A method for constructing a carbon migration and transformation model during the surface matrix deposition process based on multivariate analysis

Abstract

The present invention discloses a method for constructing a carbon migration and transformation model in a surface matrix layer deposition process based on multivariate analysis. The present invention relates to the field of data analysis and comprises the following steps: obtaining geological information, dividing regions according to the geological information, and obtaining surface matrix regions; obtaining multivariate data, corresponding time data and depth data corresponding to each surface matrix region; setting a corresponding depth data sequence according to the depth data of the multivariate data, and obtaining variable correlation factors and sequence influencing factors between multivariate data of each depth data sequence; constructing an initial transformation model based on the variable correlation factors and sequence influencing factors between the multivariate data; analyzing and processing the initial transformation model, obtaining a corresponding comprehensive transformation model, and adjusting model parameters according to the time data. The present invention improves the accuracy of constructing the carbon migration and transformation model to a certain extent.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and in particular to a method for constructing a carbon migration and transformation model in a surface matrix layer deposition process based on multivariate analysis. Background Art

[0002] Most of the existing carbon migration and transformation models are based on simple linear assumptions and empirical formulas, and often ignore the complex interactions between multiple variables. For example, a soil organic carbon density and reserve estimation method with publication number CN112782385A discloses a soil organic carbon density and reserve estimation method, which belongs to the field of quantitative soil science technology and includes the following steps: setting soil sample point locations, collecting profile samples of the soil body, and measuring the thickness of each soil layer at the same time; testing the organic carbon content of each soil layer, the volume percentage of gravel larger than 2 mm, and the bulk density index; calculating the soil organic carbon density per centimeter depth of each soil layer sample, and establishing a linear regression model between the soil organic carbon density per centimeter depth and the soil organic carbon content; based on the established linear regression model, establishing the functional relationship between soil organic carbon density and organic carbon content, and between organic carbon reserves and organic carbon content, respectively. This invention directly, quickly and accurately estimates soil carbon density and carbon reserves through organic carbon content, thereby improving the accessibility of soil organic carbon reserve data; however, when faced with actual conditions with multiple environmental factors and complex geological conditions, it often suffers from the problem of insufficient accuracy; therefore, how to improve the accuracy of the surface matrix layer carbon migration and transformation model during the analysis process under the influence of multiple environmental factors and complex geological conditions is a problem we need to solve. To this end, a method for constructing a surface matrix layer carbon migration and transformation model in the deposition process based on multivariate analysis is provided. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide a method for constructing a carbon migration and transformation model in the surface matrix layer deposition process based on multivariate analysis.

[0004] The purpose of the present invention can be achieved by the following technical solution: A method for constructing a carbon migration and transformation model in a surface matrix layer deposition process based on multivariate analysis comprises the following steps:

[0005] Step S1: obtaining geological information in a region, dividing the region according to the obtained geological information, and setting corresponding surface matrix regions according to the regional division results;

[0006] Step S2: setting a data monitoring device according to the surface matrix area, acquiring corresponding multivariate data based on the data monitoring device, and marking the depth data and time data of the corresponding multivariate data;

[0007] Step S3: setting a corresponding data set according to the depth data of the multivariate data, setting a depth data sequence, and setting a variable correlation factor and a sequence influence factor between the multivariate data according to the data set corresponding to the depth data sequence;

[0008] Step S4: Analyze and train the deep data sequence according to its corresponding variable association factor, sequence impact factor and corresponding data set to construct an initial conversion model;

[0009] Step S5: dynamically analyze the initial transformation model corresponding to the depth data sequence according to the time data, adjust the model parameters according to the dynamic analysis results, and obtain a comprehensive transformation model.

[0010] Further, the process of setting the corresponding surface matrix area includes:

[0011] Setting up a carbon migration analysis platform, and setting up a geological information collection window in the carbon migration analysis platform;

[0012] Acquire geological information in the area through the geological information acquisition window, the geological information includes geological survey information, GIS data information and geological exploration information;

[0013] The obtained geological survey information and geological exploration information are used to generate geological layer information, and the geological layer information is spatially analyzed based on GIS data information to generate a three-dimensional geological feature distribution map; image processing is performed on the obtained three-dimensional geological feature distribution map, and the three-dimensional geological feature distribution map is divided into regions according to the image processing results to obtain different types of surface matrix areas, and the obtained surface geological areas are marked.

