention are as follows:

[0058] This embodiment provides a method for predicting crude oil fluid parameters based on geochemical logging chromatograms, such as Figure 1 As shown, the method includes:

[0059] Perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve;

[0060] Calculate the curve area based on the actual coordinates of the obtained curve;

[0061] Different input sequences are constructed according to the obtained curve area. After preprocessing the input sequence, a fluid parameter prediction model is constructed using a multi-layer perceptron. Then, the fluid parameter prediction model is run to obtain the predicted crude oil fluid parameters.

[0062] In the present invention, the crude oil fluid parameters include crude oil density and / or crude oil viscosity.

[0063] The image processing includes sequentially performing HSV color space change, irrelevant factor removal, smoothing, edge detection and coordinate axis recognition.

[0064] The HSV color space change includes converting the image into the HSV color space.

[0065] The irrelevant factor removal includes applying a threshold operation based on the hue channel range to remove irrelevant parts.

[0066] Wherein, the range of the hue channel H in the threshold application operation is 100<H≤255.

[0067] The smoothing process includes smoothing the image using a Gaussian kernel function.

[0068] The expression of the Gaussian kernel function used is as follows:

[0069]

[0070] Where H(x,y) is the filtered image, σ is the standard deviation of the Gaussian distribution, x and y represent the horizontal and vertical coordinates of the image relative to the center point, and the center point is the center of the current filter.

[0071] Among them, the edge detection includes calculating the gradient amplitude and gradient direction of each pixel in the image, then performing non-maximum suppression and pixel classification to obtain weak edge pixels and strong edge pixels, and then connecting the weak edge pixels around the strong edge pixels to form edge lines to perform Canny edge detection.

[0072] The calculation formula of the gradient amplitude is as follows:

[0073]

[0074] Where G(x,y) is the gradient magnitude, G x (x, y) is the component of the gradient in the horizontal direction, G y (x, y) is the component of the gradient in the vertical direction, x is the horizontal coordinate, indicating the position from left to right on the image, and y is the vertical coordinate, indicating the position from top to bottom on the image.

[0075] The calculation formula of the gradient direction is as follows:

[0076]

[0077] Where Θ(x,y) is the gradient direction, G x (x,y) is the horizontal component of the gradient, G y (x, y) is the component of the gradient in the vertical direction, x is the horizontal coordinate, indicating the position from left to right on the image, and y is the vertical coordinate, indicating the position from top to bottom on the image.

[0078] The coordinate axis recognition includes detection and recognition using Hough transform.

[0079] The character recognition includes using character recognition technology to identify the real coordinates of the coordinate axis, obtaining the real coordinates of the geochemical logging chromatogram and the values ​​of the pixel coordinates, and obtaining the actual coordinates of the curve through scaling and conversion.

[0080] The curve area calculation includes determining the peaks and troughs according to filtering rules, and then calculating the peak area, the lower baseline ridge area and the envelope area respectively.

[0081] In the present invention, before determining the peaks and troughs according to the filtering rules, the characteristics of the peaks and troughs in the geochemical logging chromatogram can be optionally used, that is, the peak is the point with the largest ordinate among the surrounding pixels, and the trough is the opposite. The entire geochemical chromatogram is traversed to perform preliminary confirmation of the peaks and troughs.

[0082] The filtering rule includes retaining only the points with the largest ordinate as the peak and the points with the smallest ordinate as the trough for adjacent points with the horizontal coordinate.

[0083] The peak area includes the area enclosed by the broken line portion of the peak and the line connecting the two corresponding troughs.

[0084] The raised area of ​​the lower baseline includes the area enclosed by the trough line and the horizontal axis.

[0085] The envelope area includes the area enclosed by all trough lines and the horizontal axis.

[0086] Wherein, constructing different input sequences according to the obtained curve area includes constructing a density input sequence and constructing a viscosity input sequence.

[0087] The structural density input sequence divides the geochemical logging chromatogram into first equal parts, and calculates the ratio of peak areas between different regions to complete the construction of input data.

[0088] The first division includes dividing the geochemical logging chromatogram into at least three equal parts.

