Method, device and equipment for determining reservoir porosity based on rock physics knowledge

Abstract

This specification provides a method, device and equipment for determining reservoir porosity based on rock physics knowledge. The method includes obtaining a well logging data set of a target lithology category; screening a porosity index from the well logging data set using a preset method; wherein the porosity index represents a well logging curve that affects the reservoir porosity; obtaining a porosity estimation curve based on the data and porosity curve corresponding to the specified porosity index in the well logging data set; inputting the data and porosity estimation curve corresponding to the porosity index in the well logging data set into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining the training of multiple decision regressors based on rock physics information. Using the embodiments of this specification can not only better predict the porosity of complex reservoirs and the porosity of different lithologies, but also greatly improve the prediction accuracy of rock porosity.

Description

Technical Field

[0001] The present application relates to the technical field of petroleum exploration and development, and in particular to a method, device and equipment for determining reservoir porosity based on rock physics knowledge. Background Art

[0002] Reservoir evaluation is an important part of reservoir research and an important research content of oil and gas exploration and development. Porosity is one of the important physical parameters of reservoir evaluation, and its accurate evaluation is becoming more and more important for reservoir evaluation.

[0003] Since the size of porosity is affected by many different geological factors, such as structural location, burial depth, diagenesis degree, sedimentary environment, lithology changes, etc. In the prior art, reservoir porosity is mainly determined by artificial intelligence data-driven methods. This method selects logging data as input, manually finely interpreted porosity curves as labels, and uses artificial intelligence data-driven algorithms to establish nonlinear relationships between multiple logging attributes and porosity to predict porosity. However, due to the large difference in the nonlinear relationship between logging curves and porosity in different lithologies, the nonlinear relationship fitting effect of different lithologies using data-driven artificial intelligence algorithms is poor, which will reduce the accuracy of porosity prediction. In addition, since the cost of obtaining logging data through drilling is relatively high, and there are serious bottleneck problems such as insufficient labels and sample imbalance, this will also reduce the accuracy of porosity prediction using artificial intelligence data-driven methods.

[0004] Therefore, the industry is in urgent need of a technical solution that can solve the above technical problems. Summary of the invention

[0005] The embodiments of this specification provide a method, device and equipment for determining reservoir porosity based on rock physics knowledge, which can not only better predict the porosity of complex reservoirs and the porosity of different lithologies, but also greatly improve the prediction accuracy of rock porosity.

[0006] The method, device and equipment for determining reservoir porosity based on rock physics knowledge provided in this specification include the following implementation methods.

[0007] A method for determining reservoir porosity based on rock physics knowledge comprises: obtaining a well logging data set of a target lithology category; wherein each well logging data in the well logging data set corresponds to a plurality of well logging curves; screening a porosity index from the well logging data set in a preset manner; wherein the porosity index represents a well logging curve that affects the reservoir porosity; obtaining a porosity estimation curve based on data corresponding to a specified porosity index in the well logging data set and a porosity estimation curve; inputting data corresponding to the porosity index in the well logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining training of multiple decision regressors based on rock physics information.

[0008] The device for determining reservoir porosity based on rock physics knowledge comprises: an acquisition module for acquiring a logging data set of a target lithology category; wherein each logging data in the logging data set corresponds to a plurality of logging curves; a screening module for screening a porosity index from the logging data set in a preset manner; wherein the porosity index represents a logging curve that affects the reservoir porosity; a first acquisition module for obtaining a porosity estimation curve based on data and a porosity curve corresponding to a specified porosity index in the logging data set; and a second acquisition module for inputting data corresponding to the porosity index in the logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining training of multiple decision regressors based on rock physics information.

[0009] The device for determining reservoir porosity based on rock physics knowledge includes at least one processor and a memory storing computer executable instructions. When the processor executes the instructions, the steps of any method embodiment in the embodiments of this specification are implemented.

[0010] This specification provides a method, device and apparatus for determining reservoir porosity based on rock physics knowledge. In some embodiments, a well logging data set of a target lithology category can be obtained, and a porosity index can be screened from the well logging data set using a preset method; a porosity estimation curve can also be obtained based on the data and porosity curve corresponding to the specified porosity index in the well logging data set, and the data and porosity estimation curve corresponding to the porosity index in the well logging data set are input into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category. Due to the comprehensive use of well logging data with lithology interpretation and well logging data without lithology interpretation, a semi-supervised clustering algorithm is used to automatically classify multiple lithologies, which can not only add phase-controlled prior information to porosity prediction, perceive reservoir segments and non-reservoir segments based on prior information, but also focus on the porosity prediction of the reservoir segment, providing a basis for the porosity prediction of subsequent reservoir segments. Since the integrated porosity prediction model is obtained through pre-training, the efficiency of sandstone porosity prediction can be improved while improving the accuracy of sandstone porosity prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings described herein are used to provide a further understanding of this specification, constitute a part of this specification, and do not constitute a limitation of this specification. In the drawings:

[0012] Figure 1 A schematic diagram of a process for determining reservoir porosity based on rock physics knowledge provided in an embodiment of this specification;

[0013] Figure 2 Well logging data corresponding to the W1 well provided in the embodiments of this specification;

[0014] Figure 3 Well logging data corresponding to the W2 well provided in the embodiments of this specification;

[0015] Figure 4 Well logging data corresponding to the W3 well provided in the embodiments of this specification;

[0016] Figure 5 Well logging data corresponding to the W4 well provided in the embodiments of this specification;

[0017] Figure 6 Well logging data corresponding to the W5 well provided in the embodiments of this specification;

[0018] Figure 7 A schematic diagram of screening lithology sensitive attributes based on rock physics cross-plots provided in an embodiment of this specification;

[0019] Figure 8 A schematic diagram of the lithology interpretation results and lithology identification results of Well W5 provided in the embodiments of this specification;

[0020] Fig. 9 A schematic diagram of screening porosity sensitive attributes based on rock physics cross-plots provided in an embodiment of this specification;

[0021] Fig.10 A rock physics cross-plot of the acoustic time difference curve and porosity of the sandstone section provided in the embodiments of this specification and a schematic diagram of the fitting results;

