Method and device for predicting recovery reserves of shale gas well

Abstract

The invention provides a method for predicting recovery reserves of a shale gas well, and relates to the technical field of exploration. The method comprises the following steps: acquiring parameter data of a shale gas well, wherein the parameter data comprises geological parameter data, engineering parameter data and production parameter data; and performing a prediction operation based on the parameter data and a prediction model to obtain a prediction result, the prediction model being obtained by training based on training sample data, and the training sample data being obtained based on the parameter type of the parameter data.

Description

Technical Field

[0001] The present application relates to the field of exploration technology, and in particular to a method and device for predicting the production and reserves of a shale gas well. Background Art

[0002] The world's shale gas reserves are abundant and have great development potential. The shale gas well recovery reserves, namely EUR (Estimated Ultimate Recovery), are a key indicator for evaluating the technical risks and economic effects of exploration projects, and play an important role in the exploration and development of shale gas.

[0003] In the prior art, different prediction methods are usually used for EUR prediction for different shale gas wells. For old shale gas wells with a long production history and preliminarily clear production dynamic laws, the production dynamic analysis method is generally used; while for new shale gas wells with a short production history and unclear production dynamic laws, the analogy method is used.

[0004] In practical applications, no matter which of the above methods is used, manual analysis is required after data collection. In this process, the total shale gas reserves need to be calculated manually. Since the entire calculation process is relatively complex and tedious and the amount of calculation is large, it will lead to excessive manpower consumption in the process of predicting shale gas reserves. Summary of the invention

[0005] The purpose of the embodiments of the present application is to provide a method and device for predicting the production and reserves of shale gas wells, so as to solve the problem of excessive manual labor consumption in predicting the production and reserves of shale gas in the prior art.

[0006] In order to solve the above technical problems, the embodiments of the present application provide the following technical solutions:

[0007] The first aspect of the present application provides a method for predicting the reserves of a shale gas well, comprising:

[0008] Obtain parameter data of shale gas wells, including geological parameter data, engineering parameter data and production parameter data;

[0009] A prediction operation is performed based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is trained based on the training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0010] A second aspect of the present application provides a device for predicting the production and reserves of a shale gas well, comprising:

[0011] An acquisition unit, used to acquire parameter data of shale gas wells, the parameter data including geological parameter data, engineering parameter data and production parameter data;

[0012] The execution unit is used to perform a prediction operation based on parameter data and a prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0013] A third aspect of the present application provides a computer-readable storage medium, the storage medium including a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the method for predicting the reserves of shale gas wells according to the first aspect.

[0014] A fourth aspect of the present application provides an electronic device, which includes: a memory for storing programs; a processor, coupled to the memory, for running the programs to execute the method for predicting the reserves of shale gas wells according to the first aspect.

[0015] Compared with the prior art, the method and device for predicting the production and reserves of shale gas wells provided by the present application first obtain the parameter data of the shale gas well, the parameter data including geological parameter data, engineering parameter data and production parameter data; then, perform the prediction operation based on the parameter data and the prediction model to obtain the prediction result, wherein the prediction model is obtained by training based on the training sample data, and the training sample data is obtained based on the parameter type of the parameter data, thereby realizing the function of predicting the production and reserves of the shale gas well. Compared with the prior art, the present application can automatically predict the production and reserves based on the parameter data of the shale gas well, using the prediction model obtained by pre-training with the training sample data. Since the parameter type of the acquired parameter data is the same as that of the training sample data, it is ensured that the production and reserves of the shale gas well to be predicted can be correctly predicted. Compared with the existing manual method for prediction and analysis, there is no need for human calculation and analysis, thereby reducing the dependence on human labor, solving the problem of excessive human consumption when predicting production and reserves in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easy to understand. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0017] Figure 1 A flow chart of a method for predicting the production and reserves of a shale gas well provided in an embodiment of the present application is shown;

[0018] Figure 2 A flow chart of another method for predicting the production and reserves of shale gas wells provided in an embodiment of the present application is shown;

[0019] Figure 3A block diagram showing a composition of a device for predicting the production and reserves of a shale gas well provided in an embodiment of the present application is shown;

[0020] Figure 4 A block diagram showing another device for predicting the production and reserves of shale gas wells provided in an embodiment of the present application is shown;

[0021] Figure 5 A schematic diagram of an electronic device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0022] 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 described herein. On the contrary, these embodiments are provided in order 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.

[0023] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by technicians in the field to which this application belongs.

[0024] like Figure 1 As described above, an embodiment of the present application provides a method for predicting the reserves of a shale gas well, the method comprising the following steps:

[0025] 101, obtaining parameter data of shale gas wells.

[0026] Among them, parameter data includes geological parameter data, engineering parameter data and production parameter data.