[0089] For example, the geochemical logging chromatogram is divided into three regions from left to right, region ①, region ② and region ③, and the four peaks in region ① are numbered from left to right, respectively. 1-1 , p 1-2 , p 1-3 , p 1-4 The 14 peaks in region ① are numbered from left to right, namely p 2-1 , p 2-2 ,......,p 2-14 Similarly, the peak number in region ③ is p 3-1 . Calculate the ratio of the areas between the peaks in different regions to construct the final input sequence. For example, the ratio of the areas of the 4 peaks in region ① to the areas of the 14 peaks in region ② is calculated as follows:

[0090]

[0091] in, Indicates that the label is p 1-m The area of ​​the peak Indicates that the label is p 2-n The peak area, m, n represent the number of peaks in different regions, Seq ①② Represents a sequence consisting of different peak area ratios in regions ① and ②.

[0092] Similarly, get Seq ①③ and Seq ②③ , and finally Seq ①② , Seq ①③ , Seq ②③ Compose a new sequence as input data.

[0093] The structural viscosity input sequence includes dividing the geochemical logging chromatogram into a second equal part, calculating the ratio of the peak area to the lower baseline uplift area between different regions, and then using the ratio of different units between different regions as the input sequence to complete the construction of the input data.

[0094] Wherein, the second division includes dividing the geochemical logging chromatogram into at least three equal parts.

[0095] For example, if the geochemical logging chromatogram is divided into three regions from left to right, region ①, region ② and region ③, the ratio of the peak area of ​​each peak to the baseline bulge area below it (the area enclosed by the trough line and the horizontal axis) is first calculated as a unit, and then the ratio of different units between different regions is used as the input sequence similar to the structural density input sequence process. Under the definition of the peak area of ​​different regions, the definition of the baseline bulge area under the peak is newly added. Taking region ① as an example, p 1-1 is the frontmost peak in region ①. Now define u 1-1 Peak p 1-1 The corresponding area of ​​the lower baseline bulge (the area enclosed by the trough line and the horizontal axis) is as follows:

[0096]

[0097] in, It is peak 1-n The area of It is peak 1-n Lower baseline raised part u 1-n area.

[0098] And so on to calculate Sep ①③ and Sep ②③ , and then the viscosity input sequence is obtained.

[0099] In constructing the density input sequence and constructing the viscosity input sequence, the peak area in the front region (the region with a small horizontal axis) is used as the numerator, which is obtained based on the experience of experts in the petrochemical industry.

[0100] Wherein, the preprocessing includes preprocessing the data by using maximum and minimum normalization.

[0101] The calculation formula used in the maximum and minimum normalization is as follows:

[0102]

[0103] Where x′ is the normalized data value, x is the original data value, min(x) is the minimum value of the original data value, and max(x) is the maximum value of the original data value.

[0104] Wherein, the mean absolute error is used as the loss function in constructing the fluid parameter prediction model using the multi-layer perceptron.

[0105] The calculation formula of the mean absolute error is as follows:

[0106]

[0107] In the formula, MAE is the mean absolute error, n is the number of training samples, y is the label of the sample, f(x) is the data value after mapping by the multi-layer perceptron, and x is the original data value.

[0108] Furthermore, this embodiment provides a method and device for predicting crude oil fluid parameters based on geochemical logging chromatograms, such as Figure 2 As shown, the device comprises:

[0109] The actual coordinate acquisition module is used to perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve;

[0110] A curve area calculation module, used to calculate the curve area according to the actual coordinates of the obtained curve;

[0111] The crude oil fluid parameter prediction module is used to construct different input sequences according to the obtained curve area, and then use the multi-layer perceptron to build a fluid parameter prediction model after preprocessing the input sequence, and then run the fluid parameter prediction model to obtain the predicted crude oil fluid parameters.

[0112] Wherein, the actual coordinate acquisition module includes an image processing module and a character recognition module.

[0113] The image processing module includes an HSV color space change submodule, an irrelevant factor removal submodule, a smoothing processing submodule, an edge detection submodule and a coordinate axis recognition submodule.

[0114] The HSV color space change submodule is used to convert the image into an HSV color space image.

[0115] The irrelevant factor removal submodule is used to remove irrelevant parts by applying a threshold operation according to the range of the hue channel, and the range of the hue channel H in the threshold operation is 100<H≤255.

[0116] The smoothing submodule is used to perform smoothing on the image using a Gaussian kernel function.

[0117] Among them, the edge detection submodule is used to calculate the gradient amplitude and gradient direction of each pixel in the image, and then perform non-maximum suppression and pixel classification to obtain weak edge pixels and strong edge pixels, and then connect the weak edge pixels around the strong edge pixels to form an edge line to perform Canny edge detection.