[0022] Fig.11 A schematic diagram of obtaining the porosity of a reservoir section (sandstone) provided in an embodiment of this specification;

[0023] Fig.12 A schematic diagram showing the comparison between the actual sandstone porosity of Well W5 provided in the embodiments of this specification and the sandstone porosity obtained by using the integrated porosity prediction model;

[0024] Fig.13 A schematic diagram comparing the use of traditional rock physics intersection and the use of an integrated porosity prediction model to predict sandstone porosity provided in the embodiments of this specification;

[0025] Fig.14 A schematic diagram showing the comparison between the phase-controlled porosity of Well W5 obtained based on the integrated porosity prediction model provided in the embodiments of this specification and the actual phase-controlled porosity of Well W5;

[0026] Fig.15 A schematic diagram showing the comparison between the porosity of Well W5 obtained by directly using the Gradient Boosting Tree (GBDT) algorithm provided in the embodiments of this specification and the actual phase-controlled porosity of Well W5;

[0027] Fig.16 A schematic diagram of a specific process for determining reservoir porosity based on rock physics knowledge provided in an embodiment of this specification;

[0028] Fig.17 A schematic diagram of the module structure of an embodiment of a device for determining reservoir porosity based on rock physics knowledge provided in this specification;

[0029] Fig.18 A hardware structure block diagram of an embodiment of a server for determining reservoir porosity based on rock physics knowledge provided in this specification. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments in this specification, not all of the embodiments. Based on one or more embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the embodiments of this specification.

[0031] The following is an example of a specific application scenario to illustrate the implementation plan of this specification. Specifically, Figure 1 A schematic flow chart of determining reservoir porosity based on rock physics knowledge is provided for the embodiments of this specification. Although this specification provides method operation steps or device structures as shown in the following embodiments or drawings, more or fewer operation steps or module units may be included in the method or device based on routine or no creative labor.

[0032] An implementation scheme provided in this specification can be applied to a client, a server, etc. The client can include a terminal device, such as a smart phone, a tablet computer, etc. The server can include a single computer device, or a server cluster composed of multiple servers, or a server structure of a distributed system, etc.

[0033] It should be noted that the following embodiment description does not limit the technical solutions in other application scenarios that can be extended based on this specification. Figure 1 As shown, in an embodiment of a method for determining reservoir porosity based on rock physics knowledge provided in this specification, the method may include the following steps.

[0034] S0: Acquire a well logging data set of a target lithology category; wherein each well logging data in the well logging data set corresponds to a plurality of well logging curves.

[0035] Among them, the target lithology category can be sandstone (i.e., reservoir section) or mudstone (i.e., non-reservoir section), etc. The logging data set may include multiple logging data, and each logging data may correspond to multiple logging curves. Each logging curve can be understood as an attribute or indicator. It should be noted that in the embodiments of this specification, the target lithology category is sandstone as an example for exemplary description, and it does not constitute a limitation on other scenarios.

[0036] In some embodiments, the logging curve may include an acoustic time difference curve, a caliper curve, a gamma curve, a micro-gradient resistivity curve, a 2.5-meter bottom gradient resistivity curve, a natural potential curve, a mud content curve, and a porosity curve. Of course, the above is only an exemplary description, and the logging curve is not limited to the above examples. Under the inspiration of the technical essence of this application, technicians in the relevant field may also make other changes, such as the logging curve may include an acoustic time difference curve, a caliper curve, a gamma curve, a micro-gradient resistivity curve, a 2.5-meter bottom gradient resistivity curve, a natural potential curve, and a mud content curve, but as long as the functions and effects achieved are the same or similar to those of the present application, they should be covered within the scope of protection of the present application.

[0037] In some embodiments, obtaining a logging data set of a target lithology category may include: obtaining an initial logging data set; wherein each logging data in the initial logging data set corresponds to a plurality of logging curves, and the initial logging data set includes logging data of known lithology categories and logging data of unknown lithology categories; screening lithology indicators from the initial logging data set using a preset method; wherein the lithology indicators represent logging curves that affect reservoir lithology; determining the category of the logging data of the unknown lithology category in the initial logging data set based on the data corresponding to the lithology indicators in the initial logging data set and the logging data of the known lithology category, and obtaining a first logging data set; and obtaining a logging data set of a target lithology category from the first logging data set.

[0038] In some implementation scenarios, the logging data in the initial logging data set may be obtained by measuring downhole with a logging instrument, or may be obtained by other means, which is not limited in this specification.

[0039] Since it is very expensive to interpret lithology through logging, many well logging curves do not have lithology interpretation curves. Therefore, in some implementation scenarios, the initial logging data set may include a small amount of logging data with known lithology categories and a large amount of logging data with unknown lithology categories.

[0040] like Figure 2-Figure 6 As shown in the figure, the well logging data corresponding to the five wells (named W1, W2, W3, W4 and W5) in the actual work area are shown. Among them, W1 well and W5 well have lithology interpretation results (that is, known lithology categories). The well logging curves included in the well logging data mainly include AC (acoustic time difference curve), CAL (well diameter curve), GR (gamma curve), ML2 (micro-gradient resistivity curve), R25 (2.5-meter bottom gradient resistivity curve), SP (natural potential curve), VSH (shale content curve) and POR (porosity curve). Depth represents the depth, in meters (m).

[0041] In the long-term exploration and development of reservoirs, it is difficult to ensure that the logging data of all wells are measured with the same type of instruments, the same standard calibrators, and the same operating methods. Therefore, in order to reduce the errors caused by the inconsistency of instrument performance and calibration between the logging data corresponding to each well, in some implementation scenarios, after obtaining the logging data corresponding to each well, the logging data can be preprocessed. Among them, the preprocessing can include outlier removal, standardization, etc.

[0042] For example, in some implementation scenarios, after obtaining the logging data corresponding to each well, the abnormal values ​​in the logging curve (such as null values ​​Nan, -999, etc.) can be removed, and then the logging curve can be standardized according to the following formula:

[0043]

[0044] Among them, x * is the well logging curve after standardization, x is the well logging curve before standardization, is the average value of the logging curve, and σ is the standard deviation of the logging curve.

