Salt-gypsum bed geological card method and device

Abstract

This application provides a method and apparatus for identifying geological layers in salt-gypsum formations. The method includes: acquiring current drilling information of the well to be logged; inputting the acquired drilling information into a salt-gypsum formation geological layer identification model to obtain an identification result output by the model, wherein the identification result is used to characterize whether the current formation of the well to be logged is a salt-gypsum formation; if the identification result indicates that it is not a salt-gypsum formation, then searching for the next formation information corresponding to the identification result in the known formation information of the adjacent wells to be logged; if the next formation information is a salt-gypsum formation, then gradually shortening the time interval for acquiring the next drilling information of the well to be logged. By using a salt-gypsum formation geological layer identification model and drilling information to gradually shorten the time interval for acquiring the next drilling information, accurate and efficient determination of salt-gypsum formation geological layers can be achieved.

Description

Technical Field

[0001] This application relates to the field of petroleum exploration technology, and in particular to a method and apparatus for geological layer identification in salt-gypsum layers. Background Technology

[0002] In deep well drilling (depths exceeding 6000 meters), drilling through salt-gypsum formations presents numerous technical challenges, one of which is geological blockage within the salt-gypsum formation. In some oilfields, the drilling process frequently encounters problems with untimely and inaccurate geological blockages in the salt-gypsum formation, severely restricting drilling speed and safety.

[0003] Currently, two main methods are used to identify geological layers in gypsum-salt formations. Method 1: Predict the top, bottom, and thickness of the gypsum-salt layer based on seismic data of the area to be measured, and perform inversion prediction of the lithology within the gypsum-salt body. This prediction is then combined with well logging data to ultimately identify the geological layer. Method 2: Collect the first sample of rock cuttings from the first location of the thick gypsum-salt rock to be measured, and detect the content of chemical elements contained in the first sample of rock cuttings. If the changes in the content of magnesium and chloride ions in the first sample of rock cuttings compared to the previous sample of rock cuttings meet preset conditions, then calculate the first salt base index corresponding to the thick gypsum-salt rock to be measured based on the content of chemical elements contained in the first sample of rock cuttings. The salt base layer of the thick gypsum-salt rock is then identified based on the first salt base index.

[0004] However, in the first method of obtaining geological layers for salt-gypsum layers, the inversion process requires subjective judgment by on-site experts, which carries the risk of misjudgment and reduces the accuracy of the geological layers. While the second method, which uses changes in magnesium and chloride ion content in sampled rock cuttings and the first salt bottom index to determine the layers of the thick gypsum-salt rock bottom, can improve the accuracy of the geological layers, it employs elemental logging technology. Sampling rock cuttings requires a period of time for quantitative analysis, and there is a time lag between the sampled rock cuttings and the rock cuttings at the bottom of the well, resulting in untimely geological layer determination. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for geological stratification of salt-gypsum layers, so as to improve the accuracy and efficiency of geological stratification of salt-gypsum layers.

[0006] To address the aforementioned technical problems, this application provides the following technical solutions:

[0007] The first aspect of this application provides a method for identifying geological layers in salt-gypsum formations. The method includes: acquiring current drilling information of the well to be logged; inputting the drilling information into a salt-gypsum formation geological layer identification model to obtain an identification result output by the model, wherein the identification result is used to characterize whether the current formation of the well to be logged is a salt-gypsum formation; wherein the salt-gypsum formation geological layer identification model is a back propagation neural network model (BP Network) trained using drilling information and cuttings logging data corresponding to the drilled portion of the well in the area where the well to be logged is located; if the identification result indicates that the formation is not a salt-gypsum formation, then searching for the next formation information corresponding to the identification result in the known formation information of the adjacent wells of the well to be logged; if the next formation information is a salt-gypsum formation, then gradually shortening the time interval for acquiring the next drilling information of the well to be logged.

