Shale oil reservoir permeability calculation method, device and equipment based on combined driving

Abstract

The invention discloses a shale oil reservoir permeability calculation method, device and equipment based on combined driving, and relates to the field of oil-gas exploration, the method comprises the following steps: constructing a permeability physical model, and constructing a data set by using logging data and core experiment permeability data; screening target logging data from the logging data in the data set; calculating according to the permeability physical model and nuclear magnetic logging data in the logging data to obtain fitting permeability; taking the target logging data and the fitting permeability as sample data, and constructing label data by using core experiment permeability data to form a training sample set; according to the training sample set, carrying out model training by adopting a random forest algorithm to obtain a permeability intelligent model; and inputting the target logging data corresponding to the shale oil reservoir to be measured and the fitting permeability into the permeability intelligent model, and predicting the permeability of the shale oil reservoir to be measured. According to the method, the interpretation precision of the permeability of the shale oil reservoir is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of oil and gas exploration, and in particular to a shale oil reservoir permeability calculation method, device and equipment based on combined drive. Background Art

[0002] In shale oil formation evaluation, permeability logging interpretation faces challenges. The commonly used interpretation methods include the following three: 1) Mathematical statistics method: Based on rock physics experiments, empirical formulas of permeability and porosity are constructed. Most of the empirical formulas of permeability and porosity are nonlinear, and the fitting accuracy is low for tight sandstone and shale oil and gas reservoirs with low permeability, making them difficult to promote and apply.

[0003] 2) Rock physics model or knowledge model: The common model is the Timur formula, which is a function of porosity and irreducible water saturation. Porosity is in the numerator and irreducible water is in the denominator. It increases with the increase of porosity and decreases with the increase of irreducible water content, which has a clear physical meaning. In addition, for nuclear magnetic resonance logging, the Geo-Coates model was developed, which is a function of the porosity of nuclear magnetic components. The numerator is effective porosity and the denominator is clay water porosity. It increases with the increase of porosity and decreases with the increase of clay irreducible water content, which has a clear physical meaning. Moreover, the above two models are fixed in form and are both basic formulas for logging interpretation, which can be regarded as knowledge models. However, the parameters in the Timur formula and Geo-Coates model need to be refitted by rock physics experiments, and sometimes the fitting accuracy is low, which is difficult to meet the requirements of use.

[0004] 3) Machine learning method: Under the condition of sufficient rock physics experimental data, a training set consisting of logging data and label data is constructed, and a suitable intelligent algorithm is selected to construct a prediction model through repeated training. However, the machine learning method is supervised learning, which requires a relatively high amount of permeability rock physics experimental data, otherwise the amount of labels is too small to be generalized; and because the prediction of permeability parameters is carried out one by one, if there is no geological and physical model to constrain it, the prediction results are likely to exceed the geological and physical rules and it is difficult to meet the use requirements.

[0005] Shale oil and gas reservoirs are heterogeneous, with low porosity and low permeability. The relationship between their permeability and logging data is mostly nonlinear, which leads to the inability of traditional empirical formulas and knowledge theoretical models to meet actual needs. Therefore, how to accurately evaluate the permeability of shale oil reservoirs has become an urgent problem to be solved. Summary of the invention

[0006] In view of the above problems, the present invention is proposed to provide a shale oil reservoir permeability calculation method, device and equipment based on combined drive that overcome the above problems or at least partially solve the above problems.

[0007] According to one aspect of an embodiment of the present application, a method for calculating shale oil reservoir permeability based on joint driving is provided, the method comprising: Construct a permeability physical model and build a data set using well logging data and core experiment permeability data; Filter out target logging data from the logging data in the data set; Calculate the corresponding fitting permeability based on the permeability physical model and the nuclear magnetic logging data in the logging data; The target logging data and fitted permeability are used as sample data, and the core experimental permeability data are used to construct label data corresponding to the sample data to form a training sample set; Based on the training sample set, the random forest algorithm is used to train the model and obtain the permeability intelligent model; The target logging data and fitted permeability corresponding to the shale oil reservoir to be tested are input into the permeability intelligent model, and the permeability of the shale oil reservoir to be tested is predicted by the permeability intelligent model.

[0008] Furthermore, constructing the permeability physical model further includes: According to the permeability evaluation model established based on nuclear magnetic resonance logging data, the permeability physical model is determined; wherein the permeability physical model includes empirical coefficients of the study area to be determined; The empirical coefficients of the study area in the permeability physical model are determined using the permeability experimental data of the study area.

