A quantitative evaluation method for feldspar dissolution and porosity increase in reservoirs

Abstract

The present invention relates to a quantitative evaluation method for feldspar dissolution pore increase in reservoirs, belonging to the technical field of oil and gas exploration and development. By carrying out XRD whole-rock diffraction experiments to determine the mass fractions of various feldspars in rocks, and using human-computer interaction image recognition technology to extract the field-of-view ratios and pore-face ratios before and after feldspar dissolution. The feldspar dissolution pore increase amounts obtained from the pore-face ratios extracted by image recognition respectively, and the pore increase amounts calculated based on the feldspar dissolution mechanism. The feldspar dissolution pore increase rate data Q1 and Q2 calculated by different methods for the same sample point are compared, and finally Q1 and Q2 are mutually verified to determine the feldspar dissolution pore increase rate and to test the correctness of the dissolution mechanism.

Description

Technical Field

[0001] The invention relates to a quantitative evaluation method for feldspar dissolution and pore increase in a reservoir, belonging to the technical field of oil and gas exploration and development. Background Art

[0002] Feldspar is the most abundant mineral species in the upper crust and the most important mineral component in the reservoirs of oil and gas basins. For deep and ultra-deep oil and gas reservoirs, the primary pores of the reservoirs are almost all destroyed by destructive diagenesis such as compaction, so secondary pores become the main storage space of the reservoirs. Feldspar dissolution is a common geological phenomenon, accompanied by the generation of dissolution pores, which can effectively improve the physical properties of the reservoir, and therefore has an important reservoir-forming effect.

[0003] At present, the research on secondary porosity increase caused by feldspar dissolution mainly includes: using weathering indexes such as CIA to indirectly judge the degree of feldspar dissolution. This method uses the ratio of the content of different ions to characterize the intensity of feldspar dissolution based on the difficulty of ion migration during feldspar weathering and dissolution. The disadvantage is that the characteristic ion content measured today is difficult to exclude the part of the ion content change caused by diagenesis other than feldspar dissolution, such as potassium replacement of feldspar, etc., which will lead to inaccurate characterization results. Another method is to fit the relationship between feldspar dissolution pores and pore throat radius, and the relationship between feldspar content and rock porosity and permeability to evaluate the reservoir effect of feldspar dissolution. However, this method can only qualitatively describe that the less the remaining feldspar content, the stronger the dissolution degree and the better the physical properties, and cannot quantitatively characterize the porosity increase caused by feldspar dissolution. Some scholars also directly use the feldspar dissolution reaction equation to calculate the amount of dissolution porosity. The amount of dissolution porosity calculated by this method is a theoretical amount of porosity, which cannot directly obtain the amount of feldspar dissolution porosity under actual geological conditions. In addition, there is a lack of effective verification mechanism, and the calculation results cannot be verified.

[0004] For example, the Chinese patent application document with application publication number CN108363117A discloses a method for quantitatively predicting the amount of porosity increase caused by feldspar dissolution. Although this method can also quantitatively evaluate the amount of feldspar dissolution, the result obtained is only the percentage of feldspar dissolution. It can only be judged that feldspar dissolution does produce a porosity increase effect on the reservoir, but the specific amount of increased porosity cannot be directly obtained. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a quantitative evaluation method for feldspar dissolution and porosity increase in reservoirs, so as to solve the problems of lack of research on quantitative evaluation of feldspar dissolution and porosity increase and inaccurate results.

[0006] The technical solution of the present invention is as follows:

[0007] A quantitative evaluation method for feldspar dissolution and porosity increase in a reservoir comprises the following steps:

[0008] 1) Randomly select samples and carry out mineral XRD whole-rock quantitative analysis experiments to determine the mass fraction of feldspar and calculate the amount of substance;

[0009] 2) Use human-computer interactive image recognition technology to determine the viewport ratio of residual feldspar in the casting thin section, that is, the viewport ratio after removing the dissolved pores, and establish a conversion relationship between the viewport ratio and the amount of feldspar calculated by XRD measurement R 1 , and calculate the amount of residual feldspar material;

[0010] 3) Use human-computer interactive image recognition technology to determine the proportion of each type of feldspar before dissolution, that is, the proportion of the original feldspar view, and use the relationship R 1 Get the amount of original feldspar in the rock;