[0008] Compared to existing technologies, the salt-gypsum layer geological layer identification method provided in the first aspect of this application, because the salt-gypsum layer geological layer identification model is trained using known drilling information and cuttings logging data, can accurately identify whether a salt-gypsum layer geological layer has been reached based on the current drilling information. Furthermore, drilling information such as well number, wellbore parameters, and drilling parameters can be directly obtained during drilling, eliminating the need for excessive waiting time, thus improving the efficiency of salt-gypsum layer geological layer identification in deep wells. Moreover, when it is determined that the current layer is not a salt-gypsum layer, the time interval for obtaining drilling information is gradually shortened when determining the next layer as a salt-gypsum layer based on the stratigraphic information of adjacent wells. This ensures that drilling information is obtained immediately upon entering a salt-gypsum layer, thereby achieving salt-gypsum layer geological layer identification and improving the accuracy of salt-gypsum layer geological layer identification in deep wells. In summary, the salt-gypsum layer geological layer identification method provided in this application can achieve accurate and efficient identification of salt-gypsum layers in deep wells.

[0009] In other embodiments provided in this application, the error backpropagation neural network model has a network structure with three hidden layers. Before inputting drilling information into the salt-gypsum layer geological identification model, the method further includes: acquiring drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well to be logged is located; dividing the drilling information and cuttings logging data corresponding to the drilled portion according to the formation lithology and well body to obtain drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies; deleting outliers in the drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies, and supplementing missing values ​​in the drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies to obtain training data; using the drilling information in the training data as input type data and the cuttings logging data in the training data as output type data to train the error backpropagation neural network model, and determining the model whose performance parameters after training are greater than the preset parameters as the salt-gypsum layer geological identification model.

[0010] The model employs three hidden layers, enabling it to adapt to various training requirements without becoming excessively large, thus improving both accuracy and processing efficiency. Furthermore, during training, the training data is divided according to stratigraphic lithology, enhancing the accuracy and efficiency of outlier detection and missing value imputation. This, in turn, improves the efficiency and accuracy of training data preprocessing, ultimately increasing model precision and improving the accuracy of the salt-gypsum geological strata.

[0011] In other embodiments provided in this application, obtaining drilling information and cuttings logging data corresponding to the drilled portion of the well to be logged and the well in the area to be logged includes: obtaining drilling information and cuttings logging data from the drilling log, geological daily report, drilling engineering parameters and logging data of the well to be logged and the well in the area to be logged; the method further includes: obtaining one or more of the following from the drilling log, geological daily report, drilling engineering parameters and logging data of the well to be logged and the well in the area to be logged, including standpipe pressure, casing pressure, resistivity, acoustic wave, gamma wave, elemental content logging data and sidewall tack data, and using the standpipe pressure, casing pressure, resistivity, acoustic wave, gamma wave, elemental content logging data and sidewall tack data for model training.

[0012] The training data for the model is obtained through drilling logs, daily geological reports, drilling engineering parameters, and logging data, enabling accurate and rapid acquisition of training data. In addition, the training data also includes well pressure, casing pressure, resistivity, sonic logging, gamma logging, elemental content logging data, and sidewall tack data, which can improve the richness of the training data, thereby improving the model accuracy and ultimately improving the accuracy of geological tacks in the salt-gypsum layer.

[0013] In other embodiments provided in this application, there are multiple adjacent wells, and the weight corresponding to the stratigraphic information of each adjacent well is negatively correlated with the distance between the corresponding adjacent well and the well to be logged. The process of searching for the next stratigraphic information corresponding to the identified stratigraphic information in the known stratigraphic information of the adjacent wells of the well to be logged includes: searching for the next candidate stratigraphic information corresponding to the identified stratigraphic information in the known stratigraphic information of each adjacent well of the well to be logged; weighting the candidate stratigraphic information corresponding to each adjacent well with a weight to obtain different candidate stratigraphic information and their corresponding coefficients; and determining the candidate stratigraphic information corresponding to the largest coefficient as the next stratigraphic information.

[0014] By assigning weights to each neighboring well and weighting all the neighboring wells together with the next stratigraphic information, the stratigraphic information with the highest weight is selected as the next stratigraphic information, thereby improving the accuracy of determining the next stratigraphic information of the well to be measured.