[0045] In the embodiments of this specification, high-quality logging data is obtained by preprocessing the logging data, which can provide a guarantee for subsequently improving the prediction accuracy of reservoir porosity.

[0046] In some implementation scenarios, after the well logging data is preprocessed, the preprocessed well logging data can be used as the initial well logging data set, and lithology indicators can be further screened from the initial well logging data set. Among them, the lithology indicators can represent the well logging curves that affect the lithology of the reservoir. Lithology indicators can also be called lithology sensitive attributes. For example, in some implementation scenarios, after obtaining the initial well logging data set, the lithology sensitive attributes in the well logging data can be screened based on the rock physics intersection diagram. It should be noted that in the embodiments of this specification, the lithology sensitive attributes are screened from AC, CAL, GR, ML2, R25, SP, and VSH. Figure 7As shown, it is a schematic diagram of screening lithology sensitive attributes based on rock physics intersection diagram provided by the embodiment of this specification. Among them, the horizontal and vertical coordinates are AC, CAL, GR, ML2, R25, SP, and VSH, respectively, and each sub-graph represents the intersection diagram of two curves corresponding to the horizontal and vertical coordinates. According to the rock physics intersection diagram, the influence of the logging curve on the lithology classification can be intuitively known. In some implementation scenarios of this specification, the lithology indicators may include GR, ML2, R25, SP and VSH. Of course, the above is only an exemplary description, and the lithology indicators are not limited to the above examples. For example, the lithology indicators may also include ML2, R25, SP and VSH. The technical personnel in the relevant field may make other changes under the inspiration of the technical essence of this application, but as long as the functions and effects achieved are the same or similar to those of this application, they should be covered within the scope of protection of this application.

[0047] In some implementation scenarios, after determining the lithology sensitive attributes, the category of the logging data of unknown lithology category in the initial logging data set can be determined based on the data corresponding to the lithology sensitive attributes in the initial logging data set and the logging data of known lithology category.

[0048] In some implementation scenarios, the determination of the category of the logging data of unknown lithology category in the initial logging data set based on the data corresponding to the lithology index in the initial logging data set and the logging data of known lithology category to obtain the first logging data set may include: determining the number of clusters and the center point of each cluster based on the logging data of known lithology category; calculating the distance from the data corresponding to the lithology index in the initial logging data set to the center point of each cluster; determining the category of the logging data of unknown lithology category in the initial logging data set according to the distance to the center point of each cluster to obtain the first logging data set. For example, in some implementation scenarios, the number of lithology categories included in the logging data of known lithology category can be determined as the K value (i.e., the number of clusters) in the K-means algorithm, and then the initial cluster center point is determined based on the logging data of known lithology category. Further, the distance from the data point corresponding to the lithology sensitive attribute to the center point of each cluster is calculated respectively, and each data point is assigned to the cluster (lithology category) with the closest distance. Update the number of iterations, re-determine the center point of each cluster (recorded as the "new cluster center point"), and calculate the distance from the data point corresponding to the lithology sensitive attribute to each "new cluster center point", and assign each data point to the cluster (lithology category) with the closest distance. Repeat the above steps until convergence or the preset number of iterations is reached, determine the category of the logging data of unknown lithology category in the initial logging data set, and obtain the first logging data set. At this point, the lithology category of each logging data in the first logging data set is known. In some implementation scenarios, the above distance can be the Minkowski distance, of course, it can also be other distances, and this manual does not limit this.

[0049] Specifically, assuming that the lithology categories included in the logging data of known lithology categories are sandstone and mudstone, the K value in the K-means algorithm can be determined to be 2. Further, two initial cluster center points can be determined based on the logging data of known lithology categories, and then the Minkowski distances from the data points corresponding to the lithology sensitive attributes to the two initial cluster center points are calculated respectively, and each data point is assigned to the lithology category with the closest 2-norm distance. Update the number of iterations, re-determine the center points of the two clusters (recorded as "new cluster center points"), and calculate the Minkowski distances from the data points corresponding to the lithology sensitive attributes to the two "new cluster center points" respectively, and assign each data point to the lithology category with the closest 2-norm distance. Repeat the above steps until the preset number of iterations is reached, determine the lithology category of the logging data of unknown lithology categories in the initial logging data set, and obtain the first logging data set.

[0050] In some implementation scenarios, in order to verify whether the automatic classification results of multiple lithologies using the semi-supervised clustering algorithm are accurate, the logging data of Well W5 can be used to verify the lithology identification results obtained by the semi-supervised clustering algorithm. Specifically, in the verification process, the number of clusters and the center point of each cluster can be determined based on the logging data of the wells W1, W2, W3, and W4 with known lithology categories. Then, the distance from the corresponding data of the lithology index in the logging data of Well W5 to the center point of each cluster is calculated, and the category of the logging data of Well W5 is determined based on the distance to the center point of each cluster. Repeat the above steps until the preset number of iterations is reached to obtain the lithology identification results. Furthermore, since the logging data of Well W5 has lithology interpretation results (i.e., known lithology categories), the lithology identification results can be compared with the lithology interpretation results. Figure 8 As shown, it is a schematic diagram of the lithology interpretation results and lithology identification results of Well W5 provided in the embodiment of this specification. Among them, the horizontal axis represents the depth range, and the vertical axis has no practical meaning. The figure below shows the lithology identification results of Well W5. The darker color represents mudstone, and the lighter color represents sandstone. By comparing the lithology interpretation results and lithology identification results of Well W5, it can be seen that the identification accuracy of sandstone and mudstone is 91.01%. It can be seen that the intelligent division of well logging lithology can be better realized through the semi-supervised clustering algorithm, which can provide a basis for the subsequent accurate prediction of sandstone porosity.

[0051] In some implementation scenarios, after obtaining the first well logging data set, a well logging data set of a target lithology category may be obtained from the first well logging data set. Specifically, a well logging data set of a lithology category of sandstone may be obtained from the first well logging data set.