[0027] In the embodiment of the present invention, in order to realize the prediction function of the production and reserves of shale gas wells, in fact, a variety of different factors will affect its actual production and reserves. Among them, a shale gas well is a well drilled from the ground to the gas layer for the purpose of extracting shale gas during geological exploration. Among them, in the prediction process, its influencing factors mainly include geological parameter data, engineering parameter data and production parameter data.

[0028] Specifically, geological parameter data is parameter data that can reflect the generation and enrichment of shale gas resources, and may include: basin, main shale formation, burial depth, favorable area, geochemical parameters of organic-rich shale, physical parameters, gas content, brittleness parameters, etc. Basin can reflect the geographical location of the shale gas well to be predicted; main shale formations can include parameter data such as blocks and layers of shale gas wells; geochemical parameters of organic-rich shale can include parameter data such as thickness, total organic carbon (TOC), organic matter maturity, organic matter type, etc.; physical parameters can include parameter data such as porosity and permeability; brittleness parameters can include parameter data such as rock mineral composition, Poisson's ratio, and Young's modulus.

[0029] In addition, engineering parameter data may include: number of fracturing clusters, number of fracturing stages, fracture length, fracture spacing, fracture width, fracture height, fracture conductivity, horizontal well spacing and horizontal well length, etc., which have a certain degree of impact on the production capacity of shale gas wells.

[0030] In addition, production parameter data mainly include test output, daily output in the first year and production start date.

[0031] At the same time, in the process of obtaining the above-mentioned parameter data, it can be carried out by reading the uploaded data, which can be a document containing parameter data of shale gas wells. In addition, the document can be any type of text document, spreadsheet, etc. Of course, the parameter data can also be obtained in real time from the database of the shale gas exploration system through a preset data interface.

[0032] 102, performing a prediction operation based on the parameter data and the prediction model to obtain a prediction result.

[0033] The prediction model is obtained by training based on the training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0034] In this embodiment, the prediction model can be understood as a pre-trained model for predicting the production and reserves of shale gas wells, that is, predicting the EUR of shale gas wells. After the parameter data is obtained in the above steps, the parameter data can be directly used as input data and input into the prediction model, and the prediction model will perform automatic EUR prediction based on the received parameter data, that is, obtain the prediction result.

[0035] Furthermore, the prediction model of this embodiment can be obtained by training the training sample data through a preset prediction algorithm, for example, the prediction algorithm can be a random forest algorithm, a linear regression algorithm, a decision tree algorithm, etc.

[0036] At the same time, in this step, the training sample data can be understood as historical data with the same parameter type as the currently acquired parameter data, and the data form and type contained in the historical data are consistent with the form and type of the parameter data. In this way, it can be ensured that the prediction model can correctly identify the parameter data after receiving it and determine the accuracy of the prediction result.

[0037] For example, in the process of training the prediction model, if three types of parameter data, parameter A, parameter B and parameter C, are selected as training sample data, then the data type of the parameter data obtained in step 101 should also include three types of data, parameter A, parameter B and parameter C, as input data for prediction.

[0038] The embodiment of the present application provides a method for predicting the production and reserves of a shale gas well. First, the parameter data of the shale gas well is obtained, and the parameter data includes geological parameter data, engineering parameter data, and production parameter data; then, a prediction operation is performed based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is obtained by training based on the training sample data, and the training sample data is obtained based on the parameter type of the parameter data, thereby realizing the function of predicting the production and reserves of the shale gas well. Compared with the prior art, the present application can automatically predict the production and reserves based on the parameter data of the shale gas well using the prediction model obtained by pre-training with the training sample data. Since the parameter type of the acquired parameter data is the same as that of the training sample data, it is ensured that the production and reserves of the shale gas well to be predicted can be correctly predicted. Compared with the existing manual method for prediction and analysis, no human calculation and analysis is required, thereby reducing the dependence on manual labor, solving the problem of excessive labor consumption when predicting production and reserves in the prior art.

[0039] To explain in more detail below, another embodiment of the present application provides a method for predicting the reserves of shale gas wells, specifically: Figure 2 Said method comprises the following steps:

[0040] 201, obtaining parameter data of shale gas wells.

[0041] The parameter data include geological parameter data, engineering parameter data and production parameter data.

[0042] Specifically, the geological parameter data may include at least one of total organic carbon, porosity, 1-layer thickness, gas saturation, pressure coefficient, and brittle mineral content. The engineering parameter data include at least one of vertical depth, horizontal section length, fracturing section length, Class I drilling length, average section distance, fluid intensity, sand addition intensity, and average displacement. The production parameter data include at least one of test production, first-year daily production, and commissioning date.

[0043] Specifically, in actual applications, not all of the above-mentioned multiple parameter data are obtained as parameter data. They can be selected based on the actual needs of the user. The parameter type of the specific selected parameter data is consistent with the parameter type of the sample when the prediction model that performs the prediction function is trained. For example, when the prediction model of a shale gas well is trained based on the historical data of three parameters, namely, total organic carbon, horizontal section length, and first-year daily production, as training sample data, then in this embodiment, when obtaining the parameter data, this step also needs to obtain the specific data corresponding to these three parameters as parameter data.