[0009] Furthermore, the calculation formula of the permeability physical model is:

[0010] Among them, K is the permeability, the unit is mD; m, n and C are the empirical coefficients of the study area, which are decimals; is the effective porosity of NMR, in %; is the porosity of movable fluid, in %; is the bound fluid porosity, in %.

[0011] Furthermore, constructing a data set using well logging data and core test permeability data further includes: The core sample points corresponding to the core experimental permeability data and the well logging data are depth corrected, and the depth-corrected well logging data and core experimental permeability data are added to the data set.

[0012] Furthermore, before selecting the target well logging data from the well logging data in the data set, the method further includes: Each data in the dataset is preprocessed to remove outliers and standardize the data.

[0013] Furthermore, filtering out target well logging data from the well logging data in the data set further includes: Calculate the Pearson correlation coefficient between each well logging data in the data set and the core experimental permeability data; The Pearson correlation coefficient between each well logging data and the core experimental permeability data is displayed in the form of a heat map, and the well logging data with a Pearson correlation coefficient higher than a preset value are screened out from each well logging data as the target well logging data.

[0014] According to another aspect of an embodiment of the present application, a shale oil reservoir permeability calculation device based on joint driving is provided, the device comprising: A construction module, suitable for constructing a permeability physical model and constructing a data set using well logging data and core experiment permeability data; A screening module, adapted to screen out target well logging data from the well logging data in the data set; A fitting module is used to calculate according to the permeability physical model and the nuclear magnetic logging data in the logging data to obtain the corresponding fitting permeability; The sample processing module is suitable for taking the target logging data and the fitted permeability as sample data, and constructing the label data corresponding to the sample data by using the core experimental permeability data to form a training sample set; A training module, suitable for training a model using a random forest algorithm based on a training sample set to obtain a permeability intelligent model; The prediction module is suitable for inputting the target logging data and the fitted permeability corresponding to the shale oil reservoir to be tested into the permeability intelligent model, and the permeability intelligent model predicts the permeability of the shale oil reservoir to be tested.

[0015] According to another aspect of the embodiment of the present application, there is provided a computing device, including: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned shale oil reservoir permeability calculation method based on combined drive.

[0016] According to another aspect of the embodiment of the present application, a computer storage medium is provided, in which at least one executable instruction is stored, and the executable instruction enables a processor to perform operations corresponding to the above-mentioned shale oil reservoir permeability calculation method based on combined drive.

[0017] According to another aspect of an embodiment of the present application, a computer program product is provided, comprising at least one executable instruction, wherein the executable instruction enables a processor to perform operations corresponding to the above-mentioned method for calculating the permeability of a shale oil reservoir based on combined drive.

[0018] According to the technical solution provided by the present invention, a new physical model of permeability is constructed, and the machine learning of the logging data is combined for "double-wheel" driving, and a permeability intelligent model based on the joint drive of the physical model and the logging data is established; in addition, conventional logging data and nuclear magnetic logging data that are sensitive to permeability are screened out from the data set as target logging data, and the target logging data and the fitted permeability calculated using the physical model of permeability are used as input data for permeability prediction, so that the input data includes both the target logging data that is sensitive to permeability and the fitted permeability, thereby realizing the diversification of the model input data; compared with the existing Compared with the existing technology that only uses conventional logging data as input data, this solution makes full use of nuclear magnetic logging data, and uses both conventional logging data and nuclear magnetic logging data as input data. The fitted permeability obtained by fitting the nuclear magnetic resonance component porosity data and the permeability physical model in the nuclear magnetic resonance logging data is also used as input data. The random forest algorithm is used for model training. The permeability intelligent model obtained through training can predict the permeability of shale oil reservoirs conveniently, accurately and efficiently, greatly improving the interpretation accuracy of shale oil reservoir permeability. This solution has a wide range of applications and optimizes the permeability prediction method.

[0019] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings: Figure 1 A schematic flow chart of a method for calculating shale oil reservoir permeability based on combined driving according to an embodiment of the present application is shown; Figure 2 A schematic diagram showing the principle of the random forest algorithm is shown; Figure 3 A comparison chart of permeability prediction errors using the Coates model, random forest, and the permeability intelligent model of the present application based on the joint drive of the physical model and logging data is shown for a certain exploration well; Figure 4 A comparison chart showing the permeability calculated by the permeability intelligent model based on the physical model and well logging data jointly driven by the present application and the results calculated by other methods for a certain exploration well is shown; Figure 5 A structural block diagram of a shale oil reservoir permeability calculation device based on joint drive according to an embodiment of the present application is shown; Figure 6 A schematic diagram of the structure of a computing device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0021] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0022] Figure 1 A schematic diagram of a process flow of a shale oil reservoir permeability calculation method based on combined driving according to an embodiment of the present application is shown. Figure 1 As shown, the method comprises the following steps: Step S101, constructing a permeability physical model, and constructing a data set using well logging data and core test permeability data.