[0014] Furthermore, the process of obtaining multivariate data includes:

[0015] Obtaining the marking results of the corresponding surface geological areas in the three-dimensional geological feature distribution map, and setting corresponding data monitoring equipment according to the marking results of the surface geological areas;

[0016] The carbon migration analysis platform is provided with integrated monitoring schemes corresponding to different surface geological regions, and the integrated monitoring schemes are respectively provided with corresponding types of multi-sensor equipment and installation methods. Based on the integrated monitoring scheme and the marking results of the three-dimensional geological feature distribution map, corresponding data monitoring equipment is installed in the corresponding surface geological region; corresponding multivariate data is obtained through the data monitoring equipment;

[0017] The multivariate data include vegetation data, soil physical and chemical property data, meteorological data and carbon dynamic conversion data; the multivariate data obtained by the corresponding data monitoring equipment are time-stamped and mapped to a three-dimensional geological feature distribution map, and the time data and depth data corresponding to the multivariate data are obtained based on the mapping results of the time tag and the three-dimensional geological feature distribution map.

[0018] Furthermore, the process of setting the depth data sequence includes:

[0019] The obtained three-dimensional geological feature distribution map is vertically analyzed to obtain the marking results of the corresponding surface matrix area in the vertical direction, and the corresponding depth data sequence is set according to the different marking results, and the depth data sequence is marked in sequence from the surface;

[0020] Obtain regional vertical depth data corresponding to the surface matrix regional marking result corresponding to each depth data sequence, set unit depth data, sequentially perform segmentation processing on the regional vertical depth data corresponding to the corresponding depth data sequence based on the unit depth data, sequentially obtain regional segments according to the segmentation processing result, set the obtained regional segments as depth data subsequences, and sequentially mark each depth data subsequence in the corresponding depth data sequence from the surface downward;

[0021] The multivariate data mapped in the three-dimensional geological feature distribution map is obtained, a time monitoring period is preset, a corresponding data set is set for the multivariate data of the type corresponding to the corresponding depth data subsequence according to the time monitoring period, and the data set is marked according to the corresponding time data and depth data subsequence.

[0022] Furthermore, the process of setting the variable correlation factor between the multivariate data according to the data set corresponding to the depth data sequence includes:

[0023] Obtaining the geological surface region type corresponding to the depth data sequence, combining the data sets corresponding to each depth data subsequence in the depth data sequence corresponding to the corresponding geological surface region type, setting the corresponding data change curve according to the time data in turn according to the combination result corresponding to the data set, performing Fourier transform on the obtained data change curve, and obtaining the data change signal;

[0024] The obtained data change signal is subjected to scale and position analysis to obtain corresponding scale data and position data, and variable correlation factors between multivariate data types corresponding to the corresponding geological surface area are obtained based on the scale data and position data.

[0025] Furthermore, the process of setting the sequence influence factor between multivariate data includes:

[0026] Obtain the sorting type corresponding to the adjacent depth data subsequences, and set the corresponding multivariate data to the corresponding sequence association data table in turn according to the sorting type;

[0027] Based on the sequence association data table, sequentially obtain data transformation curves corresponding to the same type of multivariate data corresponding to the corresponding sorting type;

[0028] The obtained data transformation curve is analyzed and processed based on the linear regression model to obtain the corresponding sequence impact factor, and the obtained sequence impact factor is marked according to the sorting type corresponding to the adjacent depth data subsequence.

[0029] Furthermore, the process of constructing the initial conversion model includes:

[0030] Obtain the variable correlation factors, sequence influence factors and corresponding data sets corresponding to the corresponding deep data subsequences in each deep data sequence, set the initial model training sets between the corresponding deep data subsequences based on the variable correlation factors and sequence influence factors between the obtained data sets, obtain the initial model training sets corresponding to the corresponding deep data subsequences in turn based on the deep learning algorithm for analysis and training, obtain the corresponding initial conversion model, and verify the obtained initial conversion model, and correlate the initial conversion model that has completed the verification process with the corresponding deep data subsequence.