[0044] In addition, the method described in this embodiment is a method for predicting EUR of shale gas wells based on a prediction model. Therefore, in practical applications, the model can also be trained before prediction, and specifically, the model training can be performed from step 202 to step 203. Of course, if the prediction model has been trained before, step 204 can be directly executed.

[0045] 202. Determine a parameter type according to the parameter data, and acquire training sample data based on the parameter type.

[0046] In the process of training the model, in order to improve the accuracy of the model, historical data is needed for training, wherein the historical data may include, in addition to the first data having the same parameter type as the aforementioned parameter data, the EUR corresponding to the first data, that is, the data with the EUR result may be included in the process of determining the training sample data. Based on this, the training sample data includes the first data corresponding to the parameter type, and the production and reserve data corresponding to the first data.

[0047] Specifically, the parameter type is determined according to the parameter data, that is, the type corresponding to the parameter data is determined. In this embodiment, the parameter type of the training sample data is the same as the parameter type of the parameter data described in the above steps, which also includes total organic carbon, porosity, 1 small layer thickness, gas saturation, pressure coefficient, brittle mineral content, vertical depth, horizontal section length, fracturing section length, type I drilling length, average section distance, fluid intensity, sand addition intensity, average displacement, test production, first year daily production and commissioning date.

[0048] The step of obtaining training sample data based on the parameter type can obtain training sample data in the same way as step 201. Of course, the training sample data can be understood as historical data with EUR recorded. For example, the first data and the EUR corresponding to the first data, i.e., the reserve data, can be obtained by reading a historical document storing the parameter data of the shale gas well; the parameter data can also be read from a historical database storing the parameter data of the shale gas well; or the user can directly input the training sample data through an interactive interface. Of course, the training sample data can also be obtained by other means, and the method described in this embodiment is not limited to this.

[0049] It should be noted that, in this embodiment, the number of training sample data is not limited and can be determined based on user needs and historical data, but it is necessary to ensure that the parameter types contained in the training sample data are consistent with the parameter types of the parameter data obtained for the production and reserves of the shale gas well to be predicted.

[0050] Specifically, when this step is executed, it can be:

[0051] Target data corresponding to the parameter type is collected, and the target data is divided to obtain at least one training sample data and at least one test sample data.

[0052] The target data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data.

[0053] Here, the collected target data may be divided into one training sample data and one test sample data, multiple training sample data and one test sample data, or the target data may be divided into different combinations of multiple training sample data and multiple test sample data. In the division process, the sum of the number of training sample data and the number of test sample data may be equal to the number of target data; or the sum of the number of training sample data and the number of test sample data may be less than the number of target data.

[0054] Furthermore, in this step, target data corresponding to the parameter type is collected, and the target data is divided to obtain at least one training sample data and at least one test sample data, which can be specifically performed as follows:

[0055] Target data corresponding to the parameter type is collected, and the target data is divided into N data blocks, where N is an integer greater than 2.

[0056] The 1st to Nth data blocks are sequentially determined as test sample data, and the remaining N-1 data blocks are used as training sample data to obtain N combined data of training sample data and test sample data.

[0057] For example, when N is 4, in the process of dividing the target data, the target data can be divided into 4 parts, namely, four data blocks A, B, C, and D, wherein the amount of data contained in each data block can be selected based on user needs. Then, data blocks A to D are successively determined as test sample data, and the data of the remaining 3 data blocks are determined as training sample data, and then 4 combinations are obtained, wherein: A is used as test sample data, and the other data blocks B, C, and D are used as training sample data; B is used as test sample data, and the other data blocks A, C, and D are used as training sample data; C is used as test sample data, and the other data blocks A, B, and D are used as training sample data; D is used as test sample data, and the other data blocks A, B, and C are used as training sample data.

[0058] 203. Perform training based on the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain the prediction model.

[0059] Among them, the lightweight gradient boosting machine learning framework algorithm described in this embodiment, namely the LightGBM algorithm, is a tree-based integrated learning method, which uses a gradient boosting algorithm, can combine multiple weak learners into a powerful model, can continuously fit the residual of the model in an iterative manner, and gradually improve the prediction ability of the model, and it uses a tree-based model as a weak learner, wherein each tree consists of multiple nodes, each node contains a feature and a threshold, and by comparing the feature value of the sample with the size of the threshold, the sample is assigned to the left subtree or the right subtree, thereby realizing the classification or regression of the sample, and in the process of classification and regression, a feature splitting strategy called leaf node splitting is adopted, which is different from the traditional hierarchical splitting, and it can select the features and thresholds that contribute the most to the current sample set at each split, which can reduce the depth of the tree and improve the training speed. In addition, the algorithm also supports multi-threading and parallel computing, which can make full use of the computing power of multi-core CPUs and accelerate the training process of the model.