[0052] In some implementation scenarios, after obtaining the logging data set of the target lithology category from the first logging data set, a logging data set of a non-target lithology category can also be obtained from the first logging data set; the value of the porosity curve in the logging data set of the non-target lithology category is set to a preset value to obtain the porosity of the reservoir corresponding to the non-target lithology category.

[0053] In actual formations, there are usually both reservoirs (i.e. sandstone) and non-reservoir layers (i.e. mudstone). Since the porosity of mudstone is much lower than that of sandstone, researchers usually focus on the porosity curve interpretation of the reservoir segment (i.e. sandstone) in conventional logging porosity interpretation. Therefore, in some implementation scenarios, for the lithology identification results of the above-mentioned semi-supervised clustering (i.e. the first logging data set), the porosity of mudstone can be set to a very small constant (such as 0.1, 0.3, etc.). Then, the porosity is intelligently predicted for the reservoir segment (i.e. sandstone). In this way, the porosity prediction of each segment and lithology can be completed in combination with the actual expert knowledge.

[0054] The embodiment of this specification simulates the attention mechanism of the neural network, comprehensively utilizes the logging data with lithological interpretation and the logging data without lithological interpretation, and adopts the semi-supervised clustering algorithm to automatically classify multiple lithologies, so as to realize the intelligent and fine classification of logging sandstone and mudstone, which can not only add the prior information of phase control to the porosity prediction, perceive the reservoir section and non-reservoir section based on the prior information, but also focus on the porosity prediction of the reservoir section, and provide a basis for the porosity prediction of the subsequent reservoir section. Among them, the attention mechanism can be mainly used to extract important features of the image so that the machine can perceive the important and unimportant parts of the image.

[0055] S2: Filtering a porosity index from the logging data set in a preset manner; wherein the porosity index represents a logging curve that affects reservoir porosity.

[0056] In some embodiments, after obtaining a well logging data set of a target lithology category, a porosity index may be screened from the well logging data set using a preset method. The preset method may include a rock physics cross plot method, etc. The porosity index may represent a well logging curve that affects the reservoir porosity. The porosity index may also be referred to as a porosity sensitive attribute.

[0057] In some implementation scenarios, after obtaining a well logging data set with a lithology category of sandstone from a first well logging data set with a known lithology category, the porosity sensitive attributes may be screened based on a rock physics cross plot of the sandstone well logging data set.

[0058] For example, in some implementation scenarios, after obtaining a well logging data set with a lithology category of sandstone from a first well logging data set with a known lithology category, the rock physics cross-plot corresponding to each well logging curve in the well logging data set can be obtained, and then the rock physics cross-plot is fitted using the least squares method to obtain the cross-plot fitting line, and further, the porosity sensitive attributes are screened according to the slope of the cross-plot fitting line. Among them, the larger the slope value, the better the correlation between the well logging curve and the porosity. Fig. 9 As shown, it is a schematic diagram of screening porosity sensitive attributes based on rock physics intersection diagram provided by an embodiment of this specification. Among them, the ordinate represents porosity, and the abscissa corresponds to AC, CAL, GR, ML2, R25, SP, and VSH in sequence. In some implementation scenarios of this specification, the porosity index may include AC, CAL, GR, ML2, R25 and SP. Of course, the above is only an exemplary description, and the porosity index is not limited to the above examples. For example, the porosity index may also include AC, CAL, ML2 and SP. The technical personnel in the relevant field may make other changes under the inspiration of the technical essence of this application, but as long as the functions and effects achieved are the same or similar to those of this application, they should be covered within the scope of protection of this application.

[0059] S4: obtaining a porosity estimation curve based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve.

[0060] The designated porosity index may be any one of the porosity indexes, such as AC, ML2, or other indicators. It should be noted that in the embodiments of this specification, AC is used as the designated porosity index for exemplary description, and other implementation scenarios are similar, which will not be described in detail.

[0061] In some embodiments, after the porosity index is screened from the well logging data set in a preset manner, a porosity estimation curve may be obtained based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve.

[0062] In some embodiments, obtaining a porosity estimation curve based on data and a porosity curve corresponding to a specified porosity index in the logging data set may include: establishing a rock physics model based on data and a porosity curve corresponding to the specified porosity index in the logging data set; using the rock physics to estimate porosity; and inputting data corresponding to the specified porosity index in the logging data set into the rock physics model to obtain a porosity estimation curve.

[0063] In some implementation scenarios, a simple rock physics model can be obtained by modeling and fitting based on the rock physics cross-plot of the acoustic time difference curve of the sandstone section and the porosity through the least squares method. Fig.10As shown, it is a rock physics cross plot of the acoustic time difference curve and porosity of the sandstone section provided in the embodiment of this specification and a schematic diagram of the fitting result. Among them, the horizontal axis represents the acoustic time difference AC, the vertical axis represents the porosity POR, the upper side and the right side represent the sample distribution of the acoustic time difference and the porosity respectively, and the straight line represents the fitting result of the acoustic time difference and the porosity. The simple rock physics model obtained by fitting can be expressed as:

[0064] Por=0.527×AC+0.0855 (2)

[0065] Among them, Por represents the porosity of the sandstone section, and AC represents the sonic time difference of the sandstone section.

[0066] In some implementation scenarios, after obtaining a simple rock physics model, data corresponding to the acoustic time difference in the sandstone logging data set can be input into the simple rock physics model to obtain a porosity estimation curve.

[0067] Of course, the above is only an exemplary description. The embodiments of this specification can also obtain the porosity estimation curve by other means. The technicians in the relevant field may make other changes inspired by the technical essence of this application. However, as long as the functions and effects achieved are the same or similar to those of this application, they should be included in the protection scope of this application.

[0068] S6: Inputting the data corresponding to the porosity index in the logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining training of multiple decision regressors based on rock physical information.

[0069] In some embodiments, after obtaining the porosity estimation curve, the data corresponding to the porosity index in the logging data set and the porosity estimation curve can be input into the integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category. Among them, the porosity estimation curve can be used as the prior information of the integrated porosity prediction model. The integrated porosity prediction model can be pre-guided by rock physical information (such as the rock physical relationship between acoustic wave time difference and porosity), and the idea of ​​serial learning of multiple regression models and integrated evaluation is adopted to train the preset integrated learning model. The integrated porosity prediction model may include multiple decision regressors. In some implementation scenarios, the integrated porosity prediction model may be a gradient boosting tree regressor.