[0015] In other embodiments provided in this application, after searching for the next layer information corresponding to the identified layer information in the known layer information of the adjacent well to be logged, the method further includes: if the next layer information is not a salt-gypsum layer geology, then extending the time interval for obtaining the next drilling information of the well to be logged.

[0016] When it is determined that the next layer in the well to be logged may not be a salt-gypsum geological layer, the data acquisition time interval is increased to reduce the number of times the salt-gypsum geological layer identification is required, thereby reducing the resource consumption of the salt-gypsum geological layer identification model.

[0017] In other embodiments provided in this application, drilling information includes wellbore parameters and drilling parameters; wellbore parameters include: well depth and well inclination angle; drilling parameters include: hook load, torque, drilling speed, drilling pressure and rotational speed.

[0018] Well depth and inclination angle can easily and clearly characterize wellbore parameters, while hook load, torque, drilling speed, drilling pressure, and rotational speed can easily and clearly characterize drilling parameters. Therefore, it is possible to obtain model input data more quickly, and the model can more quickly and accurately calculate whether the salt-gypsum geological layer has been reached, thus improving the efficiency and accuracy of drilling into the salt-gypsum geological layer.

[0019] A second aspect of this application provides a salt-gypsum layer geological layer identification device. The device includes: an acquisition module for acquiring current drilling information of a well to be logged; a layer identification module for inputting the acquired drilling information into a salt-gypsum layer geological layer identification model to obtain an identification result output by the salt-gypsum layer geological layer identification model, wherein the identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer geological layer, wherein the salt-gypsum layer geological layer identification model is an error backpropagation neural network model trained using drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well to be logged is located; a search module for searching the next layer information corresponding to the identification result in the known layer information of the adjacent wells of the well to be logged if the identification result characterizes a salt-gypsum layer geological layer; and a processing module for shortening the time interval for acquiring the next drilling information of the well to be logged if the next layer information is a salt-gypsum layer geological layer.

[0020] A third aspect of this application provides a computer device including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the method of the first aspect.

[0021] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method of the first aspect.

[0022] The fifth aspect of this application provides a computer program product, including a computer program that, when executed by a processor, implements the method of the first aspect.

[0023] The salt-gypsum geological stratification device provided in the second aspect of this application, the computer equipment provided in the third aspect, the computer-readable storage medium provided in the fourth aspect, and the computer program product provided in the fifth aspect have the same or similar beneficial effects as the salt-gypsum geological stratification method provided in the first aspect. Attached Figure Description

[0024] The above and other objects, features, and advantages of exemplary embodiments of this application will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of this application are illustrated by way of example and not limitation, with the same or corresponding reference numerals denoteing the same or corresponding parts, wherein:

[0025] Figure 1 This is a flowchart illustrating the geological stratification method for salt-gypsum layers in the embodiments of this application. Figure 1 ;

[0026] Figure 2 This is a flowchart illustrating the geological stratification method for salt-gypsum layers in the embodiments of this application. Figure 2 ;

[0027] Figure 3This is a schematic diagram of model training in an embodiment of this application;

[0028] Figure 4 This is a schematic diagram illustrating the determination of the next layer information in an embodiment of this application;

[0029] Figure 5 This is an example diagram illustrating the determination of the next layer information in an embodiment of this application;

[0030] Figure 6 This is an example diagram showing the completion of the card layer operation in an embodiment of this application;

[0031] Figure 7 This is a schematic diagram of the structure of the salt-gypsum layer geological carding device in the embodiments of this application. Figure 1 ;

[0032] Figure 8 This is a schematic diagram of the structure of the salt-gypsum layer geological carding device in the embodiments of this application. Figure 2 ;

[0033] Figure 9 This is a schematic diagram of the structure of the computer device in the embodiments of this application. Detailed Implementation

[0034] Exemplary embodiments of this application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0035] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains.

[0036] Currently, deep well drilling frequently involves geological strata in salt-gypsum layers. In these strata, prediction is either made through seismic data inversion or through real-time sampling and quantitative analysis of rock cuttings. However, seismic inversion prediction relies on subjective human judgment, which may reduce the accuracy of geological strata in deep wells. Real-time sampling and quantitative analysis of rock cuttings is time-consuming, reducing the timeliness of geological strata analysis in deep wells.