[0023] Taking into account that physical models and data-driven methods have both advantages and disadvantages, the two methods are combined in the embodiments of the present application to achieve a "dual-wheel" drive of the model and data, and to establish an intelligent permeability model based on the joint drive of the physical model and logging data.

[0024] The permeability physical model is constructed as follows: the permeability physical model is determined according to the permeability evaluation model established based on nuclear magnetic resonance logging data; the permeability physical model includes empirical coefficients of the study area to be determined; and then the empirical coefficients of the study area in the permeability physical model are determined using the permeability experimental data of the study area.

[0025] The object of measurement of nuclear magnetic resonance logging is the fluid in the reservoir pores. Under certain conditions, it is not affected by the mineral composition of the rock, and its pore type, pore size, connectivity and other characteristics can be obtained. Based on various nuclear magnetic resonance experiments and theoretical derivations, many permeability prediction models based on nuclear magnetic characteristic parameters have been established. The Coates model is a relatively classic permeability evaluation model and is widely used. In actual application scenarios, nuclear magnetic resonance logging data is needed to perform regional permeability modeling.

[0026] In the embodiment of the present application, the Coates model is selected as the permeability evaluation model, and the Coates physical model is fitted according to the study area. The permeability is fitted using the formula of the Coates model, and the empirical coefficients m, n and C of the study area are determined by fitting to obtain the permeability physical model of the present application. Specifically, according to the component porosity cutoff value determined by the rock nuclear magnetic resonance experiment, the volume of clay water, bound water, and movable fluid is calculated, and the permeability physical model based on the Coates formula is fitted through the core permeability experiment. The present application uses the nuclear magnetic resonance component porosity data to construct a new permeability physical model and its coefficients.

[0027] Among them, the Coates model is obtained by mathematical fitting, and the formula of the Coates model is:

[0028] Among them, K is the permeability, the unit is mD; m, n and C are the empirical coefficients of the study area, which need to be calibrated by cores; is the effective porosity of NMR, in %; is the porosity of movable fluid, in %; is the bound fluid porosity, in %.

[0029] Take the logarithm of both sides of the Coates model formula as the permeability physical model. The calculation formula of the permeability physical model is:

[0030] Among them, K is the permeability, the unit is mD; m, n and C are the empirical coefficients of the study area, which are decimals; is the effective porosity of NMR, in %; is the porosity of movable fluid, in %; is the bound fluid porosity, in %.

[0031] Using the above formula for the permeability experimental data of a certain study area, the empirical coefficients m, n and C of the study area can be obtained, thereby obtaining the permeability physical model.

[0032] In an embodiment of the present application, a data set is constructed using well logging data and core test permeability data. Among them, the well logging data includes conventional well logging data and nuclear magnetic logging data. Conventional well logging data provides basic information on reservoir lithology, physical properties and electrical properties, while nuclear magnetic logging data has the advantage of its high sensitivity to fluid properties (especially oil saturation). Core test permeability data is one of the key parameters for evaluating underground reservoir performance. It reflects the ability of fluids (such as oil, gas or water) to pass through rocks. These data are crucial for understanding the production capacity of reservoirs, designing development plans, and predicting the long-term performance of oil and gas fields. The core sample points corresponding to the core test permeability data are depth-corrected with the well logging data, and the depth-corrected well logging data and core test permeability data are added to the data set to complete the construction of the data set.

[0033] Specifically, conventional logging data may include natural gamma logging, acoustic transit time logging, neutron logging, density logging, etc. Nuclear magnetic logging is an advanced geophysical logging technology used to evaluate key parameters such as formation porosity, fluid type and distribution. It obtains information by using the behavior of hydrogen nuclei (protons) in rock pores in a strong magnetic field. Nuclear magnetic logging data includes relaxation time data, nuclear magnetic resonance component porosity data, fluid property identification data, etc.