[0011] 4) Calculate the amount of dissolved feldspar using the amount of original and residual feldspar in the rock, and calculate the feldspar dissolution pore increase amount Q based on the feldspar dissolution mechanism 1 ;

[0012] 5) Use human-computer interactive image recognition technology to extract the surface ratio of cast thin sections, including various surface ratios, and establish the relationship between the surface ratio and the measured porosity R 2 ;

[0013] 6) Based on the transformation relationship R 2 , combined with the human-computer interactive image recognition technology to extract the feldspar dissolution surface ratio, calculate the feldspar dissolution pore increase amount Q 2 ;

[0014] 7) Final Q 1 With Q 2 Mutual verification was carried out to determine the porosity increase of feldspar due to dissolution and to test the feasibility of quantitatively evaluating the amount of porosity increase due to dissolution mechanism.

[0015] The present invention determines the mass fraction of various feldspars in rocks by conducting XRD whole-rock diffraction experiments, and uses human-computer interactive image recognition technology to extract the viewport ratio and dissolution surface ratio of feldspar before and after dissolution. First, a conversion relationship R is established between the viewport ratio after feldspar dissolution and the amount of feldspar material measured and calculated by XRD experiments. 1 , determine the amount of residual feldspar in the rock. According to the relationship R 1 The amount of various original feldspar materials in the rock is obtained by combining the proportion of the view field before feldspar dissolution. After calculating the amount of dissolved feldspar materials by combining the amount of residual feldspar materials, the dissolution pore increase amount Q is obtained based on the feldspar dissolution mechanism. 1 Secondly, the transformation relationship between rock surface ratio and measured porosity data was established. 2 Then, the feldspar dissolution surface ratio and transformation relationship R extracted by image recognition method were 2, and calculate the feldspar dissolution pore increase Q 2 . Last Q 1 With Q 2 Mutual verification is carried out to determine the amount of feldspar dissolution pore increase and to test the feasibility of quantitatively evaluating the amount of dissolution pore increase by dissolution mechanism and the correctness of the results.

[0016] Preferably, in step 1), a mineral XRD whole-rock quantitative analysis experiment is carried out to obtain the mass fraction ω of each type of feldspar in the rock. 钾长石 ,ω 钠长石 ,ω 钙长石 , calculate the amount of each type of feldspar material n according to the material amount calculation formula 钾长石 、n 钠长石 、n 钙长石 .

[0017] (1)

[0018] (2)

[0019] (3)

[0020] Where ω is the mass fraction of different types of feldspar, m 样品 and M represent the sample mass and the molar mass of different types of feldspar respectively, and n is the amount of substance of different types of feldspar.

[0021] Preferably, in step 2), the sheet processing identification step is as follows:

[0022] 2-1) Preliminary processing of thin slice images: TWS (Trainable Weka Segmentation), a machine learning-based image segmentation tool, is used to pre-process the thin slices. Oversized images are compressed to increase processing speed. The image size is compressed to 3M-20M. TWS will segment the image into pixel blocks.

[0023] 2-2) Dataset establishment: Use the tool's built-in box selector to label the image samples in step 2-1) with classification labels. These sample images need to be labeled through human-computer interaction. During the labeling process, you can use the tool's built-in image labeling tool selection box to select different areas in the image as training sets for machine learning. This process can be added multiple times to extract the features of the image area, including color features, contour features, area and other shape features;

[0024] 2-3) The features extracted in step 2-2) are used to repeatedly train the SVM classifier model. The goal of feature extraction is to capture important information in the image to distinguish different pixel categories. During the training process, the classifier learns based on the labeled image samples and classifies according to the features, thereby achieving accurate image segmentation and identifying thin slice images. When TWS initially completes the image segmentation, human-computer interaction is required to evaluate and optimize the model. The evaluation includes calculating the accuracy and performance indicators of the segmentation results, including pixel-level accuracy, recall rate, and F1 score. Optimization includes manually adjusting feature selection and classifier parameters to improve the quality of the segmentation results, further expanding the training set and repeating steps 2-2) and 2-3) to train again until the recognition results are completely consistent with the original image features.