[0118] The coordinate axis recognition submodule is used for detection and recognition using Hough transform.

[0119] The character recognition module is used to recognize the real coordinates of the coordinate axis using character recognition technology, obtain the real coordinates of the geochemical logging chromatogram and the values ​​of the pixel coordinates, and obtain the actual coordinates of the curve through scaling conversion.

[0120] The curve area calculation module includes a filtering rule module, a peak area calculation module, a lower baseline ridge area calculation module and an envelope area calculation module.

[0121] The filtering rule module is used to determine the peaks and troughs, including retaining only the points with the largest ordinate as the peak and the points with the smallest ordinate as the trough for adjacent points with the horizontal coordinates.

[0122] The peak area calculation module is used to calculate the area enclosed by the broken line portion of the peak and the connecting line of the two corresponding troughs.

[0123] The lower baseline bulge area calculation module is used to calculate the area enclosed by the trough line and the horizontal axis.

[0124] The envelope area calculation module is used to calculate the area enclosed by all trough lines and the horizontal axis.

[0125] The crude oil fluid parameter prediction module includes a density input sequence construction module, a viscosity input sequence construction module, a preprocessing module and a prediction model construction module.

[0126] The density input sequence construction module is used to divide the geochemical logging chromatogram into first equal parts, calculate the ratio of peak areas between different regions to complete the construction of input data, and the first equal division includes dividing the geochemical logging chromatogram into at least 3 equal parts.

[0127] Among them, the viscosity input sequence construction module is used to divide the geochemical logging chromatogram into a second equal part, calculate the ratio of the peak area to the lower baseline uplift area between different regions, and then use the ratio of different units between different regions as the input sequence to complete the construction of the input data. The second equal division includes dividing the geochemical logging chromatogram into at least 3 equal parts.

[0128] Among them, the preprocessing module is used to preprocess the input sequence data using maximum and minimum normalization.

[0129] Among them, the prediction model building module is used to use the data obtained from the preprocessing module to build a fluid parameter prediction model using a multi-layer perceptron, and the mean absolute error is used as the loss function.

[0130] Furthermore, this embodiment provides an electronic device, Figure 3A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0131] like Figure 3 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory ROM12, a random access memory RAM13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory ROM12 or the computer program loaded from the storage unit 18 to the random access memory RAM13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, ROM12 and RAM 13 are connected to each other through a bus 14. The I / O interface 15 is also connected to the bus 14.

[0132] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0133] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a method for predicting crude oil fluid parameters based on geochemical logging chromatograms.

[0134] In some embodiments, the method for predicting crude oil fluid parameters based on geochemical logging chromatograms can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for predicting crude oil fluid parameters based on geochemical logging chromatograms described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the method for predicting crude oil fluid parameters based on geochemical logging chromatograms in any other appropriate manner (e.g., by means of firmware).

[0135] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0136] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0137] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0138] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0139] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0140] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0141] In order to illustrate the good prediction effect that can be achieved by the method for predicting crude oil fluid parameters based on geochemical logging chromatogram provided by the present invention, an actual example is used for illustration, as follows:

[0142] Example 1

[0143] This embodiment provides a method for digital analysis of thermal vaporized hydrocarbon gas chromatography based on image processing, optical character recognition (OCR) technology and neural network modeling. The process is as follows: Figure 4 As shown, the details are as follows:

[0144] 1. Input geochemical logging chromatogram as follows Figure 5 As shown, first convert it from RGB color space to HSV color space, and then determine the range of blue in HSV color space.

[0145] Usually, blue is within a certain range of the hue channel. Then, a threshold operation is applied to mark the pixels within this range as the target image and the remaining pixels as the background, thus removing the irrelevant factors. The result after removal is as follows: Figure 6 shown.

[0146] Based on the image with irrelevant factors removed, its pixel coordinates are obtained by traversing the entire image.

[0147] Next, edge detection and Hough transform line detection techniques are combined to identify the coordinate axis position. The Canny operator is used to perform edge detection on the original geochemical chromatogram. The results are as follows: Figure 7 shown.