[0031] Furthermore, the process of obtaining the comprehensive transformation model includes:

[0032] Obtaining an initial conversion model corresponding to an adjacent depth data subsequence corresponding to the corresponding time data, obtaining the multivariate data and prediction analysis results that have been analyzed and processed in each depth data sequence in the vertical direction of the three-dimensional geological feature distribution map based on the initial conversion model, and setting a comprehensive model training set based on the multivariate data and prediction analysis results corresponding to the initial conversion model;

[0033] Analyze and process each comprehensive model training set based on the deep learning algorithm to obtain the comprehensive transformation model corresponding to each deep data sequence;

[0034] A dynamic monitoring cycle is set according to time data, and during the dynamic monitoring cycle, the comprehensive transformation model in each vertical direction in the three-dimensional geological feature distribution map is updated, and the model parameters of the initial transformation model and the comprehensive transformation model in the previous dynamic monitoring cycle are adjusted.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. By obtaining geological information, the underground conditions in the region are divided into regions according to the geological information, and the surface matrix region is set according to the regional division results. The corresponding depth data sequence is set in the vertical direction for the set surface matrix region, and the multivariate data in the different depth data sequences are analyzed and processed, so as to avoid the influence of the different surface matrix regions in the vertical direction on the construction of the carbon migration model to a certain extent;

[0037] 2. By setting the variable correlation factors between multivariate data and the sequence influence factors between each depth data subsequence, a model of carbon migration and transformation in the sedimentation process is constructed, which improves the accuracy of the carbon migration and transformation model to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a schematic diagram of a method for constructing a carbon migration and transformation model in a surface matrix layer deposition process based on multivariate analysis in an embodiment of the present application. DETAILED DESCRIPTION

[0039] like Figure 1 As shown, the method for constructing a carbon migration and transformation model in the surface matrix layer deposition process based on multivariate analysis includes the following steps:

[0040] Step S1: obtaining geological information in a region, dividing the region according to the obtained geological information, and setting corresponding surface matrix regions according to the regional division results;

[0041] Step S2: setting a data monitoring device according to the surface matrix area, acquiring corresponding multivariate data based on the data monitoring device, and marking the depth data and time data of the corresponding multivariate data;

[0042] Step S3: setting a corresponding data set according to the depth data of the multivariate data, setting a depth data sequence, and setting a variable correlation factor and a sequence influence factor between the multivariate data according to the data set corresponding to the depth data sequence;

[0043] Step S4: Analyze and train the deep data sequence according to its corresponding variable association factor, sequence impact factor and corresponding data set to construct an initial conversion model;

[0044] Step S5: dynamically analyze the initial transformation model corresponding to the depth data sequence according to the time data, adjust the model parameters according to the dynamic analysis results, and obtain a comprehensive transformation model.

[0045] It should be further explained that, in the specific implementation process, the process of obtaining geological information in the region, dividing the region according to the obtained geological information, and setting the corresponding surface matrix region according to the regional division result includes:

[0046] Setting up a carbon migration analysis platform, which is used to analyze and process the carbon migration situation during the deposition of surface geological layers in the corresponding area, so as to achieve carbon migration prediction;

[0047] Setting up a geological information collection window in the carbon migration analysis platform;

[0048] Acquire geological information in the area through the geological information acquisition window, the geological information includes geological survey information, GIS data information and geological exploration information;

[0049] Geological survey information includes geological survey reports, geological maps and academic literature;

[0050] GIS data information includes remote sensing information and terrain, landform, soil and vegetation information;

[0051] Geological exploration information includes the geological samples and measured data obtained, rock and soil samples collected, and laboratory analyses such as petrology, mineralogical and geochemical analyses to determine the composition and properties of geological sedimentary layers;

[0052] Generate geological layer information from the obtained geological survey information and geological exploration information;

[0053] The geological layer information includes geological layers of different geological units, including rock types, sedimentary layer distribution and structural characteristics;

[0054] Perform spatial analysis on geological layer information based on GIS data information to generate a three-dimensional geological feature distribution map; perform image processing on the obtained three-dimensional geological feature distribution map, and divide the three-dimensional geological feature distribution map into regions according to the image processing results;

[0055] The three-dimensional geological feature distribution map includes the different characteristics of each geological unit (such as lithology, age, genesis, etc.). Regional division is carried out according to the different characteristics of the geological units, different types of surface matrix areas are obtained, and the obtained surface geological areas are marked.