[0037] In view of this, embodiments of this application provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for identifying geological layers in salt-gypsum formations. This method trains an error backpropagation neural network model using drilled data from the well to be logged and other wells in the area to obtain a salt-gypsum formation identification model. Then, using currently obtained drilling information such as well number, wellbore parameters, and drilling parameters, the model determines whether a salt-gypsum formation has been reached. This achieves accurate identification of salt-gypsum formations in deep wells without wasting time, thus improving the efficiency of identifying salt-gypsum formations in deep wells.

[0038] It should be noted that all components, data, and related processing methods involved in this application are authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0039] First, the method for geological stratification of salt-gypsum layers provided in the embodiments of this application will be described in detail.

[0040] Figure 1 This is a flowchart illustrating the geological stratification method for salt-gypsum layers in the embodiments of this application. Figure 1 See Figure 1 As shown, the method may include:

[0041] S11: Obtain the current drilling information of the well to be logged.

[0042] In one possible implementation, the drilling information includes drilling parameters, such as hook load, torque, drilling speed, drilling pressure, and rotational speed, and may also include the geographical location of the well to be logged, and further may include the well depth and well inclination angle.

[0043] In another possible implementation, the drilling information includes well number, well parameters, and drilling parameters; where the well number is used to identify the area to which the well belongs and different wells in the same area. Specifically, the well number is used to characterize the geographical location information of the well to be logged, and the well number corresponding to the same geographical location information is unique.

[0044] Optionally, the well number includes a first well number identifier and a second well number identifier. The first well number identifier is used to identify the area to which the well belongs, and the second well number identifier is used to identify different wells within the same area.

[0045] Specifically, this application provides an example of drilling information including well number, wellbore parameters, and drilling parameters.

[0046] The wells to be tested here are those currently being drilled and require geological testing of the salt-gypsum layer.

[0047] During the drilling process, when it is necessary to test for salt-gypsum geological formations, the well number, wellbore parameters, and drilling parameters can be obtained from the drilling data. This drilling data can refer to the data recorded during the drilling design phase. Alternatively, the well number, wellbore parameters, and drilling parameters can be obtained directly from the field. Field acquisition can refer to acquisition through various measuring devices both above and below ground. For example, the well number can be obtained through a camera, wellbore parameters through wellbore sensors, and drilling parameters through drill pipe sensors. Of course, the well number, wellbore parameters, and drilling parameters can also be obtained by combining drilling data with field data. Combined acquisition can mean acquiring them simultaneously and then selecting the appropriate one, or selecting one of them.

[0048] The well number here refers to the number of the well to be logged, which is determined based on the drilling project.

[0049] The wellbore parameters mentioned here refer to parameters related to the wellbore, and are not specifically limited here.

[0050] The drilling parameters mentioned here refer to parameters related to the drilling process, and are not specifically limited here.

[0051] S12: Input the drilling information into the salt-gypsum layer geological card layer identification model to obtain the identification result output by the salt-gypsum layer geological card layer identification model. The identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer. The salt-gypsum layer geological card layer identification model is an error backpropagation neural network model trained with drilling information and cuttings logging data corresponding to the well to be logged and the drilled part of the well in the area where the well to be logged is located.

[0052] Optionally, during specific training, an initial error backpropagation neural network model is first constructed, and then the error backpropagation neural network model is trained using drilling information corresponding to the well to be logged and the portion of the well already drilled in the area where the well is located.

[0053] Optionally, when formally using the salt-gypsum layer geological layer identification model for salt-gypsum layer geological layer identification, model training is required first to obtain the salt-gypsum layer geological layer identification model. Specifically, during training, an initial error backpropagation neural network model is first constructed, and then the error backpropagation neural network model is trained using the well to be logged and the well number, wellbore parameters, drilling parameters, and cuttings logging data corresponding to the drilled portion of the well in the area to be logged.