[0060] Based on the above characteristics of the LightGBM algorithm, prediction model training can ensure that the obtained prediction model can efficiently and accurately predict the corresponding results, and occupy less memory, so that model training can be achieved in a shorter time; at the same time, due to the small memory it occupies, this method can be implemented on more devices. For example, this method can be directly applied on exploration equipment that collects data to predict reserves while collecting data.

[0061] Specifically, when this step is executed, it may include: training by combining each of the training sample data with a lightweight gradient boosting machine learning framework algorithm to obtain an initial model corresponding to each of the training samples, and using at least one of the test sample data to determine the prediction model from multiple initial models.

[0062] As an example, the training samples may include N training samples, Tr1, Tr2, Tr3…TrN; the test samples may include M test samples, Te1, Te2, Te3…TeM. When executing this step: train N initial models for the N training samples using the LightGBM algorithm, namely initial model 1, initial model 2, initial model 3…initial model n. Test the initial model 1 using the test sample Te1, test the initial model 2 using the test sample Te2, and test the initial model n using the test sample TeN; alternatively, the initial model 1, initial model 2, initial model 3…initial model n may also be tested using the test sample Te1. This application does not make any specific limitations on this.

[0063] After obtaining multiple initial models, it is necessary to determine the prediction model from the multiple initial models. Specifically, the test result can be determined based on the situation of hitting the actual EUR. Of course, the test results obtained from multiple test sample data can also be used to evaluate the initial model in the form of regression evaluation indicators, where the regression evaluation indicators can include but are not limited to the sum of squared errors, the square root of the sum of squared errors, the fitting coefficient, the total sum of squares, and the residual sum of squares, etc., which are not limited here.

[0064] Furthermore, in practical applications, the process of determining the training sample data and the test sample data from the target data can be performed by dividing the data into N parts as in the aforementioned steps. Based on this, the process of determining the prediction model based on the initial model in this step can be:

[0065] The initial model corresponding to each combined data is obtained by training the training sample data in each combined data in combination with a lightweight gradient boosting machine learning framework algorithm.

[0066] A test operation is performed on the initial model corresponding to each combination of data according to the test sample data in each combination of data to obtain a test result.

[0067] Among the multiple test results, an optimal test result is determined, and an initial model corresponding to the optimal test result is determined as the prediction model.

[0068] In combination with the above example, the method of determining the prediction model based on the initial model in this step can be carried out as follows:

[0069] After obtaining 4 combinations, they are: A is used as the test sample data, and the B, C, and D data blocks are used as the training sample data; B is used as the test sample data, and the A, C, and D data blocks are used as the training sample data; C is used as the test sample data, and the A, B, and D data blocks are used as the training sample data; D is used as the test sample data, and the A, B, and C data blocks are used as the training sample data.

[0070] Here, when data block A is used as test sample data, and data blocks B, C, and D are used as training sample data, data blocks B, C, and D are trained in combination with the lightweight gradient boosting machine learning framework algorithm to obtain model 1, and then data block A is used as test sample data to test model 1 to obtain test result a of model 1;

[0071] When data block B is used as test sample data, and data blocks A, C, and D are used as training sample data, data blocks A, C, and D are trained in combination with the lightweight gradient boosting machine learning framework algorithm to obtain model 2, and then data block B is used as test sample data to test model 2 to obtain test result b of model 1;

[0072] When data block C is used as test sample data, data blocks A, B, and D are used as training sample data. In this way, data blocks A, B, and D are trained in combination with the lightweight gradient boosting machine learning framework algorithm to obtain model 3. Then, data block C is used as test sample data to test model 3 to obtain test result c of model 3.

[0073] When data block D is used as test sample data, and data blocks A, B, and C are used as training sample data, data blocks A, B, and C are trained in combination with the lightweight gradient boosting machine learning framework algorithm to obtain model 4; then data block D is used as test sample data to test model 4 to obtain the test result d of model 4;

[0074] Finally, the best test result is determined based on test results a to test results d, and the initial model corresponding to the test result is determined as the test model. Of course, in practical applications, test results a to test results d can be the hit rate of the reserves in the actual test sample data, or can be the score after the result obtained from each test is operated by the regression evaluation index as described in the above step. Here, the specific type of the test result is not limited, and can be selected based on the actual needs of the user.

[0075] 204, performing a prediction operation based on the parameter data and the prediction model to obtain a prediction result.

[0076] The prediction model is obtained by training based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0077] The process of obtaining the prediction result based on the prediction model in this step may be consistent with the execution method of step 102 in the aforementioned embodiment, and will not be described in detail here.

[0078] 205. Generate a shale gas reserve map based on the parameter data and the prediction result.

[0079] The shale gas reserves map includes shale gas entities and hierarchical relationships between shale gas entities, and the shale gas entities include shale gas wells and shale gas blocks.