[0025] 2-4) Export the data for statistical analysis and synthesize the area of ​​the regions marked in the classification map, and then conduct quantitative analysis of the face ratio and feldspar field proportion;

[0026] The Excel software was used to draw the intersection diagram of the residual feldspar viewport ratio after dissolution and the amount of feldspar material, and the conversion relationship between the residual feldspar viewport ratio after dissolution and the amount of feldspar material was fitted. 1 : Y 1 =a 1 ×X 1 +b 1 When the amount of residual feldspar material needs to be calculated, the residual feldspar view area ratio is substituted into the formula Y 1 =a 1 ×X 1 +b 1 Replace X in 1 , the proportion of residual feldspar in the field of view can be converted into the amount of residual feldspar material.

[0027] where Y 1 is the amount of feldspar, X 1 is the viewing area proportion of residual feldspar, a 1 is the coefficient, b 1 is a constant.

[0028] Preferably, in step 4), the feldspar dissolution pore increase amount Q is calculated based on the feldspar dissolution mechanism. 1 , that is, the volume difference before and after the dissolution is calculated using the amount of material dissolved by feldspar, and the amount of dissolution pore increase is determined based on the ratio of the volume difference before and after to the total volume of the sample;

[0029] (4)

[0030] (5)

[0031] (6)

[0032] (7)

[0033] (8)

[0034] (9)

[0035] (10)

[0036] (11)

[0037] (12)

[0038] (13)

[0039] (14)

[0040] (15)

[0041] (16)

[0042] (17)

[0043] (18)

[0044] (19)

[0045] (20)

[0046] (twenty one)

[0047] Therefore, the quantitative characterization of feldspar dissolution and pore increase can be carried out. 1 =φ 钾长石溶蚀 +φ 钙长石溶蚀 +φ 钠长石溶蚀 ;

[0048] Among them, ρ represents the density of different types of rocks, V represents the volume difference caused by the dissolution of different types of feldspar (potassium feldspar, calcium feldspar, sodium feldspar), and V 总 is the total volume of rock, and φ represents the amount of pore increase caused by dissolution of different types of feldspar (potassium feldspar, calcium feldspar, and sodium feldspar).

[0049] Preferably, in step 5), Excel software is used to draw a cross-plot of the measured porosity and the face ratio extracted by the image recognition technology to establish a relationship between the face ratio and the measured porosity. 2 : Y 2 =a 2 ×X 2 +b2 ;

[0050] where Y 2 is the porosity, X 2 is the face rate, a 2 is the coefficient, b 2 is a constant.

[0051] Preferably, in step 7), SPSS software is used to analyze Q 1 With Q 2 To conduct a difference analysis, we first used SPSS software to analyze Q 1 With Q 2 The difference is tested for normal distribution. If Q 1 With Q 2 If the difference is in accordance with the normal distribution, SPSS software is used to analyze Q 1 With Q 2 A t-test was performed on the paired sample data. When the significant P value was > 0.05, it indicated that there was no significant difference in the results calculated by the two methods. If P < 0.05, it indicated that there was a significant difference in the results calculated by the two methods. 1 With Q 2 When the difference does not conform to the normal distribution, the Wilcoxon test method of two paired sample data is used to test whether there is a difference between the two methods; similarly, when the significant P value>0.05, it indicates that there is no significant difference in the results calculated by the two methods. If P<0.05, it proves that there is a significant difference in the results calculated by the two methods.

[0052] The beneficial effects of the present invention are:

[0053] The present invention is a method for calculating the amount of feldspar dissolution pore increase by utilizing the dissolution mechanism of feldspar, which solves the problem that the amount of feldspar dissolution pore increase cannot be directly and quantitatively evaluated, and provides an effective testing method (confidence level 95%) to verify whether the calculation result is accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of the quantitative evaluation method of feldspar dissolution and pore increase of the present invention;

[0055] Figure 2 is a diagram of the steps of extracting the feldspar visual field proportion and face rate by image recognition technology in an embodiment of the present invention;

[0056] Figure 3 This is a result diagram of the surface ratio of feldspar dissolution pores extracted by the image recognition technology in an embodiment of the present invention. The left figure is the original figure, and the right figure is the effect diagram of feldspar dissolution pore recognition;

[0057] Figure 4 This is a result diagram of the visual field ratio after feldspar dissolution identified by the image recognition technology in an embodiment of the present invention. The left picture is the original picture, and the right picture is the residual feldspar identification effect picture;

[0058] Figure 5 is a graph showing the relationship between the proportion of the viewing area after the feldspar is dissolved and the amount of the feldspar material converted in the embodiment of the present invention;

[0059] Figure 6 is a graph showing the conversion relationship between the surface porosity and the measured porosity in the embodiment of the present invention;

[0060] Figure 7 is a comparison chart of the porosity increase due to dissolution calculated based on the feldspar dissolution mechanism and the porosity increase due to dissolution calculated using the image recognition technology in the embodiment of the present invention;

[0061] Figure 8 Q in the embodiment of the present invention 1 With Q 2 Normality test results of differences. DETAILED DESCRIPTION

[0062] The present invention will be further described below by way of embodiments in conjunction with the accompanying drawings, but is not limited thereto.