[0070] like Fig.11 FIG. 1 is a schematic diagram of obtaining the porosity of a reservoir section (sandstone) provided in an embodiment of this specification. Fig.10The simple rock physics model obtained calculates the prior information of porosity (i.e., the porosity estimation curve POR_lin), and then the data AC, CAL, GR, ML2, R25, and SP corresponding to POR_lin and the porosity index are input into the gradient boosting tree regressor for intelligent prediction of sandstone porosity to obtain the sandstone porosity. Among them, the gradient boosting tree regressor includes n decision regressors (such as Fig.10 in Tree_1,…,Tree_n).

[0071] In some embodiments, the integrated porosity prediction model can be determined by: obtaining a well logging data set of a target lithology category in a training data set; screening a porosity index from the well logging data set using a preset method; obtaining a porosity estimation curve based on the data and porosity curve corresponding to the specified porosity index in the well logging data set; using the data and porosity estimation curve corresponding to the porosity index in the well logging data set as input data, using the target lithology category porosity curve of the real interpretation as a label, training a preset integrated learning model, and obtaining an integrated porosity prediction model. The preset integrated learning model includes multiple decision regressors.

[0072] For example, in some implementation scenarios, Figure 2-Figure 6 Taking the well logging data obtained as an example, after preprocessing the well logging data corresponding to well W1, well W2, well W3, well W4 and well W5, the well logging data of well W1, well W2, well W3 and well W4 are selected as the training data set, and the well logging data of well W5 is selected as the test data set. The lithology categories include sandstone and mudstone.

[0073] After obtaining the training data set, the rock physics intersection diagram can be used to screen the lithology indicators from the training data set. Based on the data corresponding to the lithology indicators in the training data set and the logging data of the W1 well, the categories of the logging data corresponding to the W2 well, the W3 well, and the W4 well in the training data set are determined. In this way, all the logging data in the training data set have known lithology categories. Further, a logging data set with a lithology category of sandstone is obtained from the training data set, and the porosity indicators are screened from the sandstone logging data set using the rock physics intersection diagram. Based on the data corresponding to AC in the sandstone logging data set and the porosity curve, the porosity estimation curve POR_lin is obtained. Furthermore, AC, CAL, GR, ML2, R25, SP and POR_lin can be used as input data, and the real interpreted sandstone porosity curve can be used as a label to train the preset integrated learning model to obtain an integrated porosity prediction model.

[0074] In some implementation scenarios, the ensemble porosity prediction model can be expressed as an ensemble boosting device with decision trees as base regressors:

[0075]

[0076] Among them, F M (x) represents the predicted porosity result, T(x; φ m ) represents the mth decision regressor, x represents the input data, φ m is the parameter of the decision regressor, and M is the total number of decision regressors.

[0077] Accordingly, the above-mentioned training of the preset integrated learning model may include the following steps:

[0078] (1) Determine the initial boost regressor:

[0079] F0(x)=0 (4)

[0080] (2) Using the forward distribution algorithm, the model of the m-th step (i.e., integrating the m-th decision regressor) is:

[0081] F m (x) = F m-1 (x)+T(x;φ m ) (5)

[0082] Among them, F m-1 (x) is the current model.

[0083] (3) Use the empirical risk structure minimization as the evaluation criterion to determine the parameter φ of the next decision regressor m :

[0084]

[0085] Among them, L is the loss function, N is the number of decision regressors, and y i For each decision regressor’s prediction, x i represents the input data, x i =x.

[0086] When the square error loss function is used as the evaluation criterion:

[0087] L[y,F(x)]=[yF m-1 (x)] 2 (7)

[0088] Among them, y is the porosity label (i.e., label), and F(x) represents the result predicted by the ensemble learning model (i.e., F M (x)).

[0089] At this time, the loss is:

[0090] L[y,F m-1 (x)+T(x;φ m )]=[yF m-1 (x)-T(x;φm )] 2 (8)

[0091] (4) Repeat operations (2) and (3) until the residual converges or the preset number of iterations is reached to obtain the integrated porosity prediction model.

[0092] It should be noted that the relevant steps S0-S6 involved in the process of training the integrated porosity prediction model can refer to the aforementioned embodiment and will not be described in detail.

[0093] Due to the characteristics of the decision regressor itself, even if there is a complex relationship between the input and output in the data, the linear combination of multiple decision regressors can well fit and express all the training data. Therefore, the final integrated porosity prediction model can be understood as a high-performance predictor.

[0094] In the embodiments of this specification, a strong learner with multiple regressors in parallel is constrainedly trained based on rock physical information to obtain an integrated porosity prediction model. Since it fully mines and integrates all existing data and expert knowledge, it can improve the prediction accuracy of the porosity of complex oil and gas reservoirs and provide an important reference for subsequent high-quality reservoir evaluation and drilling design.

[0095] In some implementation scenarios, after the integrated porosity prediction model is obtained through training based on the training data set, the integrated porosity prediction model can be verified using the logging data of the W5 well in the test data set.

[0096] Specifically, the porosity estimation curve POR_lin can be obtained based on the data corresponding to the acoustic time difference curve in the W5 well logging data and the porosity curve. The AC, CAL, GR, ML2, R25, SP and POR_lin in the W5 well logging data are input into the integrated porosity prediction model to obtain the sandstone porosity. Since the W5 well logging data has lithology interpretation results, the obtained sandstone porosity can be compared with the actual sandstone porosity. Fig.12 As shown, it is a schematic diagram comparing the actual sandstone porosity of Well W5 provided in the embodiment of this specification with the sandstone porosity obtained by using the integrated porosity prediction model. Among them, the abscissa represents depth, the ordinate represents POR represents porosity, Predict represents the actual sandstone porosity, Real represents the sandstone porosity obtained by using the integrated porosity prediction model, and the accuracy of the sandstone porosity predicted by the integrated porosity prediction model is 91.42%.