[0034] The method also includes: preprocessing each data in the data set to remove outliers and standardize the data. Data preprocessing is to perform reasonable optimization processing on the data in the data set, which is an important step before logging interpretation. The main contents of preprocessing include outlier removal and data standardization.

[0035] In actual application scenarios, the collected logging data has different data lengths and dimensions, and there are certain outliers, that is, dirty data, which need to be cleaned and unified. Among them, for outlier and missing value processing, for missing values ​​in the logging curve where the data is NULL (i.e., null value), the sample can be directly deleted. For outliers in the data, the Laida rule can be used to judge the outliers in the logging curve data, and the outliers are replaced by the mean of each logging curve. The Laida rule is a statistical method based on normal distribution, which is used to detect outliers in a data set.

[0036] Before training intelligent algorithms such as machine learning, it is necessary to pre-process the logging data so that the logging data between different wells can be quantitatively comparable horizontally. In actual application scenarios, data standardization can use the Min-Max deviation standardization method to linearly change the original logging data so that the results are mapped to the [0,1] interval, thereby achieving data standardization.

[0037] Step S102: Filter out target logging data from the logging data in the data set.

[0038] After the outlier removal and data standardization are completed, sensitive logging data that is strongly correlated with permeability can be screened out from the logging data in the data set as target logging data for constructing a training sample set, thereby achieving the optimization of feature input. In step S102, the Pearson correlation coefficient can be selected as an evaluation indicator, and the correlation relationship can be intuitively displayed in the form of a correlation heat map to facilitate the selection of feature input.

[0039] Specifically, the Pearson correlation coefficient between each well logging data and the core experiment permeability data in the data set is calculated; the Pearson correlation coefficient between each well logging data and the core experiment permeability data is displayed in the form of a heat map, and the well logging data with a Pearson correlation coefficient higher than a preset value is screened out from each well logging data as the target well logging data, so that the target well logging data can be subsequently used as the feature input of the permeability intelligent model.

[0040] The Pearson Correlation Coefficient is a statistical indicator that measures the degree of linear correlation between two continuous variables. Its value range is [-1,1] and is used to quantify the trend of coordinated changes between variables. The quotient of the covariance and standard deviation between two variables is the Pearson coefficient. The overall correlation coefficient is usually defined as The specific expression is:

[0041] in, is the covariance between variable X and variable Y; is the expected value of variable X; is the expected value of variable Y; is the standard deviation of variable X; is the standard deviation of variable Y.

[0042] Calculate the data sample covariance and standard deviation to obtain the Pearson correlation coefficient r:

[0043] Where r is the Pearson correlation coefficient; n is the number of samples; is the sample mean of variable X; is the sample mean of variable Y; is the value of the i-th sample in variable X; is the value of the i-th sample in the variable Y.

[0044] After the calculation of the Pearson correlation coefficient between each well logging data and the core test permeability data is completed, the correlation relationship between each well logging data and the core test permeability data is displayed in the form of a correlation heat map to guide the optimization of feature input.

[0045] For conventional logging data, natural gamma logging (GR) data, acoustic logging (AC) data, density logging (DEN) data and neutron logging (CNL) data can be selected as target logging data as feature input for permeability prediction. For nuclear magnetic resonance logging data, nuclear magnetic resonance component porosity data can be selected as target logging data.

[0046] Step S103, performing calculations based on the permeability physical model and the nuclear magnetic logging data in the logging data to obtain the corresponding fitting permeability.

[0047] For the nuclear magnetic resonance logging data, three parameter data, namely, nuclear magnetic resonance effective porosity, movable fluid porosity and bound fluid porosity, can be selected and the fitted permeability can be calculated using the constructed permeability physical model as the characteristic input for permeability prediction.

[0048] Step S104: taking the target logging data and the fitted permeability as sample data, and using the core experiment permeability data to construct label data corresponding to the sample data, to form a training sample set.

[0049] The target logging data and fitted permeability are used as sample data for feature input of permeability prediction. The core experimental permeability data are used to construct label data corresponding to the sample data, that is, the label data required for permeability prediction, thereby completing the construction of the training sample set required for model training.

[0050] Step S105, based on the training sample set, a random forest algorithm is used to perform model training to obtain a permeability intelligent model.