[0063] Embodiment 1:

[0064] Feldspar dissolution pores have an important reservoir-forming effect on deep and ultra-deep tight reservoirs. The present invention proposes a quantitative evaluation method for feldspar dissolution pores. This method determines the mass fraction of various feldspars in the rock by conducting XRD whole-rock diffraction experiments, and uses human-computer interactive image recognition technology to extract the viewport ratio and dissolution surface ratio of feldspar before and after dissolution. First, a conversion relationship R is established between the viewport ratio after feldspar dissolution and the amount of feldspar material calculated by XRD data. 1 , and determine the amount of residual feldspar in the rock. According to the relationship R 1 Combined with the proportion of the field of view before feldspar dissolution, the original amount of each type of feldspar in the rock is obtained. The amount of dissolved feldspar is obtained by subtracting the amount of original feldspar and the amount of residual feldspar, and then the dissolution pore increase amount Q is calculated based on the feldspar dissolution mechanism. 1 Secondly, the conversion relationship between the rock thin section porosity and the measured porosity data was established. 2 Then, the feldspar dissolution surface ratio and transformation relationship R extracted by image recognition method were 2 , calculate the feldspar dissolution pore increase Q 2 . Last Q 1 With Q 2 Mutual verification is carried out to determine the amount of feldspar dissolution pore increase and to verify the feasibility of quantitatively evaluating the amount of dissolution pore increase by the dissolution mechanism. The implementation process of this method is as follows Figure 1 As shown, the implementation process of the invention is explained by taking a specific work area as an example, and the specific implementation process is as follows.

[0065] 1) Randomly select samples, wash them with oil, then dry them at a temperature below 60°C, cool and crush them until the total particle size is less than 40μm, and carry out mineral XRD whole-rock quantitative analysis experiments to obtain the mass fraction of various feldspars in the rock. 钾长石 ,ω 钠长石 ,ω 钙长石 , calculate the amount of each type of feldspar material n according to the material amount calculation formula 钾长石 、n 钠长石 、n 钙长石 .

[0066] (1)

[0067] (2)

[0068] (3)

[0069] Where ω is the mass fraction of different types of feldspar, m 样品 and M represent the sample mass and the molar mass of different types of feldspar respectively, and n is the amount of substance of different types of feldspar.

[0070] 2) Use human-computer interactive image recognition technology to determine the viewport ratio of residual feldspar in the casting thin section, that is, the viewport ratio after removing the dissolved pores, such as Figure 4 As shown, Figure 4 The left picture in the middle is the original picture, and the right picture is the residual feldspar identification effect picture, with the field of view accounting for 9.82%. The conversion relationship between the field of view percentage and the amount of feldspar calculated by XRD determination is established. 1 , and calculate the amount of residual feldspar, use a microscope to take thin sections of the rock, import the thin sections into the image recognition software, and use the segmentation tool Trainable Weka Segmentation to process the image. The thin section processing and recognition steps are as follows:

[0071] 2-1) Preliminary processing of thin slice images: TWS (Trainable Weka Segmentation), a machine learning-based image segmentation tool, is used to pre-process the thin slices. Oversized images are compressed to increase processing speed. The image size is compressed to 3M-20M. TWS will divide the image into pixel blocks.

[0072] 2-2) Dataset establishment: Use the tool's built-in box selector to label the image samples in step 2-1) with classification labels. These sample images need to be labeled through human-computer interaction. During the labeling process, you can use the tool's built-in image labeling tool selection box to select different areas in the image as training sets for machine learning. This process can be added multiple times to extract the features of the image area, including color features, contour features, area and other shape features.