[0080] Among them, the shale gas reserves map can be understood as a knowledge map containing shale gas related data, in which each entity involved can be specifically a shale gas well and a shale gas block, and the arrows between the entities represent the hierarchical relationship between the entities. For example, when a shale gas well and a shale gas block are connected and the arrow points to the shale gas well, it means that the shale gas block contains the shale gas well. Of course, in practical applications, shale gas exploration also includes different levels such as basins and shale gas reservoirs, and the relationship between them is that basins include shale gas blocks, and shale gas blocks include shale gas reservoirs and shale gas wells. Therefore, in this map, entities can also involve shale gas reservoirs, basins, and other data representing different levels between shale gas entities. In this case, the connection between entities can also represent the relationship between shale gas reservoirs and basins, and the relationship between shale gas reservoirs and shale gas blocks.

[0081] Furthermore, in practical applications, since the user may also need to additionally display the above parameter data and prediction results in the determined map, in this case, the shale gas reserves map includes the initial map and shale gas entity attributes. When this step is executed, it may include:

[0082] constructing an initial map based on the relationships among the shale gas entities;

[0083] respectively determining the corresponding relationship between the parameter data and the shale gas entity, and the corresponding relationship between the prediction result and the shale gas entity;

[0084] The parameter data and the prediction results are determined as the attributes of the shale gas entity and added to the corresponding shale gas entity in the initial map to obtain the shale gas reserves map.

[0085] In an embodiment, the initial map can be understood as a map that only includes various entities of shale gas and the relationships between them. Specifically, based on the above description, it can be seen that the shale gas entity may include at least one basin, at least one shale gas block, at least one shale gas well and at least one shale gas reservoir; the basin may include various shale gas blocks; the shale gas block may include various shale gas wells and / or shale gas reservoirs. The initial map may also include other entities and hierarchical relationships between entities, and this application does not make specific restrictions on this.

[0086] Specifically, since the various parameter data corresponding to the shale gas well can reflect the relevant characteristics of the shale gas well, and the results predicted by the prediction model can be understood as the predicted reserves of the shale gas well, the parameter data and the prediction results can be used as the attributes of the main characteristics of the shale gas well. Therefore, the parameter data and the prediction results can be added to the corresponding entity in the initial map, so as to ensure that when the user clicks on the entity, the attribute information of the shale gas well or shale gas block corresponding to the entity can be displayed, that is, the parameter data and the predicted reserves, that is, EUR. In this way, by adding the attribute information of the shale gas well or shale gas block corresponding to the entity in the initial map to obtain the shale gas reserves map, it can be ensured that the function of quickly and intuitively querying various parameters related to shale gas can be realized directly through the shale gas reserves map.

[0087] In addition, the attribute information in the shale gas reserves map can also include other parameters to display various information about shale gas and predicted reserves. The specific display form can be based on the dialog box that pops up in response to the user's click command. Among them, the dialog box of the shale gas reserves map can intuitively display various shale gas information related to the currently selected entity. Taking the shale gas well as an example, the dialog box may include but is not limited to the shale gas well name, ID, wellhead coordinate X, wellhead coordinate Y, core filling altitude, geographical location, drilling site, well type, well type, production time, brittle mineral content, TOC, porosity, permeability, average daily gas production, sand addition intensity, fluid intensity, fracturing fluid volume, proppant dosage, gas content, etc. Quantity, gas saturation, reserves, number of fracturing clusters, average cluster spacing, company, pressure coefficient, horizontal section length, Poisson's ratio, Young's modulus, brittleness index, test production, wellhead pressure, first year daily production, 1 small layer thickness, vertical depth, type I drilling length, type I drilling rate, well depth, horizontal section length of fracturing, number of transformation sections, length of fracturing section, main fracturing fluid volume, sand addition volume, construction displacement, average section spacing, and block to which it belongs; when the selected entity corresponds to a shale gas block, the dialog box can display the name of the shale gas block, shale Gas block ID, area, recoverable resources, geological resources, reserve abundance, average recoverable coefficient, average gas content, average reservoir depth, minimum reservoir depth, maximum reservoir depth, average horizontal well length, average vitrinite reflectance (Ro), minimum vitrinite reflectance (Ro), maximum vitrinite reflectance (Ro), average TOC, minimum TOC, maximum TOC, average net thickness, minimum net thickness, maximum net thickness, average total thickness, minimum total thickness, maximum total thickness, average single well test production, average Wellhead pressure of single well test, geographical location, formation name, sedimentary environment, main lithology of reservoir, type of organic matter in reservoir, degree of development of reservoir fractures, minimum gas content, maximum gas content, average porosity, maximum porosity, minimum porosity, average reservoir pressure coefficient, maximum reservoir pressure coefficient, minimum thermal evolution degree of reservoir, maximum horizontal section length, average daily gas production of single well, average EUR of single well, recovery factor, geological reserves, reservoir pressure, quartz mineral content, brittleness index, gas saturation, rock density, block type. Of course, among the various parameters described above, users can actually select any one or more of the above as attribute information to add to the map for display based on their needs. There is no limitation here, and actual needs shall prevail.