[0148] After that, before locating and identifying the coordinate axis, the image needs to be smoothed by Gaussian filtering. The Gaussian kernel function smoothes the image, and the formula is as follows:

[0149]

[0150] Where H(x,y) is the filtered image, σ is the standard deviation of the Gaussian distribution, x and y represent the horizontal and vertical coordinates of the image relative to the center point, and the center point is the center of the current filter.

[0151] Calculate the gradient magnitude and direction of each pixel in the image. The gradient magnitude indicates the degree of change in the pixel value, and the gradient direction indicates the direction of the change. The gradient calculation formula is as follows:

[0152]

[0153] Among them, G(x,y) is the gradient magnitude, Θ(x,y) is the gradient direction, G x (x, y) is the component of the gradient in the horizontal direction, G y (x, y) is the component of the gradient in the vertical direction, x is the horizontal coordinate, indicating the position from left to right on the image, and y is the vertical coordinate, indicating the position from top to bottom on the image.

[0154] The calculated gradient magnitude and direction are subjected to non-maximum suppression to retain the pixels with the largest gradient magnitude and suppress other pixels. According to the set high and low thresholds, the pixels are classified, and the pixels with gradient magnitudes higher than the high threshold are marked as strong edges, and the pixels with gradient magnitudes between the low and high thresholds are marked as weak edges. Finally, a continuous edge line is formed by connecting the strong edge pixels with the surrounding weak edge pixels. The result of Canny edge detection is a binary image, in which strong edge pixels represent the edges in the image, and other pixels are regarded as background.

[0155] Next, apply Hough Transform to the result of Canny edge detection to detect straight lines. For a straight line, each point on the line can be represented by ρ and θ in the polar coordinate system. ρ is the vertical distance from the coordinate origin (0,0) to the straight line, and θ is the angle (in radians) between the normal from the x-axis to the straight line and the positive x-axis. For each point (ρ, θ), an accumulator is initialized to count the number of edge points on the straight line defined by the parameter pair. Whenever an edge point is on a straight line, the accumulator value corresponding to (ρ, θ) increases. Look for high-value points in the accumulator array. These points represent the straight lines that pass through the most edge points in the parameter space. For the coordinate axis, the vertical axis is the leftmost vertical line, and the horizontal axis is the bottommost horizontal line. The coordinate axis positioning results are as follows. Figure 8 shown.

[0156] Then, the optical character recognition (OCR) technology is combined to identify the real coordinates of the coordinate axis. The numerical area on the upper side of the scale is intercepted, data enhancement and data denoising are completed, and OCR digital recognition is used to obtain the corresponding scale value and scale position.

[0157] At this point, the values ​​of the real coordinates and pixel coordinates of the geochemical chromatogram have been obtained, and their mutual conversion relationship has been calculated. The pixel coordinates of the curve can be scaled and converted to obtain the actual coordinates of the curve, completing the mutual conversion between the two coordinates.

[0158] 2. Complete the peak and trough detection and area calculation of different parts according to the obtained coordinate values. Peaks and troughs have their own characteristics in geochemical chromatography, that is, the peak is the point with the largest vertical coordinate among the surrounding pixels, and the trough is the opposite. Traverse the entire geochemical chromatogram and preliminarily detect the peaks and troughs. Fig. 9 shown.

[0159] Design filtering rules for adjacent peaks and troughs, that is, for adjacent points on the horizontal axis, only the point with the largest vertical coordinate is retained as the peak, and the point with the smallest vertical coordinate is retained as the trough. The results after filtering are as follows Fig.10 shown.

[0160] Next, various area calculations are completed based on the obtained broken line coordinates and the detected peaks and troughs. Here, the peak area, lower baseline ridge area, and envelope area are defined respectively. The peak area is the area enclosed by the broken line part of the peak and the corresponding two troughs, the lower baseline ridge area is the area enclosed by the trough line and the horizontal axis, and the envelope area is the area enclosed by all trough lines and the horizontal axis. Finally, the corresponding area calculation can be completed according to the curve integration method.

[0161] 3. According to the different characteristics of density and viscosity, different input sequences are constructed, and a neural network model is built to establish the correspondence between geochemical chromatograms, viscosity and density, and finally complete the prediction of viscosity and density.