[0054] The cuttings logging data here can refer to the analysis of cuttings carried by drilling fluid, the analysis of cuttings sampled from the wellbore, or the information on rock formations obtained through sounding wave analysis.

[0055] During training, each set of well number, wellbore parameters, drilling parameters, and cuttings logging data serves as one training dataset, with multiple training datasets in total. These multiple training datasets are divided into training, testing, and validation sets according to a preset ratio. The specific value of this preset ratio can be determined based on actual needs and is not specifically limited here. The well number, wellbore parameters, and drilling parameters in the training dataset serve as the input data, while the cuttings logging data serves as the output data, ultimately completing the model training.

[0056] The currently acquired well number, wellbore parameters, and drilling parameters are input into the salt-gypsum layer geological layer identification model. The model then predicts whether the current formation is a salt-gypsum layer. If the model predicts yes, it determines that the well has entered a salt-gypsum layer, thus achieving salt-gypsum layer geological layer identification. If the model predicts no, it determines that the well has not yet entered a salt-gypsum layer, and drilling needs to continue, while the well number, wellbore parameters, and drilling parameters are continuously acquired to perform salt-gypsum layer geological layer identification.

[0057] S13: Determine whether the identification result represents a salt-gypsum layer geology. If yes, proceed to S14; otherwise, proceed to S15.

[0058] Since the identification results output by the salt-gypsum layer geological strata identification model are generally yes or no, or 0 or 1, a pre-defined correspondence can be used to determine whether the identification result represents a salt-gypsum layer geological layer. This correspondence includes each specific identification result and its corresponding information on whether it represents a salt-gypsum layer geological layer. For example, the correspondence could be: 1 - salt-gypsum layer geological layer, 0 - non-salt-gypsum layer geological layer (or other specific geological strata). This allows for the determination of whether the identification result represents a salt-gypsum layer geological layer.

[0059] S14: Determine the geological conditions of the salt-gypsum layer currently drilled to the well to be logged.

[0060] S15: Search for the next layer information corresponding to the identified layer information in the known layer information of the adjacent wells of the well to be logged.

[0061] The adjacent wells of the well to be logged can refer to the one or more wells that are closest to the well to be logged, or to the one or more wells that are most similar to the well as a whole. The whole here can include: well design, drill pipe parameters used, and one or more of the mining projects involved.

[0062] For adjacent wells, the drilled sections contain records of formation information at different depths, i.e., known stratigraphic information. Formation information, i.e., stratigraphic information, can also be obtained from the identification results. The process involves searching for the stratigraphic information corresponding to the identification result within the known stratigraphic information. If found, the next stratigraphic level below that level is the next stratigraphic level to be acquired. If not found, it indicates that the formation at this location of the well to be logged is unusual, and the time interval for obtaining the well number, wellbore parameters, and drilling parameters should be shortened.

[0063] S16: Determine if the next stratigraphic level is a salt-gypsum layer. If yes, proceed to S17; otherwise, proceed to S18.

[0064] Specifically, the information of the next stratigraphic level can be matched with the geological features of the salt-gypsum layer. If the match is successful, the next stratigraphic level is determined to be a salt-gypsum layer. A successful match here can mean that the textual similarity between the description of the next stratigraphic level and the geological features of the salt-gypsum layer is greater than a preset similarity. If the match fails, the next stratigraphic level is determined not to be a salt-gypsum layer.

[0065] S17: Gradually shorten the time interval for obtaining drilling information from the well to be logged.

[0066] Optionally, the time interval between obtaining the well number, wellbore parameters, and drilling parameters of the well to be tested can be gradually shortened.

[0067] The next stratum information is salt-gypsum geology, which means that the next stratum drilled by the well to be logged may be salt-gypsum geology. In order to avoid the next salt-gypsum geology block when the well has been drilled for a period of time, and to ensure that the salt-gypsum geology block can be achieved as soon as the well is drilled, the time interval for obtaining the well number, wellbore parameters and drilling parameters of the next well to be logged can be gradually shortened.