[0097] Furthermore, in order to verify the feasibility of the integrated porosity prediction model obtained in this application, this specification compares the prediction of sandstone porosity using traditional rock physics intersection with the prediction of sandstone porosity using the integrated porosity prediction model. Fig.13As shown, it is a comparative schematic diagram of the prediction of sandstone porosity using traditional rock physics intersection and integrated porosity prediction model provided in the embodiment of this specification. Among them, the horizontal axis represents depth, the vertical axis represents POR represents porosity, Predict represents the prediction of sandstone porosity using the integrated porosity prediction model, and Real represents the prediction of sandstone porosity using traditional rock physics intersection. The accuracy of predicting sandstone porosity using traditional rock physics intersection is 56.86%, which is much lower than the accuracy of predicting sandstone porosity using the integrated porosity prediction model in this application.

[0098] In summary, the integrated porosity prediction model obtained by the rock physics information guided training in this application has a high nonlinear fitting ability and can effectively improve the prediction accuracy of sandstone porosity.

[0099] In some implementation scenarios, after obtaining the porosity of the reservoir corresponding to the target lithology category based on the integrated porosity prediction model, the porosity of the reservoir corresponding to the target lithology category and the porosity of the reservoir corresponding to the non-target lithology category can be combined to obtain the phase-controlled porosity of the target reservoir. For example, in some implementation scenarios, after obtaining the sandstone porosity, the sandstone porosity and the mudstone porosity can be combined based on the clustering results to obtain the phase-controlled porosity.

[0100] like Fig.14 As shown, it is a schematic diagram of the comparison between the phase-controlled porosity of Well W5 obtained based on the integrated porosity prediction model provided in the embodiment of this specification and the real phase-controlled porosity of Well W5. Among them, the horizontal axis represents depth, the vertical axis represents POR represents porosity, Real represents the real phase-controlled porosity of Well W5, and Predict represents the phase-controlled porosity obtained by combining the sandstone porosity and mudstone porosity of Well W5 predicted by the integrated porosity prediction model. The accuracy of the phase-controlled porosity obtained by combining the sandstone porosity and mudstone porosity of Well W5 predicted by the integrated porosity prediction model is 78.85%.

[0101] like Fig.15 As shown, it is a schematic diagram of the comparison between the porosity of Well W5 obtained by directly using the gradient boosting tree (GBDT) algorithm provided in the embodiment of this specification and the real phase-controlled porosity of Well W5. Among them, the horizontal axis represents depth, the vertical axis represents POR, and the porosity represents Real, which represents the real phase-controlled porosity of Well W5, and Predict represents the porosity of Well W5 obtained directly using the gradient boosting tree (GBDT) algorithm. Since the porosity of Well W5 obtained directly using the gradient boosting tree (GBDT) algorithm is the porosity predicted by a pure data-driven device using the gradient boosting tree without distinguishing between reservoir and non-reservoir parts, and without considering the changes in lithology, its accuracy is only 16%.

[0102] In summary, the embodiments of this specification can not only better predict the porosity of complex reservoirs and the porosity of different lithologies, but also greatly improve the prediction accuracy of rock porosity.

[0103] In the embodiments of this specification, the integrated porosity prediction model is obtained by pre-training, which can not only improve the prediction accuracy of complex reservoir porosity and the efficiency of sandstone porosity prediction, but also provide important reference for subsequent high-quality reservoir evaluation, reservoir description, and drilling design. By combining sandstone porosity and mudstone porosity based on clustering results to obtain phase-controlled porosity, accurate prediction of porosity of different sedimentary formations in different regions can be achieved.

[0104] The embodiments of this specification comprehensively utilize logging data with lithological interpretation and logging data without lithological interpretation, automatically model multiple measured sensitive logging curves and logging lithological labels through a semi-supervised clustering algorithm, and apply the established model to areas without lithological interpretation, so as to realize automatic classification of multiple types of lithology and extract prior knowledge of phase control.

[0105] In the embodiments of this specification, for sandstone reservoir sections, a rock physical relationship between multiple logging curves and porosity is established through rock physical theory, and the porosity is estimated based on the rock physical relationship. Then, this model-driven rock physical relationship is indirectly used to take the estimated rough porosity and the measured logging curves as multivariate input data, and the idea of ​​ensemble learning is used to make joint decisions by weighted combination of multiple statistical prediction models, thereby realizing the prediction of sandstone porosity. This can not only predict the porosity of complex reservoirs with high precision, but also distinguish between reservoirs and non-reservoirs at different depths in different regions.

[0106] Of course, the above is only an exemplary description, and the embodiments of this specification are not limited to the above examples. Technical personnel in the relevant field may make other changes based on the technical essence of this application. However, as long as the functions and effects achieved are the same or similar to those of this application, they should be included in the scope of protection of this application.

[0107] From the above description, it can be seen that the embodiment of the present application can obtain a well logging data set of the target lithology category, and use a preset method to filter the porosity index from the well logging data set; it can also obtain a porosity estimation curve based on the data and porosity curve corresponding to the specified porosity index in the well logging data set, and input the data and porosity estimation curve corresponding to the porosity index in the well logging data set into the integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category. Due to the comprehensive use of well logging data with lithology interpretation and well logging data without lithology interpretation, a semi-supervised clustering algorithm is used to automatically classify multiple lithologies, which can not only increase the prior information of phase control for porosity prediction, perceive the reservoir section and non-reservoir section based on the prior information, but also focus on the porosity prediction of the reservoir section, providing a basis for the porosity prediction of the subsequent reservoir section. Since the integrated porosity prediction model is obtained by pre-training, it can improve the efficiency of sandstone porosity prediction while improving the accuracy of sandstone porosity prediction. By combining the sandstone porosity and mudstone porosity based on the clustering results to obtain the phase-controlled porosity, accurate prediction of the porosity of different sedimentary strata in different regions can be achieved.