[0162] First, construct the input data. Figure 3 Equally divided, the schematic diagram is as follows Fig.11 In density prediction, the input data is constructed by the ratio of peak areas between different regions. The four peaks in region ① are numbered from left to right, namely p 1-1 , p 1-2 , p 1-3 , p 1-4 The 14 peaks in region ① are numbered from left to right, namely p 2-1 , p 2-2 ,......,p 2-14 Similarly, the peak number in region ③ is p 3-1 . Calculate the ratio of the areas between the peaks in different regions to construct the final input sequence. For example, the ratio of the areas of the 4 peaks in region ① to the areas of the 14 peaks in region ② is calculated as follows:

[0163]

[0164] in, Indicates that the label is p 1-m The area of ​​the peak Indicates that the label is p 2-nThe peak area, m, n represent the number of peaks in different regions. Fig.11 In the example, m is 4, n is 14, and Seq ①② Represents a sequence consisting of different peak area ratios in regions ① and ②.

[0165] Similarly, get Seq ①③ and Seq ②③ , and finally Seq ①② , Seq ①③ , Seq ②③ To form a new sequence as input data. Fig.11 For example, Seq ①② The length of the sequence represented is 56, Seq ①③ The length of the sequence represented is 4, Seq ②③ The length of the represented sequence is 14, so the length of the final sequence after concatenation is 74;

[0166] In viscosity prediction, according to the viscosity-related characteristics, the ratio of the peak area of ​​each peak to the baseline ridge area below it (the area enclosed by the trough line and the horizontal axis) is used as a unit. Then, similar to the process of constructing the density input sequence, the ratio of different units between different regions is used as the input sequence. Taking region ① as an example, p 1-1 is the most forward peak in region ①. Now define u 1-1 Peak p 1-1 The corresponding area of ​​the lower baseline bulge (the area enclosed by the trough line and the horizontal axis) is as follows:

[0167]

[0168] in, It is peak 1-n The area of It is peak 1-n Lower baseline raised part u 1-n In addition, the input sequence construction method is the same as that of density prediction. Similarly, Sep ①③ and Sep ②③ , and finally Sep ①② , Sep ①③ , Sep ②③ Compose a new sequence as the viscosity input sequence.

[0169] In constructing the density input sequence and constructing the viscosity input sequence, the peak area in the front region (the region with a small horizontal axis) is used as the numerator, which is based on the experience of experts in the petrochemical industry.

[0170] Use maximum and minimum value normalization to preprocess the data. The formula is as follows:

[0171]

[0172] Where x′ is the normalized data value, x is the original data value, min(x) is the minimum value of the original data value, and max(x) is the maximum value of the original data value.

[0173] Based on the multi-layer perceptron to construct the viscosity and density prediction model, the mean absolute error is used as the loss function for predicting viscosity and density. The formula is as follows:

[0174]

[0175] In the formula, MAE is the mean absolute error, n is the number of training samples, y is the label of the sample, f(x) is the data value after mapping by the multi-layer perceptron, and x is the original data value.

[0176] The mean absolute error provides an intuitive and robust way to quantify how far the predicted values ​​are from the actual values, making the error estimate uniformly sensitive to deviations from all forecasts.

[0177] Based on the above method flow, the viscosity and density prediction results on some data are shown in Table 1.

[0178] Table 1

[0179] Model Mean absolute error Mean relative error Density prediction model 0.021 0.026 Viscosity prediction model 9.517 0.398

[0180] The predicted values ​​of density and viscosity of some spectra are shown in Table 2 and Table 3 respectively.

[0181] Table 2

[0182] Predicted density Actual density Absolute error 0.8407 0.8529 0.0122 0.8750 0.8498 0.0252 0.8027 0.8083 0.0056 0.7948 0.7926 0.0022 0.7996 0.8038 0.0042 0.8489 0.87 0.0211

[0183] Table 3

[0184]

[0185]

[0186] It can be seen from Tables 1, 2 and 3 that in the solution provided by the present invention, a mapping of geochemical maps to density and viscosity is constructed through a multi-layer perceptron in a neural network, which verifies the effectiveness of the design of the method. It can be seen that the solution provided by the present invention can achieve efficient and accurate prediction of the viscosity and density of crude oil.

[0187] It is stated that the present invention illustrates the detailed structural features of the present invention through the above embodiments, but the present invention is not limited to the above detailed structural features, that is, it does not mean that the present invention must rely on the above detailed structural features to be implemented. Those skilled in the art should understand that any improvement of the present invention, equivalent replacement of the components selected by the present invention, addition of auxiliary components, selection of specific methods, etc., all fall within the protection scope and disclosure scope of the present invention.