[0068] The term "gradually shortening" here refers to the fact that the time interval between each acquisition of well number, wellbore parameters, and drilling parameters is shorter than the time interval between the previous acquisitions. For example, when the next stratigraphic information is first determined to be a salt-gypsum layer, the well number, wellbore parameters, and drilling parameters are acquired again after 5 seconds, and steps S12-S16 are executed. When the next stratigraphic information is determined to be a salt-gypsum layer for the second time, the well number, wellbore parameters, and drilling parameters are acquired again after 4 seconds, and steps S12-S16 are executed again, until the model's identification result is confirmed as a salt-gypsum layer in step S13.

[0069] S18: Continue acquiring drilling information after the preset time interval.

[0070] This section uses drilling information, including well number, wellbore parameters, and drilling parameters, as an example. If the next layer information is not a salt-gypsum geological layer, it means that the well to be logged still has a long way to go to reach the salt-gypsum geological layer. In order to reduce the resource consumption of the model, the time interval for obtaining the well number, wellbore parameters, and drilling parameters for the next well to be logged can be increased or gradually increased. Alternatively, the well number, wellbore parameters, and drilling parameters can be obtained again according to the previous fixed time interval. That is, after the preset time interval, the well number, wellbore parameters, and drilling parameters can continue to be obtained.

[0071] As described above, the salt-gypsum layer geological layer identification method provided in this application can accurately identify whether a salt-gypsum layer has been reached based on the current drilling information, since the salt-gypsum layer geological layer identification model is trained using known drilling information and cuttings logging data. Furthermore, the drilling information can be obtained directly during drilling without excessive waiting time, thus improving the efficiency of salt-gypsum layer geological layer identification in deep wells. Additionally, when it is determined that the current layer is not a salt-gypsum layer, the time interval for obtaining drilling information is gradually shortened when determining the next layer as a salt-gypsum layer based on the stratigraphic information of adjacent wells. This ensures that drilling information is obtained immediately upon entering a salt-gypsum layer, thereby achieving salt-gypsum layer geological layer identification and improving the accuracy of salt-gypsum layer geological layer identification in deep wells. In summary, the salt-gypsum layer geological layer identification method provided in this application can achieve accurate and efficient identification of salt-gypsum layers in deep wells.

[0072] Furthermore, as Figure 1 As a refinement and extension of the method shown, this application embodiment also provides a method for geological stratification of salt-gypsum layers.

[0073] Figure 2 This is a flowchart illustrating the geological stratification method for salt-gypsum layers in the embodiments of this application. Figure 2 See Figure 2 As shown, this method can include two main parts. The first part is the geological identification model of salt-gypsum layers, and the second part is the application of the geological identification model of salt-gypsum layers.

[0074] I. Training of the geological strata identification model for salt-gypsum layers:

[0075] S21: Obtain well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic logging, gamma ray, elemental content logging data, cuttings logging data, and sidewall retrieval data from the drilling log, geological daily report, drilling engineering parameters, and logging data of the well to be logged and the wells in the area where the well is located.

[0076] The term "wells within the area to be logged" refers to all wells within that area. This "area" can refer to a region within a predetermined distance, the area of ​​the project where the well is located, a region with geological features, or an administratively defined region.

[0077] For each well to be logged and the wells in its area, there are corresponding drilling logs, geological reports, drilling engineering parameters, and logging data. The well number, depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotational speed, standpipe pressure, casing pressure, resistivity, acoustic logging, gamma ray logging, elemental content logging data, cuttings logging data, and sidewall retrieval data are obtained from the drilling logs, geological reports, drilling engineering parameters, and logging data.

[0078] If drilling logs, geological reports, drilling engineering parameters, or logging data are not available for a particular well, then the well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, standpipe pressure, casing pressure, resistivity, acoustic logging, gamma ray logging, elemental content logging data, cuttings logging data, and sidewall retrieval data should be obtained from the available drilling logs, geological reports, drilling engineering parameters, and logging data. Alternatively, the well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, standpipe pressure, casing pressure, resistivity, acoustic logging, gamma ray logging, elemental content logging data, cuttings logging data, and sidewall retrieval data can be obtained from other available well data.