[0051] The basis of the random forest algorithm is the decision tree. The core idea is to construct a new sample set by extracting samples with replacement and randomly extracting features without replacement, and then train multiple CART decision trees to finally form a random forest model. The processing flow is as follows: First, it is necessary to clarify the training parameters of random forests, which include but are not limited to the potential subsets of input features, feature attribute sets, the number of decision trees, and the number of features randomly extracted each time. The setting of these parameters directly affects the performance and generalization ability of the model. The algorithm will extract multiple sub-sample sets (i.e., subsets) from the original training sample set with replacement. For each sub-sample set, the algorithm will further randomly extract its features, and then train the corresponding decision tree based on these features and samples. Figure 2 A schematic diagram of the principle of the random forest algorithm is shown, as Figure 2 As shown in the figure, all decision trees will jointly predict or classify the new data and summarize the results of each tree to output the final prediction or classification result.

[0052] After the random forest algorithm is constructed, parameters such as the number of iterations, maximum tree depth, and minimum number of node samples are adjusted, and the optimal parameters are selected for model training. Based on the data set, the target logging data and the fitted permeability are used as feature inputs. That is to say, on the basis of conventional logging data, the embodiment of the present application also adds the fitted permeability calculated by the permeability physical model as a feature input for constraint; the permeability intelligent model is obtained by training using the random forest algorithm and optimal parameters, thereby completing the training of the permeability intelligent model driven jointly by the physical model and logging data. The permeability intelligent model obtained through training can conveniently, accurately, and efficiently predict the permeability of shale oil reservoirs.

[0053] Step S106, inputting the target logging data and the fitted permeability corresponding to the shale oil reservoir to be tested into the permeability intelligent model, and using the permeability intelligent model to predict the permeability of the shale oil reservoir to be tested.

[0054] When it is necessary to predict the permeability of a shale oil reservoir to be tested, the target logging data corresponding to the shale oil reservoir to be tested is obtained, and calculations are performed based on the permeability physical model and the nuclear magnetic resonance logging data of the shale oil reservoir to be tested, and the corresponding fitting permeability is obtained by fitting; the target logging data and the fitting permeability corresponding to the shale oil reservoir to be tested are input into the permeability intelligent model, and the permeability intelligent model is used to predict the permeability of the shale oil reservoir to be tested, and the permeability prediction result of the shale oil reservoir to be tested is output.

[0055] The following takes the Weixinan Sag as the study area to apply the method provided in this application and analyze its application effect.

[0056] The Eocene Liushagang Formation is the main source rock system in the depression. The sandstones of the bottom flow section 3 and the top flow section 1 are the main reservoirs, and the deep lake mudstone of the middle flow section 2 is the main source rock. Due to the complex geological structure and strong reservoir heterogeneity in this area, the traditional conventional logging interpretation has a low compliance rate and is difficult to meet the requirements of reservoir evaluation. To this end, the shale oil reservoir permeability calculation method based on the joint drive of physical model and logging data proposed in this application is used to carry out permeability prediction. For the study area of ​​the Weixinan Depression, after multivariate linear fitting, the formula expression of the fitted Coates model is obtained as follows:

[0057] Among them, K is permeability, unit is mD; after core calibration, m=-0.2673, n=2.5361, C=1.8; is the effective porosity of NMR, in %; is the porosity of movable fluid, in %; is the bound fluid porosity, in %.

[0058] For a certain exploration well in the Southwest Wei Sag, the target layers are the lower sequence of the second section of the flow and the upper sequence of the third section of the flow, and the main lithologies are shale and oil shale. The logging project includes conventional logging series and nuclear magnetic resonance logging. The random forest algorithm is selected as the intelligent algorithm, which is the basis of the data-driven algorithm; on the basis of conventional logging data, the fitted permeability obtained by using the permeability physical model is added as the feature input, which is the basis of the physical model drive. The data-driven algorithm is organically combined with the permeability physical model driven by the physical model to form a shale oil reservoir permeability calculation method based on the joint drive of the physical model and logging data.

[0059] For this exploration well, the Coates model, random forest, and the permeability intelligent model based on the physical model and logging data jointly driven by this application are used to predict the permeability. The permeability prediction error comparison diagram is shown in the figure below. Figure 3 As shown, the logarithmic error of the permeability calculated by the Coates model is 0.47, and the logarithmic error of the permeability calculated by the random forest model is 0.41; RF-Coates joint drive refers to the permeability intelligent model based on the joint drive of physical model and logging data proposed in this application, which adds the fitted permeability obtained by the permeability physical model to the random forest model for constraint, and the calculated permeability logarithmic error is 0.33. Figure 3 It can be seen intuitively that the logarithmic error of the permeability predicted by a single random forest model or Coates model is relatively large, while when the fitted permeability obtained by using the permeability physical model is added as a feature input constraint on the basis of the random forest model, the error is significantly reduced. It can be seen that the shale oil reservoir permeability calculation method based on the joint drive of the physical model and logging data proposed in this application can significantly improve the interpretation accuracy of the shale oil reservoir permeability.