[0188] The preferred embodiments of the present invention are described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

[0189] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0190] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the present invention, they should also be regarded as the contents disclosed by the present invention.

Claims

1. A method for predicting crude oil fluid parameters based on geochemical logging chromatograms, characterized in that: The method comprises: Perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve; Calculate the curve area based on the actual coordinates of the obtained curve; Different input sequences are constructed according to the obtained curve area. After preprocessing the input sequence, a fluid parameter prediction model is constructed using a multi-layer perceptron. Then, the fluid parameter prediction model is run to obtain the predicted crude oil fluid parameters.

2. The method according to claim 1, characterized in that: The image processing includes sequentially performing HSV color space change, irrelevant factor removal, smoothing, edge detection and coordinate axis recognition.

3. The method according to claim 2, characterized in that: The HSV color space change includes converting the image into the HSV color space; Preferably, the irrelevant factor removal includes applying a threshold operation to remove irrelevant parts according to the hue channel range; Preferably, the range of the hue channel H in the threshold application operation is 100<H≤255; Preferably, the smoothing process includes smoothing the image using a Gaussian kernel function; Preferably, the edge detection includes calculating the gradient amplitude and gradient direction of each pixel in the image, then performing non-maximum suppression and pixel classification to obtain weak edge pixels and strong edge pixels, then connecting the weak edge pixels around the strong edge pixels to form edge lines, and performing Canny edge detection; Preferably, the coordinate axis identification includes detection and identification using Hough transform.

4. The method according to any one of claims 1 to 3, characterized in that: The character recognition includes using character recognition technology to recognize the real coordinates of the coordinate axis, obtaining the real coordinates of the geochemical logging chromatogram and the values ​​of the pixel coordinates, and obtaining the actual coordinates of the curve through scaling conversion.

5. The method according to any one of claims 1 to 4, characterized in that: The curve area calculation includes determining the peaks and troughs according to the filtering rules, and then calculating the peak area, the lower baseline ridge area and the envelope area respectively.

6. The method according to claim 5, characterized in that The filtering rule includes retaining only the points with the largest ordinate as the peak and the points with the smallest ordinate as the trough for the adjacent points with the horizontal coordinate; Preferably, the peak area includes the area enclosed by the broken line portion of the peak and the connecting line of the corresponding two troughs; Preferably, the lower baseline raised area includes the area enclosed by the trough line and the horizontal axis; Preferably, the envelope area includes the area enclosed by all trough lines and the horizontal axis.

7. The method according to any one of claims 1 to 6, characterized in that: The constructing of different input sequences according to the obtained curve area includes constructing a density input sequence and constructing a viscosity input sequence; Preferably, the structural density input sequence divides the geochemical logging chromatogram into first equal parts, and calculates the ratio of peak areas between different regions to complete the construction of the input data; Preferably, the first division includes dividing the geochemical logging chromatogram into at least three equal parts; Preferably, the structural viscosity input sequence includes dividing the geochemical logging chromatogram into a second equal part, calculating the ratio of the peak area to the lower baseline uplift area between different regions, and then using the ratio of different units between different regions as the input sequence to complete the construction of the input data; Preferably, the second division comprises dividing the geochemical logging chromatogram into at least three equal parts.

8. The method according to any one of claims 1 to 7, characterized in that: The preprocessing includes preprocessing the data by using maximum and minimum normalization; Preferably, the mean absolute error is used as the loss function in constructing the fluid parameter prediction model using the multi-layer perceptron; Preferably, the crude oil fluid parameters include crude oil density and / or crude oil viscosity.

9. A method and device for predicting crude oil fluid parameters based on geochemical logging chromatograms, characterized in that: The device comprises: The actual coordinate acquisition module is used to perform image processing and character recognition on the geochemical logging chromatogram to obtain the actual coordinates of the curve; A curve area calculation module, used to calculate the curve area according to the actual coordinates of the obtained curve; The crude oil fluid parameter prediction module is used to construct different input sequences according to the obtained curve area, and then use the multi-layer perceptron to build a fluid parameter prediction model after preprocessing the input sequence, and then run the fluid parameter prediction model to obtain the predicted crude oil fluid parameters.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for predicting crude oil fluid parameters based on geochemical logging chromatograms as described in any one of claims 1-8.