[0088] Furthermore, as a response to the above Figure 1In order to realize the method shown in the figure, another embodiment of the present application also provides a device for predicting the production and reserves of shale gas wells. The embodiment of the device for predicting the production and reserves of shale gas wells corresponds to the aforementioned method embodiment. For the sake of ease of reading, the embodiment of the device for predicting the production and reserves of shale gas wells will no longer repeat the details of the aforementioned method embodiment one by one, but it should be clear that the device in this embodiment can correspond to all the contents of the aforementioned method embodiment. Specifically, Figure 3 As shown, the device for predicting the reserves of shale gas wells includes:

[0089] An acquisition unit 31 may be used to acquire parameter data of a shale gas well, wherein the parameter data includes geological parameter data, engineering parameter data and production parameter data;

[0090] The execution unit 32 can be used to perform a prediction operation based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0091] Furthermore, as a response to the above Figure 2 In order to realize the method shown in the figure, another embodiment of the present application also provides a device for predicting the production and reserves of shale gas wells. The embodiment of the device for predicting the production and reserves of shale gas wells corresponds to the aforementioned method embodiment. For the sake of ease of reading, the embodiment of the device for predicting the production and reserves of shale gas wells will no longer repeat the details of the aforementioned method embodiment one by one, but it should be clear that the device in this embodiment can correspond to all the contents of the aforementioned method embodiment. Specifically, Figure 4 As shown, the device for predicting the reserves of shale gas wells includes:

[0092] An acquisition unit 41 may be used to acquire parameter data of a shale gas well, wherein the parameter data includes geological parameter data, engineering parameter data, and production parameter data;

[0093] The execution unit 42 can be used to perform a prediction operation based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0094] Furthermore, the device also includes:

[0095] The generation unit 43 can be used to generate a shale gas reserve map based on the parameter data and the prediction results, wherein the shale gas reserve map includes shale gas entities and hierarchical relationships between shale gas entities, and the shale gas entities include shale gas wells and shale gas blocks.

[0096] Furthermore, the shale gas reserves map includes an initial map and shale gas entity attributes, and the generation unit 43 includes:

[0097] A construction module 431 may be used to construct an initial map based on the relationship between the shale gas entities;

[0098] The determination module 432 may be used to respectively determine the corresponding relationship between the parameter data and the shale gas entity, and the corresponding relationship between the prediction result and the shale gas entity;

[0099] The adding module 433 can be used to determine the parameter data and the prediction results as the attributes of the shale gas entity and add them to the corresponding shale gas entity in the initial map to obtain the shale gas reserves map.

[0100] Furthermore, the device also includes:

[0101] The sample acquisition unit 44 may be used to determine the parameter type according to the parameter data, and acquire training sample data based on the parameter type, wherein the training sample data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data;

[0102] The training unit 45 can be used to perform training based on the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain the prediction model.

[0103] Furthermore, the sample acquisition unit 44 includes:

[0104] The partitioning module 441 may be used to collect target data corresponding to the parameter type and partition the target data into N data blocks, where N is an integer greater than 2;

[0105] The determination module 442 may be used to sequentially determine the first to Nth data blocks as test sample data, and use the remaining N-1 data blocks as training sample data to obtain N combined data of training sample data and test sample data;

[0106] The training unit 45 comprises:

[0107] The training module 451 can be used to perform training by combining the training sample data in each of the combined data with a lightweight gradient boosting machine learning framework algorithm to obtain the initial model corresponding to each of the combined data;

[0108] The testing module 452 may be used to perform a testing operation on the initial model corresponding to each combination of data according to the test sample data in each combination of data to obtain a test result;

[0109] The determination module 453 may be used to determine an optimal test result among the multiple test results, and determine the initial model corresponding to the optimal test result as the prediction model.

[0110] Further, the geological parameter data includes at least one of: total organic carbon, porosity, thickness of a small layer, gas saturation, pressure coefficient and brittle mineral content;

[0111] The engineering parameter data includes at least one of: vertical depth, horizontal section length, fracturing section length, type I drilling length, average section distance, fluid intensity, sand addition intensity, and average displacement;

[0112] The production parameter data includes at least one of: test output, first-year daily output, and production start date.