[0056] It should be further explained that, in the specific implementation process, the process of setting up a data monitoring device according to the surface matrix area, acquiring corresponding multivariate data based on the data monitoring device, and marking the depth data and time data of the corresponding multivariate data includes:

[0057] Obtaining the marking results of the corresponding surface geological areas in the three-dimensional geological feature distribution map, and setting corresponding data monitoring equipment according to the marking results of the surface geological areas;

[0058] The carbon migration analysis platform is provided with integrated monitoring solutions corresponding to different surface geological regions, and the integrated monitoring solutions are respectively provided with corresponding types of multi-sensor equipment and installation methods;

[0059] Install corresponding data monitoring equipment in the corresponding surface geological area based on the integrated monitoring scheme and the marking results of the three-dimensional geological feature distribution map; obtain corresponding multivariate data through the data monitoring equipment;

[0060] Multivariate data include vegetation data, soil physical and chemical property data, meteorological data, and carbon dynamic transformation data;

[0061] Vegetation data include vegetation index and vegetation type;

[0062] Soil physical and chemical properties data include soil carbon data, microbial data and organic matter decomposition data, soil physical properties and soil chemical properties;

[0063] Meteorological data include temperature, humidity, precipitation data, radiation data, and wind speed and direction data;

[0064] Carbon dynamic conversion data includes carbon emission data and carbon capture data;

[0065] The multivariate data obtained by the corresponding data monitoring equipment are time-tagged and mapped to a three-dimensional geological feature distribution map, and the time data and depth data corresponding to the multivariate data are obtained according to the mapping results of the time tag and the three-dimensional geological feature distribution map.

[0066] It should be further explained that, in a specific implementation process, the process of setting a corresponding data set according to the depth data of the multivariate data, setting a depth data sequence, and setting a variable correlation factor and a sequence influence factor between the multivariate data according to the data set corresponding to the depth data sequence includes:

[0067] The obtained three-dimensional geological feature distribution map is vertically analyzed to obtain the marking results of the corresponding surface matrix area in the vertical direction, and the corresponding depth data sequence is set according to the different marking results, and the depth data sequence is marked in sequence from the surface;

[0068] Obtain regional vertical depth data corresponding to the surface matrix regional marking result corresponding to each depth data sequence, set unit depth data, sequentially perform segmentation processing on the regional vertical depth data corresponding to the corresponding depth data sequence based on the unit depth data, sequentially obtain regional segments according to the segmentation processing result, set the obtained regional segments as depth data subsequences, and sequentially mark each depth data subsequence in the corresponding depth data sequence from the surface downward;

[0069] The multivariate data mapped in the three-dimensional geological feature distribution map is obtained, a time monitoring period is preset, a corresponding data set is set for the multivariate data of the type corresponding to the corresponding depth data subsequence according to the time monitoring period, and the data set is marked according to the corresponding time data and depth data subsequence.

[0070] Obtaining the geological surface region type corresponding to the depth data sequence, combining the data sets corresponding to each depth data subsequence in the depth data sequence corresponding to the corresponding geological surface region type, setting the corresponding data change curve according to the time data in turn according to the combination result corresponding to the data set, performing Fourier transform on the obtained data change curve, and obtaining the data change signal;

[0071] Performing scale position analysis on the acquired data change signal;

[0072] The obtained data change signal is marked as x(t); the scale parameter corresponding to the data change signal is marked as a, and the position parameter corresponding to the data change signal is marked as b; the variable correlation factor is marked as BG, (t) is the wavelet function; where:

[0073] .

[0074] Obtain the sorting type corresponding to the adjacent depth data subsequences, and set the corresponding multivariate data to the corresponding sequence association data table in turn according to the sorting type;

[0075] Based on the sequence association data table, sequentially obtain data transformation curves corresponding to the same type of multivariate data corresponding to the corresponding sorting type;

[0076] The obtained data transformation curve is analyzed and processed based on the linear regression model to obtain the corresponding sequence impact factor, and the obtained sequence impact factor is marked according to the sorting type corresponding to the adjacent depth data subsequence;

[0077] Establish a linear relationship model between dependent variables and independent variables, wherein the dependent variable is the multivariate data of the type corresponding to the next depth data subsequence, and the independent variable is the multivariate data of the type corresponding to the previous depth data subsequence adjacent to the next depth data subsequence;

[0078] Sequence impact factor analysis based on data transformation curve and corresponding linear regression model:

[0079] The sequence influencing factor is recorded as XY, and the multivariate data corresponding to the previous depth data subsequence is marked as ; The multivariate data corresponding to the previous depth data subsequence adjacent to the next depth data subsequence is marked as The error term is ; where i is the corresponding unit coordinate in the corresponding data transformation curve, where:

[0080] .