[0060] Figure 4A comparison chart showing the calculation results of the permeability of a certain exploration well using the permeability intelligent model based on the joint drive of the physical model and logging data of this application and other methods, wherein "data-model joint calculation of permeability" refers to the calculation of the permeability of the exploration well using the shale oil reservoir permeability calculation method based on the joint drive of the physical model and logging data proposed in this application. For conventional logging data, natural gamma ray logging (GR) data, acoustic logging (AC) data, density logging (DEN) data and neutron logging (CNL) data can be selected as target logging data as feature inputs for permeability prediction. For nuclear magnetic resonance logging data, three parameter data, namely nuclear magnetic resonance effective porosity, movable fluid porosity and bound fluid porosity, can be selected, and the constructed permeability physical model can be used to calculate the fitted permeability as the feature input for permeability prediction. As shown Figure 4 As shown in the figure, the permeability errors calculated by the two single methods, Coates model and random forest, are relatively large, with the logarithmic errors of the calculated permeabilities being 0.48 and 0.43 respectively. After adding the fitted permeability obtained by the permeability physical model as a feature input constraint on the basis of the random forest model, the logarithmic error of the permeability obtained is reduced to 0.31. At the depth point where there are core test samples, it can be seen that the shale oil reservoir permeability calculation method proposed in this application based on the joint drive of physical model and logging data predicts a permeability that is more consistent with the core test permeability, and this scheme has significant advantages.

[0061] In addition, the above method was used to process and interpret data of several other exploration wells in the study area. The interpretation results were compared with the core test results, and the logarithmic error range of the permeability of this scheme was 0.30 to 0.38. It can be seen that this scheme is accurate and feasible, and its effect has obvious advantages over traditional methods.

[0062] According to the shale oil reservoir permeability calculation method based on joint drive provided in the embodiment of the present application, a new permeability physical model is constructed, and the machine learning of the logging data is combined for "double-wheel" drive, and a permeability intelligent model based on the joint drive of the physical model and the logging data is established; and conventional logging data and nuclear magnetic logging data that are sensitive to permeability are screened out from the data set as target logging data, and the target logging data and the fitted permeability calculated using the permeability physical model are used as input data for permeability prediction, so that the input data includes both the target logging data that is sensitive to permeability and the fitted permeability, thereby realizing the model input Diversification of data; compared with the existing technology that only uses conventional logging data as input data, this solution makes full use of nuclear magnetic logging data, and uses both conventional logging data and nuclear magnetic logging data as input data. The fitted permeability obtained by fitting the nuclear magnetic resonance component porosity data and the permeability physical model in the nuclear magnetic logging data is also used as input data. The random forest algorithm is used for model training. The permeability intelligent model obtained through training can predict the permeability of shale oil reservoirs conveniently, accurately and efficiently, greatly improving the interpretation accuracy of shale oil reservoir permeability. This solution has a wide range of applications and optimizes the permeability prediction method.

[0063] Figure 5 FIG. 1 shows a structural block diagram of a shale oil reservoir permeability calculation device based on joint drive according to an embodiment of the present application. Figure 5 As shown, the device includes: a construction module 510, a screening module 520, a fitting module 530, a sample processing module 540, a training module 550 and a prediction module 560.

[0064] The construction module 510 is adapted to: construct a permeability physical model, and construct a data set using well logging data and core test permeability data.

[0065] The screening module 520 is adapted to screen out target well logging data from the well logging data in the data set.

[0066] The fitting module 530 is adapted to perform calculations based on the permeability physical model and the nuclear magnetic logging data in the logging data to obtain the corresponding fitting permeability.

[0067] The sample processing module 540 is adapted to: use the target logging data and the fitted permeability as sample data, and construct label data corresponding to the sample data using the core experiment permeability data to form a training sample set.

[0068] The training module 550 is suitable for: performing model training using a random forest algorithm according to a training sample set to obtain a permeability intelligent model.

[0069] The prediction module 560 is suitable for: inputting the target logging data and the fitted permeability corresponding to the shale oil reservoir to be tested into the permeability intelligent model, and the permeability intelligent model predicts the permeability of the shale oil reservoir to be tested.