[0113] The embodiment of the present application provides a method and device for predicting the production and reserves of a shale gas well. First, the parameter data of the shale gas well is obtained, and the parameter data includes geological parameter data, engineering parameter data and production parameter data; then, a prediction operation is performed based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is obtained by training based on the training sample data, and the training sample data is obtained based on the parameter type of the parameter data, thereby realizing the function of predicting the production and reserves of the shale gas well. Compared with the prior art, the present application can automatically predict the production and reserves based on the parameter data of the shale gas well, using the prediction model obtained by pre-training with the training sample data. Since the parameter type of the acquired parameter data is the same as that of the training sample data, it is ensured that the production and reserves of the shale gas well to be predicted can be correctly predicted. Compared with the existing manual method for prediction and analysis, no human calculation and analysis is required, thereby reducing the dependence on manual labor, solving the problem of excessive labor consumption when predicting production and reserves in the prior art.

[0114] An embodiment of the present application provides a computer-readable storage medium, which includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the method for predicting the reserves of shale gas wells.

[0115] The storage medium may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0116] An embodiment of the present application also provides a data extraction device, which includes a storage medium; and one or more processors, wherein the storage medium is coupled to the processor, and the processor is configured to execute program instructions stored in the storage medium; when the program instructions are executed, the method for predicting the reserves of shale gas wells is executed.

[0117] Furthermore, an embodiment of the present invention also provides an electronic device, the electronic device includes a processor and a memory, the acquisition unit 31, the execution unit 32, etc. are stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory.

[0118] One or more kernels may be set, and the method for predicting the production and reserves of shale gas wells may be implemented by adjusting kernel parameters to solve the problem of excessive manual labor consumption when predicting the production and reserves of shale gas in the prior art.

[0119] An embodiment of the present invention provides a storage medium on which a program is stored. When the program is executed by a processor, the method for predicting the production and reserves of a shale gas well is implemented.

[0120] An embodiment of the present invention provides a processor, which is used to run a program, wherein the method for predicting the production and reserves of a shale gas well is executed when the program is run.

[0121] The embodiment of the present invention provides a device 50, such as Figure 5 As shown, the device 50 includes at least one processor 501, at least one memory 502 and a bus 503 connected to the processor 501; the processor 501 and the memory 502 communicate with each other through the bus 503; the processor 501 is used to call the program instructions in the memory 502 to execute the above-mentioned method for predicting the production and reserves of shale gas wells. The device in this article can be a server, PC, PAD, mobile phone, etc.

[0122] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialized program having the following method steps: obtaining parameter data of a shale gas well, wherein the parameter data includes geological parameter data, engineering parameter data, and production parameter data; performing a prediction operation based on the parameter data and a prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

[0123] Furthermore, after performing the prediction operation based on the parameter data and the prediction model to obtain the prediction result, the method further includes:

[0124] A shale gas reserve map is generated according to the parameter data and the prediction result, wherein the shale gas reserve map includes shale gas entities and hierarchical relationships between shale gas entities, and the shale gas entities include shale gas wells and shale gas blocks.

[0125] Further, the shale gas reserves map includes an initial map and shale gas entity attributes; generating a shale gas reserves map according to the parameter data and the prediction result includes:

[0126] constructing an initial map based on the relationships among the shale gas entities;

[0127] respectively determining the corresponding relationship between the parameter data and the shale gas entity, and the corresponding relationship between the prediction result and the shale gas entity;

[0128] The parameter data and the prediction results are determined as the attributes of the shale gas entity and added to the corresponding shale gas entity in the initial map to obtain the shale gas reserves map.

[0129] Furthermore, before performing a prediction operation based on the parameter data and the prediction model to obtain a prediction result, the method further includes:

[0130] Determine a parameter type according to the parameter data, and acquire training sample data based on the parameter type, wherein the training sample data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data;

[0131] Based on the training sample data combined with a lightweight gradient boosting machine learning framework algorithm, training is performed to obtain the prediction model.

[0132] Further, determining the parameter type according to the parameter data, and acquiring training sample data based on the parameter type, includes:

[0133] Target data corresponding to the parameter type is collected, and the target data is divided to obtain at least one training sample data and at least one test sample data; wherein the target data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data.

[0134] Furthermore, the training is performed based on the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain the prediction model, including:

[0135] By combining each of the training sample data with a lightweight gradient boosting machine learning framework algorithm, training is performed to obtain an initial model corresponding to each of the training samples, and using at least one of the test sample data, the prediction model is determined from multiple initial models.

[0136] Furthermore, the collecting target data corresponding to the parameter type and dividing the target data to obtain at least one training sample data and at least one test sample data includes:

[0137] Collecting target data corresponding to the parameter type, and dividing the target data into N data blocks, where N is an integer greater than 2;

[0138] Sequentially determine the 1st to Nth data blocks as test sample data, and use the remaining N-1 data blocks as training sample data, to obtain N combined data of training sample data and test sample data;

[0139] The step of training each of the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain an initial model corresponding to each of the training samples, and determining the prediction model from a plurality of the initial models using at least one of the test sample data, comprises:

[0140] The initial model corresponding to each combination of data is obtained by training the training sample data in each combination of data in combination with a lightweight gradient boosting machine learning framework algorithm;

[0141] Perform a test operation on the initial model corresponding to each combination of data according to the test sample data in each combination of data to obtain a test result;

[0142] Among the multiple test results, an optimal test result is determined, and an initial model corresponding to the optimal test result is determined as the prediction model.