[0073] 2-3) The features extracted in step 2-2) are used to repeatedly train the SVM classifier model. The goal of feature extraction is to capture important information in the image to distinguish different pixel categories. During the training process, the classifier learns based on the labeled image samples and classifies according to the features, thereby achieving accurate image segmentation and then identifying thin slice images. When TWS initially completes the image segmentation, human-computer interaction is required to evaluate and optimize the model. The evaluation includes calculating the accuracy and performance indicators of the segmentation results, including pixel-level accuracy, recall rate, and F1 score. Optimization includes manually adjusting feature selection and adjusting classifier parameters to improve the quality of the segmentation results, further expanding the training set, repeating steps 2-2), 2-3) and training again until the recognition results are completely consistent with the original image features. Through repeated adjustments, the software can accurately identify the features of the target to be extracted and automatically extract the features.

[0074] 2-4) Set the threshold to calculate the area of ​​the extracted target, export the data to analyze the area of ​​the area marked in the classification map, and then conduct quantitative analysis of the face rate and feldspar field ratio; Figure 2 shown.

[0075] The Excel software was used to draw the intersection diagram of the residual feldspar viewport ratio after dissolution and the amount of feldspar material, and the conversion relationship between the residual feldspar viewport ratio after dissolution and the amount of feldspar material was fitted. 1 : Y 1 =a 1 ×X 1 +b 1 ; When the amount of residual feldspar material needs to be calculated, the residual feldspar field ratio is substituted into the formula Y 1 =a 1 ×X 1 +b 1 Replace X in 1 , the proportion of residual feldspar in the field of view can be converted into the amount of residual feldspar. Figure 5 shown.

[0076] where Y 1 is the amount of residual feldspar, X 1 is the viewing area proportion of residual feldspar, a 1 is the coefficient, b 1 is a constant.

[0077] 3) Use human-computer interactive image recognition technology to determine the proportion of each type of feldspar before dissolution, that is, the proportion of the original feldspar view. According to the relationship R 1 : Y 1 =a 1 ×X 1 +b 1, put the original feldspar field ratio into formula Y 1 =a 1 ×X 1 +b 1 Replace X in 1 , to obtain the amount of original feldspar in the rock. The original feldspar field ratio is the sum of the residual feldspar and feldspar dissolution pore field ratios.

[0078] 4) Calculate the amount of dissolved feldspar by subtracting the amount of original feldspar from the amount of residual feldspar in the rock, and calculate the feldspar dissolution pore increase amount Q based on the feldspar dissolution mechanism. 1 , that is, the volume difference before and after the dissolution is calculated using the amount of material dissolved by feldspar, and the amount of dissolution pore increase is determined based on the ratio of the volume difference before and after to the total volume of the sample;

[0079] (4)

[0080] (5)

[0081] (6)

[0082] (7)

[0083] (8)

[0084] (9)

[0085] (10)

[0086] (11)

[0087] (12)

[0088] (13)

[0089] (14)

[0090] (15)

[0091] (16)

[0092] (17)

[0093] (18)

[0094] (19)

[0095] (20)

[0096] (twenty one)

[0097] Therefore, the quantitative characterization of feldspar dissolution and pore increase can be carried out. 1 =φ 钾长石溶蚀 +φ 钙长石溶蚀 +φ 钠长石溶蚀 ;

[0098] Among them, ρ represents the density of different types of rocks, V represents the volume difference caused by the dissolution of different types of feldspar (potassium feldspar, calcium feldspar, sodium feldspar), and V 总 is the total volume of rock, and φ represents the amount of pore increase caused by dissolution of different types of feldspar (potassium feldspar, calcium feldspar, and sodium feldspar).

[0099] 5) Use human-computer interactive image recognition technology to extract the surface ratio of the casting thin section, including the total surface ratio of various pores and the surface ratio of feldspar dissolution. The total surface ratio includes the dissolution surface ratio, but both cannot be extracted at once and need to be extracted separately. Figure 3 As shown, Figure 3 The left picture in the middle is the original picture, and the right picture is the feldspar dissolution pore identification effect picture, with a surface porosity of 4.42%. It is matched with the measured porosity of the corresponding sample point, and the intersection diagram of the measured porosity and the surface porosity extracted by image recognition technology is drawn using Excel software to establish the relationship between the surface porosity and the measured porosity R 2 : Y 2 =a 2 ×X 2 +b 2 ;like Figure 6 shown.