[0079] If not all necessary data, including well number, depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotational speed, standpipe pressure, casing pressure, resistivity, sonic logging, gamma ray logging, elemental content logging data, cuttings logging data, and sidewall retrieval data, are obtained from the drilling log, geological report, drilling engineering parameters, and logging data, then ensuring the acquisition of essential data such as well number, depth, drilling speed, inclination angle, hook load, torque, drilling speed, drilling pressure, rotational speed, and cuttings logging data is sufficient. Unavailable, non-essential data can be ignored or obtained from other sources. For essential data that is not readily available, it is necessary to obtain it from other sources or by adding additional measuring equipment.

[0080] The next step is to preprocess the acquired data, namely, to divide, clean, and fill it.

[0081] S22: The acquired well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, elemental content logging data, cuttings logging data, and sidewall retrieval data are divided according to formation lithology and well body to obtain well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, elemental content logging data, cuttings logging data, and sidewall retrieval data corresponding to different formation lithologies and different well bodies.

[0082] The lithology of the strata here can refer to the properties or names of the rocks that the strata mainly contain, such as mudstone, sandstone, etc.

[0083] The "well body" here can refer to an independent well.

[0084] Based on the lithology of the formation and the well body, the acquired data, including well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic logging, gamma ray logging, elemental content logging data, cuttings logging data, and sidewall retrieval data, are further subdivided to facilitate the removal of outliers and the supplementation of missing values ​​in subsequent data processing, thereby improving data preprocessing efficiency.

[0085] Among the logging data (well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotational speed, standpipe pressure, casing pressure, resistivity, acoustic logging, gamma ray, elemental content, cuttings logging data, and sidewall retrieval data), these data can be further categorized into input and output types based on their purpose in model training. Input type data includes: logging data for well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotational speed, standpipe pressure, casing pressure, resistivity, acoustic logging, gamma ray, and elemental content. Output type data includes: cuttings logging data and sidewall retrieval data. Lithology can be determined using cuttings logging data and sidewall retrieval data. That is, lithology can be used as output type data during training.

[0086] Table 1 below shows a schematic diagram of the data partitioning.

[0087] Table 1

[0088]

[0089] Table 2 below shows a schematic diagram of the data partitioning.

[0090] Table 2

[0091]

[0092] Where n is any positive integer.

[0093] S23: Remove outliers from logging data, cuttings logging data, and sidewall retrieval data corresponding to different formation lithologies and different well bodies, including well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, elemental content, and related data. Also, supplement missing values ​​from logging data, cuttings logging data, and sidewall retrieval data corresponding to different formation lithologies and different well bodies to obtain training data.

[0094] In outlier removal, for each group of data after partitioning, abnormally high and abnormally low values ​​can be deleted. Abnormally high values ​​can refer to values ​​exceeding the normal range or the highest value among all values. Abnormally low values ​​can refer to values ​​falling below the normal range or the lowest value among all values.

[0095] After deleting outliers, continue with missing value imputation. Missing values ​​here can refer to values ​​that were never retrieved initially, or values ​​that were subsequently deleted as outliers.

[0096] When filling in missing values, you can do so based on the adjacent values ​​of the missing value. For example, you can use the average of the adjacent values ​​to fill in missing values, or you can fill in missing values ​​based on the pattern of multiple adjacent values.

[0097] For example, suppose in a certain set of data, the drilling pressure data for the 4010m interval in the 4000m-4020m section is missing, and it needs to be supplemented based on the average of the adjacent data. The specific formula is as follows:

[0098] Formula (1)

[0099] Among them, WOB 缺失 This indicates the drilling pressure at the missing location, in MPa (WOB). 上 This indicates the drilling pressure (WP) of the well section above the missing location, in MPa (WOB). 下 This indicates the drilling pressure for the section of the well below the missing location, expressed in MPa.

[0100] S24: The well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic logging, gamma ray and element content logging data in the training data are used as input data, and the cuttings logging data and sidewall retrieval data in the training data are used as output data. The error backpropagation neural network model is trained, and the model whose performance parameters after training are greater than the preset parameters is identified as the salt-gypsum layer geological identification model.