[0081] It should be further explained that, in the specific implementation process, the process of analyzing and training the deep data sequence according to its corresponding variable association factor, sequence influence factor and corresponding data set to construct the initial conversion model includes:

[0082] Obtain the variable correlation factor, sequence influence factor and corresponding data set corresponding to the corresponding deep data subsequence in each deep data sequence, set the initial model training set between the corresponding deep data subsequences based on the variable correlation factor and sequence influence factor between the obtained data sets, obtain the initial model training set corresponding to the corresponding deep data subsequence in turn based on the deep learning algorithm for analysis and training, obtain the corresponding initial conversion model, and verify the obtained initial conversion model, and associate the initial conversion model that has completed the verification process with the corresponding deep data subsequence;

[0083] Among them, the initial conversion model is used to predict and analyze the carbon transfer data of the depth data subsequence during the deposition of the surface matrix layer based on the multivariate data corresponding to the previous depth data subsequence in the depth data sequence, and obtain the prediction analysis results corresponding to the depth data subsequence.

[0084] It should be further explained that, in the specific implementation process, the process of dynamically analyzing the initial conversion model corresponding to the depth data sequence according to the time data, adjusting the model parameters according to the dynamic analysis results, and obtaining the comprehensive conversion model includes:

[0085] Obtaining an initial conversion model corresponding to an adjacent depth data subsequence corresponding to the corresponding time data, obtaining the multivariate data and prediction analysis results that have been analyzed and processed in each depth data sequence in the vertical direction of the three-dimensional geological feature distribution map based on the initial conversion model, and setting a comprehensive model training set based on the multivariate data and prediction analysis results corresponding to the initial conversion model;

[0086] Analyze and process each comprehensive model training set based on the deep learning algorithm to obtain the comprehensive transformation model corresponding to each deep data sequence;

[0087] A dynamic monitoring cycle is set according to time data, and during the dynamic monitoring cycle, the comprehensive transformation model in each vertical direction in the three-dimensional geological feature distribution map is updated, and the model parameters of the initial transformation model and the comprehensive transformation model in the previous dynamic monitoring cycle are adjusted.

[0088] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis, characterized in that: The following steps are involved: Step S1: obtaining geological information in the Dongting Lake area, performing regional division according to the obtained geological information, and setting corresponding surface matrix areas according to the regional division results; Step S2: setting a data monitoring device according to the surface matrix area, acquiring corresponding multivariate data based on the data monitoring device, and marking the depth data and time data of the corresponding multivariate data; Step S3: setting a corresponding data set according to the depth data of the multivariate data, setting a depth data sequence, and setting a variable correlation factor and a sequence influence factor between the multivariate data according to the data set corresponding to the depth data sequence; Step S4: Analyze and train the deep data sequence according to its corresponding variable association factor, sequence impact factor and corresponding data set to construct an initial conversion model; Step S5: dynamically analyzing the initial transformation model corresponding to the depth data sequence according to the time data, adjusting the model parameters according to the dynamic analysis results, and obtaining a comprehensive transformation model; The process of obtaining the comprehensive transformation model includes: Obtaining an initial conversion model corresponding to an adjacent depth data subsequence corresponding to the corresponding time data, obtaining the multivariate data and prediction analysis results that have been analyzed and processed in each depth data sequence in the vertical direction of the three-dimensional geological feature distribution map based on the initial conversion model, and setting a comprehensive model training set based on the multivariate data and prediction analysis results corresponding to the initial conversion model; Analyze and process each comprehensive model training set based on the deep learning algorithm to obtain the comprehensive transformation model corresponding to each deep data sequence; A dynamic monitoring cycle is set according to time data, and during the dynamic monitoring cycle, the comprehensive transformation model in each vertical direction in the three-dimensional geological feature distribution map is updated, and the model parameters of the initial transformation model and the comprehensive transformation model in the previous dynamic monitoring cycle are adjusted.

2. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 1 is characterized in that: The process of setting the corresponding surface matrix area includes: Setting up a carbon migration analysis platform, and setting up a geological information collection window in the carbon migration analysis platform; Obtain geological information in the Dongting Lake area through the geological information collection window, the geological information includes geological survey information, GIS data information and geological exploration information; The obtained geological survey information and geological exploration information are used to generate geological layer information, and the geological layer information is spatially analyzed based on GIS data information to generate a three-dimensional geological feature distribution map; image processing is performed on the obtained three-dimensional geological feature distribution map, and the three-dimensional geological feature distribution map is divided into regions according to the image processing results to obtain different types of surface matrix areas, and the obtained surface geological areas are marked.

3. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 2 is characterized in that: The process of obtaining multivariate data includes: Obtaining the marking results of the corresponding surface geological areas in the three-dimensional geological feature distribution map, and setting corresponding data monitoring equipment according to the marking results of the surface geological areas; The carbon migration analysis platform is provided with integrated monitoring schemes corresponding to different surface geological regions, and the integrated monitoring schemes are respectively provided with corresponding types of multi-sensor equipment and installation methods. Based on the integrated monitoring scheme and the marking results of the three-dimensional geological feature distribution map, corresponding data monitoring equipment is installed in the corresponding surface geological region; corresponding multivariate data is obtained through the data monitoring equipment; The multivariate data include vegetation data, soil physical and chemical property data, meteorological data and carbon dynamic conversion data; the multivariate data obtained by the corresponding data monitoring equipment are time-stamped and mapped to a three-dimensional geological feature distribution map, and the time data and depth data corresponding to the multivariate data are obtained based on the mapping results of the time tag and the three-dimensional geological feature distribution map.

4. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 3 is characterized in that: The process of setting the depth data sequence includes: The obtained three-dimensional geological feature distribution map is vertically analyzed to obtain the marking results of the corresponding surface matrix area in the vertical direction, and the corresponding depth data sequence is set according to the different marking results, and the depth data sequence is marked in sequence from the surface; Obtain regional vertical depth data corresponding to the surface matrix regional marking result corresponding to each depth data sequence, set unit depth data, sequentially perform segmentation processing on the regional vertical depth data corresponding to the corresponding depth data sequence based on the unit depth data, sequentially obtain regional segments according to the segmentation processing result, set the obtained regional segments as depth data subsequences, and sequentially mark each depth data subsequence in the corresponding depth data sequence from the surface downward; The multivariate data mapped in the three-dimensional geological feature distribution map is obtained, a time monitoring period is preset, a corresponding data set is set for the multivariate data of the type corresponding to the corresponding depth data subsequence according to the time monitoring period, and the data set is marked according to the corresponding time data and depth data subsequence.

5. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 4 is characterized in that: The process of setting the variable correlation factor between the multivariate data according to the data set corresponding to the depth data sequence includes: Obtaining the geological surface region type corresponding to the depth data sequence, combining the data sets corresponding to each depth data subsequence in the depth data sequence corresponding to the corresponding geological surface region type, setting the corresponding data change curve according to the time data in turn according to the combination result corresponding to the data set, performing Fourier transform on the obtained data change curve, and obtaining the data change signal; The obtained data change signal is subjected to scale and position analysis to obtain corresponding scale data and position data, and variable correlation factors between multivariate data types corresponding to the corresponding geological surface area are obtained based on the scale data and position data.

6. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 5 is characterized in that: The process of setting the serial influence factor between multivariate data includes: Obtain the sorting type corresponding to the adjacent depth data subsequences, and set the corresponding multivariate data to the corresponding sequence association data table in turn according to the sorting type; Based on the sequence association data table, sequentially obtain data transformation curves corresponding to the same type of multivariate data corresponding to the corresponding sorting type; The obtained data transformation curve is analyzed and processed based on the linear regression model to obtain the corresponding sequence impact factor, and the obtained sequence impact factor is marked according to the sorting type corresponding to the adjacent depth data subsequence.

7. The method for constructing a carbon migration and transformation model in the surface matrix deposition process based on multivariate analysis according to claim 6 is characterized in that: The process of constructing the initial conversion model includes: Obtain the variable correlation factors, sequence influence factors and corresponding data sets corresponding to the corresponding deep data subsequences in each deep data sequence, set the initial model training sets between the corresponding deep data subsequences based on the variable correlation factors and sequence influence factors between the obtained data sets, obtain the initial model training sets corresponding to the corresponding deep data subsequences in turn based on the deep learning algorithm for analysis and training, obtain the corresponding initial conversion model, and verify the obtained initial conversion model, and correlate the initial conversion model that has completed the verification process with the corresponding deep data subsequence.

Images