[0143] Further, the geological parameter data includes at least one of: total organic carbon, porosity, thickness of a small layer, gas saturation, pressure coefficient and brittle mineral content;

[0144] The engineering parameter data includes at least one of: vertical depth, horizontal section length, fracturing section length, type I drilling length, average section distance, fluid intensity, sand addition intensity, and average displacement;

[0145] The production parameter data includes at least one of: test output, first-year daily output, and production start date.

[0146] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0147] In a typical configuration, the device includes one or more processors (CPU), memory and bus. The device may also include input / output interface, network interface and the like.

[0148] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip. The memory is an example of a computer-readable medium.

[0149] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0150] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0151] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0152] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for predicting the production and reserves of shale gas wells, characterized in that: include: Acquiring parameter data of shale gas wells, wherein the parameter data includes geological parameter data, engineering parameter data and production parameter data; A prediction operation is performed based on the parameter data and the prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

2. The method according to claim 1, characterized in that After performing the prediction operation based on the parameter data and the prediction model to obtain the prediction result, the method further includes: A shale gas reserve map is generated according to the parameter data and the prediction results, wherein the shale gas reserve map includes shale gas entities and hierarchical relationships between shale gas entities, and the shale gas entities include shale gas wells and shale gas blocks.

3. The method according to claim 2, characterized in that The shale gas reserves map includes an initial map and shale gas entity attributes; Generating a shale gas reserves map according to the parameter data and the prediction result includes: constructing an initial map based on the relationships among the shale gas entities; respectively determining the corresponding relationship between the parameter data and the shale gas entity, and the corresponding relationship between the prediction result and the shale gas entity; The parameter data and the prediction results are determined as the attributes of the shale gas entity and added to the corresponding shale gas entity in the initial map to obtain the shale gas reserves map.

4. The method according to claim 3, characterized in that Before performing a prediction operation based on the parameter data and the prediction model to obtain a prediction result, the method further includes: Determine a parameter type according to the parameter data, and acquire training sample data based on the parameter type, wherein the training sample data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data; Based on the training sample data combined with a lightweight gradient boosting machine learning framework algorithm, training is performed to obtain the prediction model.

5. The method according to claim 4, characterized in that The determining the parameter type according to the parameter data, and acquiring the training sample data based on the parameter type, includes: Target data corresponding to the parameter type is collected, and the target data is divided to obtain at least one training sample data and at least one test sample data; wherein the target data includes first data corresponding to the parameter type and production and reserve data corresponding to the first data.

6. The method according to claim 5, characterized in that The training is performed based on the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain the prediction model, including: By combining each of the training sample data with a lightweight gradient boosting machine learning framework algorithm, training is performed to obtain an initial model corresponding to each of the training samples, and using at least one of the test sample data, the prediction model is determined from multiple initial models.

7. The method according to claim 6, characterized in that The collecting target data corresponding to the parameter type and dividing the target data to obtain at least one training sample data and at least one test sample data comprises: Collecting target data corresponding to the parameter type, and dividing the target data into N data blocks, where N is an integer greater than 2; Sequentially determine the 1st to Nth data blocks as test sample data, and use the remaining N-1 data blocks as training sample data, to obtain N combined data of training sample data and test sample data; The step of training each of the training sample data in combination with a lightweight gradient boosting machine learning framework algorithm to obtain an initial model corresponding to each of the training samples, and determining the prediction model from a plurality of the initial models using at least one of the test sample data, comprises: The initial model corresponding to each combination of data is obtained by training the training sample data in each combination of data in combination with a lightweight gradient boosting machine learning framework algorithm; Perform a test operation on the initial model corresponding to each combination of data according to the test sample data in each combination of data to obtain a test result; Among the multiple test results, an optimal test result is determined, and an initial model corresponding to the optimal test result is determined as the prediction model.

8. A device for predicting the production and reserves of shale gas wells, characterized in that: include: An acquisition unit, used to acquire parameter data of a shale gas well, wherein the parameter data includes geological parameter data, engineering parameter data and production parameter data; An execution unit is used to perform a prediction operation based on the parameter data and a prediction model to obtain a prediction result, wherein the prediction model is trained based on training sample data, and the training sample data is obtained based on the parameter type of the parameter data.

9. A computer-readable storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the method for predicting the reserves of a shale gas well as recited in any one of claims 1 to 7.

10. An electronic device, characterized in that: The electronic device comprises: Memory, used to store programs; A processor, coupled to the memory, is used to run the program to execute the method for predicting the reserves of a shale gas well as described in any one of claims 1-7.

Images