[0100] where Y 2 is the porosity, X 2 is the face rate, a 2 is the coefficient, b 2 is a constant.

[0101] 6) According to the transformation relationship R 2 : Y 2 =a 2 ×X 2 +b 2 , the feldspar dissolution surface ratio extracted by image recognition technology is brought into the formula to replace X 2 , the porosity Y is calculated 2 Feldspar dissolution pore increase Q 2 .

[0102] 7) The feldspar dissolution porosity data Q calculated by different methods at the same sample point 1 With Q 2 For comparison, such as Figure 7As shown in Figure 2. The porosity increase of feldspar due to dissolution obtained by the surface ratio extracted by image recognition is objective, so it is used to verify the porosity increase calculated based on the feldspar dissolution mechanism. 1 With Q 2 SPSS software was used to analyze Q 1 With Q 2 To conduct a difference analysis, we first used SPSS software to analyze Q 1 With Q 2 The difference is tested for normal distribution. If Q 1 With Q 2 If the difference is in accordance with the normal distribution, SPSS software is used to analyze Q 1 With Q 2 A t-test was performed on the paired sample data. When the significant P value was > 0.05, it indicated that there was no significant difference in the results calculated by the two methods. If P < 0.05, it indicated that there was a significant difference in the results calculated by the two methods. 1 With Q 2 When the difference does not conform to the normal distribution, the Wilcoxon test method of two paired sample data is used to test whether there is a difference between the two methods; similarly, when the significant P value>0.05, it indicates that there is no significant difference in the calculation results of the two methods, and if P<0.05, it proves that there is a significant difference in the calculation results of the two methods. In this way, the amount of feldspar dissolution pore increase can be determined and the feasibility of quantitatively evaluating the amount of dissolution pore increase by dissolution mechanism can be tested.

[0103] like Figure 8 As shown in the figure, in this embodiment, the data points are all near the diagonal line, indicating that the difference between the two is in accordance with the normal distribution according to the SPSS software analysis. Therefore, the difference between the two is analyzed by the t-test distribution of the two paired sample data. The analysis result shows that P>0.05, indicating that Q 1 With Q 2 There was no significant difference, as shown in Table 1.

[0104] Table 1 Q 1 With Q 2 Two paired sample t-test results

[0105]

[0106] This proves the feasibility of the method of the present invention, and can provide assistance for the quantitative evaluation of the amount of pore increase caused by feldspar dissolution in reservoirs.