[0108] like Fig.16 As shown, it is a specific flow chart of determining reservoir porosity based on rock physics knowledge provided by an embodiment of this specification. Among them, the integrated learning model constructed by multiple decision regressors is equivalent to the integrated porosity prediction model. In the specific implementation process, the well logging data and the corresponding partially labeled lithology curves can be input, and then the rock sensitive attributes can be screened by rock physics intersection analysis, and then the K value of the K-means algorithm is determined based on the labeled lithology curves, and the rock sensitive attributes are unsupervisedly classified by the K-means algorithm. Further, on the one hand, the non-reservoir porosity is set to a constant value of 0.1 based on expert experience. On the other hand, the reservoir segment logging curve and the corresponding porosity label are extracted, and the porosity sensitive attributes are screened by rock physics intersection analysis, and the porosity is estimated based on a simple rock physics model as the model prior information, and the corresponding data of the porosity sensitive attributes and the model prior information are used as the input of the integrated learning model constructed by multiple decision regressors to obtain the reservoir porosity prediction result. Finally, the non-reservoir porosity set based on expert experience and the obtained reservoir porosity prediction result are combined to obtain the phase-controlled porosity prediction result.

[0109] The embodiments of this specification can not only improve the problem of few wells and few labels, but also greatly improve the accuracy of porosity prediction of complex reservoirs by utilizing all measured information, rock physical information and expert knowledge.

[0110] The embodiments of this specification simulate the operating principle of the attention mechanism algorithm in computer vision. Through the semi-supervised clustering algorithm, the reservoir segment and the non-reservoir segment can be perceived, and the attention can be focused on the porosity prediction of the reservoir segment, thereby better meeting the needs of actual reservoir parameter prediction.

[0111] The embodiments of this specification combine the respective advantages of supervised and unsupervised methods, and develop a semi-supervised clustering method for logging sandstone and shale identification, which can maximize the use of all labeled and unlabeled logging data and avoid the situation where the classification accuracy is reduced due to limited labels.

[0112] In this specification, each embodiment of the above method is described in a progressive manner, and the same or similar parts between the embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments. For relevant parts, refer to the partial description of the method embodiment.

[0113] Based on the above-mentioned method for determining reservoir porosity based on rock physics knowledge, one or more embodiments of this specification also provide a device for determining reservoir porosity based on rock physics knowledge. The device may include a system (including a distributed system), software (application), module, component, server, client, etc. using the method described in the embodiment of this specification and a device combined with necessary implementation hardware. Based on the same innovative concept, the device in one or more embodiments provided in the embodiment of this specification is as described in the following embodiments. Since the implementation scheme and method of the device to solve the problem are similar, the implementation of the specific device in the embodiment of this specification can refer to the implementation of the aforementioned method, and the repetitions will not be repeated. As used below, the term "unit" or "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0114] Specifically, Fig.17 A schematic diagram of the module structure of an embodiment of a device for determining reservoir porosity based on rock physics knowledge provided in this specification is shown in FIG. Fig.17 As shown, a device for determining reservoir porosity based on rock physics knowledge provided in this specification may include: an acquisition module 120 , a screening module 122 , a first acquisition module 124 , and a second acquisition module 126 .

[0115] The acquisition module 120 may be used to acquire a well logging data set of a target lithology category; wherein each well logging data in the well logging data set corresponds to a plurality of well logging curves;

[0116] The screening module 122 may be used to screen the porosity index from the well logging data set in a preset manner; wherein the porosity index represents the well logging curve that affects the reservoir porosity;

[0117] A first obtaining module 124 may be used to obtain a porosity estimation curve based on data corresponding to a specified porosity index in the well logging data set and a porosity curve;

[0118] The second acquisition module 126 can be used to input the data corresponding to the porosity index in the logging data set and the porosity estimation curve into the integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining the training of multiple decision regressors based on rock physical information.

[0119] It should be noted that the above-mentioned device may also include other implementation modes according to the description of the method embodiment. The specific implementation mode may refer to the description of the relevant method embodiment, which will not be described one by one here.

[0120] This specification also provides an embodiment of a device for determining reservoir porosity based on rock physics knowledge, including a processor and a memory for storing processor executable instructions, and the processor can implement any of the above method embodiments when executing the instructions. For example, when the instructions are executed by the processor, the following steps are implemented: obtaining a well logging data set of a target lithology category; wherein each well logging data in the well logging data set corresponds to a plurality of well logging curves; screening a porosity index from the well logging data set using a preset method; wherein the porosity index represents a well logging curve that affects the reservoir porosity; obtaining a porosity estimation curve based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve; inputting the data corresponding to the porosity index in the well logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by training multiple decision regressors based on rock physics information.

[0121] It should be noted that the above-mentioned device may also include other implementations according to the description of the method or device embodiment. The specific implementation methods can refer to the description of the relevant method embodiments, which will not be described one by one here.

[0122] The method embodiments provided in this specification can be executed in a mobile terminal, a computer terminal, a server or a similar computing device. Taking running on a server as an example, Fig.18 This is a hardware structure block diagram of an embodiment of a server for determining reservoir porosity based on rock physics knowledge provided in this specification. The server may be the device for determining reservoir porosity based on rock physics knowledge or the system for determining reservoir porosity based on rock physics knowledge in the above embodiments. Fig.18As shown, the server 10 may include one or more (only one is shown in the figure) processors 100 (the processor 100 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 200 for storing data, and a transmission module 300 for communication functions. A person skilled in the art can understand that Fig.18 The structure shown is only for illustration and does not limit the structure of the above electronic device. Fig.18 More or fewer components shown in the figure may also include other processing hardware, such as a database or multi-level cache, GPU, or other hardware with Fig.18 Different configurations are shown.