[0070] Optionally, the construction module 510 is further suitable for: determining a permeability physical model based on a permeability evaluation model established based on nuclear magnetic resonance logging data; wherein the permeability physical model includes empirical coefficients of the study area to be determined; and determining the empirical coefficients of the study area in the permeability physical model using the permeability experimental data of the study area.

[0071] Optionally, the calculation formula of the permeability physical model is:

[0072] Among them, K is the permeability, the unit is mD; m, n and C are the empirical coefficients of the study area, which are decimals; is the effective porosity of NMR, in %; is the porosity of movable fluid, in %; is the bound fluid porosity, in %.

[0073] Optionally, the construction module 510 is further adapted to: perform depth correction on the core sample points and the well logging data corresponding to the core experiment permeability data, and add the depth-corrected well logging data and the core experiment permeability data to the data set.

[0074] Optionally, the construction module 510 is further adapted to: preprocess each data in the data set to remove outliers and standardize the data.

[0075] Optionally, the screening module 520 is further adapted to: calculate the Pearson correlation coefficient between each well logging data in the data set and the core experiment permeability data; display the Pearson correlation coefficient between each well logging data and the core experiment permeability data in the form of a heat map, and screen out the well logging data with a Pearson correlation coefficient higher than a preset value from each well logging data as the target well logging data.

[0076] According to the shale oil reservoir permeability calculation device based on joint drive provided in the embodiment of the present application, a new permeability physical model is constructed, and the machine learning of the logging data is combined for "double-wheel" drive, and a permeability intelligent model based on the joint drive of the physical model and the logging data is established; and conventional logging data and nuclear magnetic logging data that are sensitive to permeability are screened out from the data set as target logging data, and the target logging data and the fitted permeability calculated using the permeability physical model are used as input data for permeability prediction, so that the input data includes both the target logging data that is sensitive to permeability and the fitted permeability, thereby realizing the model input Diversification of data; compared with the existing technology that only uses conventional logging data as input data, this solution makes full use of nuclear magnetic logging data, and uses both conventional logging data and nuclear magnetic logging data as input data. The fitted permeability obtained by fitting the nuclear magnetic resonance component porosity data and the permeability physical model in the nuclear magnetic logging data is also used as input data. The random forest algorithm is used for model training. The permeability intelligent model obtained through training can predict the permeability of shale oil reservoirs conveniently, accurately and efficiently, greatly improving the interpretation accuracy of shale oil reservoir permeability. This solution has a wide range of applications and optimizes the permeability prediction method.

[0077] The present invention also provides a non-volatile computer storage medium, which stores at least one executable instruction, and the executable instruction can execute the shale oil reservoir permeability calculation method based on combined drive in any of the above method embodiments.

[0078] An embodiment of the present invention provides a computer program product, which includes at least one executable instruction or computer program, which enables a processor to perform operations corresponding to the shale oil reservoir permeability calculation method based on combined drive in any of the above-mentioned method embodiments.

[0079] Figure 6 A schematic diagram of the structure of a computing device according to an embodiment of the present application is shown. The specific embodiment of the present application does not limit the specific implementation of the computing device.

[0080] like Figure 6 As shown, the computing device may include: a processor 602 , a communication interface 604 , a memory 606 , and a communication bus 608 .

[0081] The processor 602, the communication interface 604, and the memory 606 communicate with each other via a communication bus 608. The communication interface 604 is used to communicate with other devices such as a client or other server network elements. The processor 602 is used to execute a program 610, which can specifically execute the relevant steps in the embodiment of the shale oil reservoir permeability calculation method based on joint drive for the computing device.

[0082] Specifically, the program 610 may include program codes, which include computer operation instructions.

[0083] The processor 602 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in the computing device may be processors of the same type, such as one or more CPUs; or may be processors of different types, such as one or more CPUs and one or more ASICs.

[0084] The memory 606 is used to store the program 610. The memory 606 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0085] Program 610 can be specifically used to enable processor 602 to execute the shale oil reservoir permeability calculation method based on combined drive in any of the above-mentioned method embodiments. The specific implementation of each step in program 610 can refer to the corresponding descriptions in the corresponding steps and units in the above-mentioned shale oil reservoir permeability calculation embodiment based on combined drive, which will not be repeated here. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described equipment and modules can refer to the corresponding process description in the above-mentioned method embodiment, which will not be repeated here.

[0086] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific languages ​​is for disclosing the best mode of the present invention.