Claims

1. A quantitative evaluation method for feldspar dissolution and porosity increase in reservoirs, characterized in that: The steps include: 1) Randomly select samples and carry out mineral XRD whole-rock quantitative analysis experiments to determine the mass fraction of feldspar and calculate the amount of substance; 2) The human-computer interactive image recognition technology is used to determine the viewport ratio of residual feldspar in the casting thin section, that is, the viewport ratio after removing the dissolved pores, and the conversion relationship R1 between the viewport ratio and the amount of feldspar material calculated by XRD measurement is established, and the amount of residual feldspar material is calculated; The steps for thin film processing identification are as follows: 2-1) Preliminary processing of thin slice images: TWS, an image segmentation tool based on machine learning, is used to pre-process the thin slices and compress the images to a size of 3M-20M; 2-2) Dataset establishment: Use the tool's built-in box selector to label the image samples in step 2-1) with classification labels. The labeling process uses the tool's built-in image labeling tool selection box to select different areas in the image as training sets for machine learning. This process is added multiple times to extract the features of the image area, including color features, contour features, and shape features. 2-3) The features extracted in step 2-2) are used to repeatedly train the SVM classifier model. During the training process, the classifier learns based on the labeled image samples and classifies according to the features, thereby achieving image segmentation and then identifying the thin slice image; when TWS initially completes the image segmentation, human-computer interaction is required to evaluate and optimize the model itself; the evaluation includes calculating the accuracy and performance indicators of the segmentation results, including pixel-level accuracy, recall rate, and F1 score; the optimization includes manually adjusting the feature selection and adjusting the classifier parameters, further expanding the training set and repeating steps 2-2) and 2-3) to train again until the recognition result is completely consistent with the original image features; 2-4) Export the data to analyze the area of ​​the regions marked in the classification map, and then conduct quantitative analysis of the face ratio and feldspar field ratio; The Excel software was used to draw the intersection diagram of the residual feldspar viewport ratio after dissolution and the amount of feldspar material, and the conversion relationship between the residual feldspar viewport ratio after dissolution and the amount of feldspar material was fitted R1: Y1 = a1 × X1 + b1; Among them, Y1 is the amount of feldspar material, X1 is the visual area proportion of residual feldspar, a1 is the coefficient, and b1 is a constant; 3) Use human-computer interactive image recognition technology to determine the proportion of each type of feldspar before dissolution, that is, the proportion of the original feldspar, and use the relationship R1 to obtain the amount of the original feldspar in the rock; 4) Calculate the amount of dissolved feldspar using the amount of original and residual feldspar in the rock, and calculate the feldspar dissolution pore increase amount Q1 based on the feldspar dissolution mechanism; <h2 style=";text-align:left;direction:ltr">2KAlSi3O8(AlO)+2H<h2 style=";text-align:left;direction:ltr"> + <h2 style=";text-align:left;direction:ltr"> +H2O = Al2Si2O5(OH)4 + 4SiO2 + 2K<h2 style=";text-align:left;direction:ltr"> + <h2 style=";text-align:left;direction:ltr"> (4) CaAl2Si2O8(calcium feldspar)+2H + +H2O=Al2Si2O5(OH)4+Ca 2+ (10) n:n 钙长石 n 钙长石 (11) <h2 style=";text-align:left;direction:ltr">2NaAlSi3O8(AlO)+2H<h2 style=";text-align:left;direction:ltr"> + <h2 style=";text-align:left;direction:ltr"> +H2O = Al2Si2O5(OH)4 + 4SiO2 + 2Na<h2 style=";text-align:left;direction:ltr"> + <h2 style=";text-align:left;direction:ltr"> (16) The quantitative characterization of feldspar dissolution and pore expansion is thus obtained. Among them, ρ represents the density of different types of rocks, V represents the volume difference caused by the dissolution of different types of feldspar, and V 总 is the total volume of rock, It indicates the amount of pore increase caused by dissolution of different types of feldspar; 5) Use human-computer interactive image recognition technology to extract the surface ratio of casting thin sections, including various surface ratios, and establish the relationship between the surface ratio and the measured porosity R2; The Excel software was used to draw the intersection diagram of the measured porosity and the surface ratio extracted by the image recognition technology, and the relationship between the surface ratio and the measured porosity was established: R2: Y2 = a2 × X2 + b2; Where Y2 is the porosity, X2 is the surface ratio, a2 is the coefficient, and b2 is the constant; 6) Based on the conversion relationship R2, combined with the feldspar dissolution surface ratio extracted by the human-computer interactive image recognition technology, the feldspar dissolution pore increase amount Q2 is calculated; 7) Finally, Q1 and Q2 are mutually verified to determine the porosity increase rate of feldspar dissolution and to test the feasibility of quantitatively evaluating the amount of dissolution porosity increase by the dissolution mechanism; Finally, SPSS software was used to analyze the differences between Q1 and Q2. SPSS software was first used to perform a normal distribution test on the difference between Q1 and Q2. If the difference between Q1 and Q2 conformed to the normal distribution, SPSS software was used to perform a paired sample data t test on Q1 and Q2. When the significant P value was >0.05, it indicated that there was no significant difference in the calculation results of the two methods. If P<0.05, it proved that there was a significant difference in the calculation results of the two methods. When the difference between Q1 and Q2 did not conform to the normal distribution, the Wilcoxon test of two paired sample data was used to test whether there was a difference between the two. Similarly, when the significant P value was >0.05, it indicated that there was no significant difference in the calculation results of the two methods. If P<0.05, it proved that there was a significant difference in the calculation results of the two methods.

2. The quantitative evaluation method for feldspar dissolution and pore increase in reservoir according to claim 1, characterized in that: In step 1), a mineral XRD whole-rock quantitative analysis experiment is carried out to obtain the mass fraction of various feldspars in the rock. 钾长石 ,ω 钠长石 ,ω 钙长石 , calculate the amount of each type of feldspar material n according to the material amount calculation formula 钾长石 、n 钠长石 、n 钙长石 ; Where ω is the mass fraction of different types of feldspar, m 样品 and M represent the sample mass and the molar mass of different types of feldspar respectively, and n is the amount of substance of different types of feldspar.

Images