[0123] The memory 200 can be used to store software programs and modules of application software, such as the program instructions / modules corresponding to the method for determining reservoir porosity based on rock physics knowledge in the embodiment of this specification. The processor 100 executes various functional applications and data processing by running the software programs and modules stored in the memory 200. The memory 200 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 200 may further include a memory remotely arranged relative to the processor 100, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0124] The transmission module 300 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of a computer terminal. In one example, the transmission module 300 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission module 300 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0125] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0126] The methods or devices described in the above embodiments provided in this specification can implement business logic through computer programs and record them on storage media, and the storage media can be read and executed by computers to achieve the effects of the solutions described in the embodiments of this specification. The storage medium may include a physical device for storing information, which is usually to digitize the information and then store it in a medium using electrical, magnetic or optical methods. The storage medium may include: devices that use electrical energy to store information, such as various types of memories, such as RAM, ROM, etc.; devices that use magnetic energy to store information, such as hard disks, floppy disks, magnetic tapes, magnetic core memories, bubble memories, and U disks; devices that use optical methods to store information, such as CDs or DVDs. Of course, there are other types of readable storage media, such as quantum memories, graphene memories, and so on.

[0127] The above-mentioned method or device embodiment for determining reservoir porosity based on rock physics knowledge provided in this specification can be implemented by a processor in a computer executing corresponding program instructions, such as using the C++ language of the Windows operating system to implement it on a PC, a Linux system, or other methods such as using Android or iOS system programming languages ​​to implement it on a smart terminal, as well as based on the processing logic of a quantum computer.

[0128] It should be noted that the devices, equipment, and systems described in the specification may also include other implementation methods according to the description of the relevant method embodiments. The specific implementation methods can refer to the description of the corresponding method embodiments, which will not be described one by one here.

[0129] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

Claims

1. A method for determining reservoir porosity based on rock physics knowledge, characterized in that: include: Acquire a well logging data set of a target lithology category; wherein each well logging data in the well logging data set corresponds to a plurality of well logging curves; and select a porosity index from the well logging data set using a preset method; wherein the porosity index represents a well logging curve that affects the reservoir porosity; Obtaining a porosity estimation curve based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve; Inputting the data corresponding to the porosity index in the well logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining multiple decision regressors based on rock physical information; The integrated porosity prediction model can be expressed as an integrated boosting model with decision tree as the base regressor. Placement: Where: FM(x) represents the predicted porosity result, T(x, φ m ) represents the mth decision regressor, x represents the input data, φ m is the parameter of the decision regressor, M is the total number of decision regressors; The step of obtaining a well logging data set of a target lithology category includes: Acquire an initial well logging data set; wherein each well logging data in the initial well logging data set corresponds to a plurality of well logging curves, and the initial well logging data set includes well logging data of known lithology categories and well logging data of unknown lithology categories; Using a preset method to select lithology indicators from the initial well logging data set; wherein the lithology indicators represent well logging curves that affect reservoir lithology; Based on the data corresponding to the lithology index in the initial well logging data set and the well logging data of known lithology categories, determining the category of the well logging data of unknown lithology categories in the initial well logging data set, and obtaining a first well logging data set; Acquire a well logging data set of a target lithology category from the first well logging data set; The multiple logging curves include acoustic wave time difference curve, well diameter curve, gamma curve, micro-gradient resistivity curve, 2.5-meter bottom gradient resistivity curve, natural potential curve, shale content curve and porosity curve; the lithology indicators include gamma curve, micro-gradient resistivity curve, 2.5-meter bottom gradient resistivity curve, natural potential curve, shale content curve and porosity curve; After obtaining the logging data, first remove the outliers in the logging curve, and then standardize the logging curve according to the following formula: Among them, x * is the well logging curve after standardization, x is the well logging curve before standardization, is the mean value of the logging curve, σ is the standard deviation of the logging curve; The step of determining the category of the logging data of unknown lithology category in the initial logging data set based on the data corresponding to the lithology index in the initial logging data set and the logging data of known lithology category to obtain the first logging data set includes: Based on the well logging data of known lithology categories, determine the number of clusters and the center point of each cluster; Calculate the distance between the data corresponding to the lithology index in the initial well logging data set and the center point of each cluster; Determining the category of the logging data of unknown lithology category in the initial logging data set according to the distance to the center point of each cluster, and obtaining a first logging data set; The porosity indicators include acoustic time difference curve, wellbore curve, gamma curve, micro-gradient resistivity curve, 2.5-meter bottom gradient resistivity curve and natural potential curve; The step of obtaining a porosity estimation curve based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve comprises: Establishing a rock physics model based on the data corresponding to the specified porosity index in the well logging data set and the porosity curve, wherein the rock physics is used to estimate the porosity; The rock physics model is: Por=0.527×AC+0.0855 (2) Wherein, Por represents the porosity of the sandstone section, AC represents the acoustic time difference of the sandstone section; The rock physics model is used to estimate porosity; Inputting data corresponding to a specified porosity index in the well logging data set into the rock physics model to obtain a porosity estimation curve; After obtaining the well logging data set of the target lithology category from the first well logging data set, the method further includes: Acquire a well logging data set of a non-target lithology category from the first well logging data set; Setting the value of the porosity curve in the well logging data set of the non-target lithology category to obtain the porosity of the reservoir corresponding to the non-target lithology category; After obtaining the porosity of the reservoir corresponding to the target lithology category, the method further includes: The porosity of the reservoir corresponding to the target lithology category and the porosity of the reservoir corresponding to the non-target lithology category are combined to obtain the phase-controlled porosity of the target reservoir.

2. A device used in the method for determining reservoir porosity based on rock physics knowledge as claimed in claim 1, characterized in that: include: An acquisition module is used to acquire a logging data set of a target lithology category; wherein each logging data in the logging data set corresponds to a plurality of logging curves; a screening module is used to screen a porosity index from the logging data set in a preset manner; wherein the porosity index represents a logging curve that affects the reservoir porosity; a first acquisition module is used to obtain a porosity estimation curve based on data and a porosity curve corresponding to a specified porosity index in the logging data set; a second acquisition module is used to input data corresponding to the porosity index in the logging data set and the porosity estimation curve into an integrated porosity prediction model to obtain the porosity of the reservoir corresponding to the target lithology category; wherein the integrated porosity prediction model is obtained by constraining training of multiple decision regressors based on rock physical information.

3. A device for determining reservoir porosity based on rock physics knowledge, characterized in that: The method comprises at least one processor and a memory storing computer executable instructions, wherein the method of claim 1 is implemented when the processor executes the instructions.

Images