[0087] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0088] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the intention that the claimed invention requires more features than those expressly recited in each claim. Rather, as reflected in the claims, inventive aspects lie in less than all of the features of the individual embodiments previously disclosed. Therefore, the claims that follow the detailed description are hereby expressly incorporated into the detailed description, with each claim itself serving as a separate embodiment of the present invention.

[0089] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device so disclosed may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0090] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present invention and form different embodiments. For example, in the claims, any one of the claimed embodiments may be used in any combination.

[0091] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the embodiments of the present invention. The present invention may also be implemented as a device or apparatus program (e.g., a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0092] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

Claims

1. A method for calculating shale oil reservoir permeability based on combined driving, characterized in that: The method comprises: Construct a permeability physical model and build a data set using well logging data and core experiment permeability data; Filter out target well logging data from the well logging data in the data set; Calculating according to the permeability physical model and the nuclear magnetic logging data in the logging data to obtain the corresponding fitting permeability; The target well logging data and the fitted permeability are used as sample data, and the core experimental permeability data are used to construct label data corresponding to the sample data to form a training sample set; According to the training sample set, a random forest algorithm is used to perform model training to obtain a permeability intelligent model; The target logging data and fitted permeability corresponding to the shale oil reservoir to be tested are input into the permeability intelligent model, and the permeability of the shale oil reservoir to be tested is predicted by the permeability intelligent model.

2. The method for calculating shale oil reservoir permeability based on combined driving according to claim 1, characterized in that: The constructing of the permeability physical model further comprises: Determine a permeability physical model based on a permeability evaluation model established based on nuclear magnetic resonance logging data; wherein the permeability physical model includes empirical coefficients of the study area to be determined; The empirical coefficients of the study area in the permeability physical model are determined using the permeability experimental data of the study area.

3. The method for calculating shale oil reservoir permeability based on combined driving according to claim 2, characterized in that: The calculation formula of the permeability physical model is: Among them, K is the permeability; m, n and C are the empirical coefficients of the study area; is the NMR effective porosity; is the movable fluid porosity; is the bound fluid porosity.

4. The method for calculating shale oil reservoir permeability based on combined driving according to claim 1, characterized in that: The use of well logging data and core test permeability data to construct a data set further includes: The core sample points corresponding to the core experimental permeability data and the well logging data are depth corrected, and the well logging data and core experimental permeability data after depth correction are added to the data set.

5. The method for calculating shale oil reservoir permeability based on combined driving according to claim 1, characterized in that: Before selecting the target well logging data from the well logging data in the data set, the method further includes: Each data in the data set is preprocessed to remove outliers and standardize the data.

6. The method for calculating shale oil reservoir permeability based on combined driving according to any one of claims 1 to 5, characterized in that: The step of selecting target well logging data from the well logging data in the data set further comprises: Calculating the Pearson correlation coefficient between each well logging data in the data set and the core experimental permeability data; The Pearson correlation coefficient between each well logging data and the core experimental permeability data is displayed in the form of a heat map, and the well logging data with the Pearson correlation coefficient higher than a preset value is screened out from each well logging data as the target well logging data.

7. A shale oil reservoir permeability calculation device based on joint drive, characterized in that: The device comprises: A construction module, suitable for constructing a permeability physical model and constructing a data set using well logging data and core experiment permeability data; A screening module, adapted to screen out target well logging data from the well logging data in the data set; A fitting module, adapted to perform calculations based on the permeability physical model and the nuclear magnetic logging data in the logging data to obtain a corresponding fitting permeability; A sample processing module, adapted to use the target logging data and the fitted permeability as sample data, and construct label data corresponding to the sample data using the core experiment permeability data to form a training sample set; A training module, adapted to perform model training using a random forest algorithm according to the training sample set to obtain a permeability intelligent model; The prediction module is suitable for inputting the target logging data and the fitted permeability corresponding to the shale oil reservoir to be tested into the permeability intelligent model, and the permeability intelligent model predicts the permeability of the shale oil reservoir to be tested.

8. A computing device, characterized in that include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the shale oil reservoir permeability calculation method based on combined drive as described in any one of claims 1-6.

9. A computer storage medium, characterized in that The computer storage medium stores at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the shale oil reservoir permeability calculation method based on combined drive as described in any one of claims 1-6.

10. A computer program product, characterized in that It includes at least one executable instruction, which enables the processor to perform operations corresponding to the shale oil reservoir permeability calculation method based on combined drive as described in any one of claims 1-6.

Images