[0101] Here, the error backpropagation neural network model employs a three-hidden-layer network structure. This structure can meet the learning needs of data with various dimensions.

[0102] In practical applications, the training data can be divided into training data, test data, and validation data in a ratio of 7:2:1 to train the error backpropagation neural network model, so that the trained model has good lithology identification function.

[0103] Figure 3 This is a schematic diagram of model training in an embodiment of this application. See also... Figure 3As shown, the model employs a three-hidden-layer structure. During training, input type data is input into the input layer and output through the output layer. Output type data resides in the output layer. During model training, the lithological data output from the output layer is compared with the output type data, and relevant parameters in the model are adjusted based on the comparison results until convergence, at which point training ends, yielding the trained salt-gypsum layer geological strata identification model.

[0104] The salt-gypsum layer geological card recognition model can only be used in practical applications after its accuracy reaches the preset accuracy. If the model does not reach the preset accuracy, it must be retrained using the previous training data or newly acquired training data until it reaches the preset accuracy before being put into practical use.

[0105] II. Application of the Geological Strata Identification Model for Salt-Gypsum Layers

[0106] S25: Obtain logging data for the current well number, well depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray and element content of the well to be logged.

[0107] When some data in the logging data, such as well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic logging, gamma ray, and element content, cannot be obtained, these data can be discontinued, and the remaining available data can be used to determine the geological layer of the salt-gypsum layer.

[0108] S26: Input the logging data of well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic wave, gamma and element content into the salt-gypsum layer geological card layer identification model to obtain the identification results output by the salt-gypsum layer geological card layer identification model.

[0109] S27: Determine whether the identification result represents a salt-gypsum layer geology. If yes, proceed to S28; otherwise, proceed to S29.

[0110] S28: Determine the geological conditions of the salt-gypsum layer currently drilled to the well to be logged.

[0111] S29: Find the next candidate stratigraphic information in the known stratigraphic information of each adjacent well to be logged, which corresponds to the stratigraphic information of the identification result.

[0112] For a well to be logged, there may be multiple neighboring wells. For example, there may be multiple wells within the preset range of the well to be logged. Or, multiple wells may be equidistant from the well to be logged, and all of them may be the closest to it.

[0113] For each adjacent well, it is necessary to search for the corresponding stratigraphic information in its known stratigraphic information, and then use the next stratigraphic information of the found stratigraphic information as the candidate stratigraphic information. That is, the number of candidate stratigraphic information can be found as many as there are adjacent wells, or less than that number (it is possible that the stratigraphic information corresponding to the identification result cannot be found in the known stratigraphic information of some adjacent wells).

[0114] Figure 4 This is a schematic diagram illustrating the determination of the next layer information in an embodiment of this application. See [link / reference] Figure 4 As shown, the well to be logged is a drilling well, and the adjacent wells include adjacent well A and adjacent well B. During drilling, the stratigraphic information for depths 1, 2, 3, and 4-1 has been determined, namely stratigraphic position 1, stratigraphic position 2, stratigraphic position 3, and stratigraphic position 4-1, respectively. When drilling reaches depth 4-2, geological stratigraphic analysis of the salt-gypsum layer continues. The known stratigraphic information for adjacent well A includes: depth A1 - stratigraphic position A1, depth A2 - stratigraphic position A2, depth A3 - stratigraphic position A3, and depth A4 - stratigraphic position A4. When drilling reaches depth 4-2, the stratigraphic information is calculated based on the model. If this stratigraphic information is still stratigraphic position A3, then the next stratigraphic information determined from adjacent well A is stratigraphic position A4. Similarly, the known stratigraphic information for adjacent well B includes: depth B1 - stratigraphic position B1, depth B2 - stratigraphic position B2, depth B3 - stratigraphic position B3, and depth B4 - stratigraphic position B4. When drilling reaches a depth of 4-2, the stratigraphic information is calculated based on the model. If the stratigraphic information is still stratigraphic B3, then the next stratigraphic information determined from the adjacent well B is stratigraphic B4.

[0115] Figure 5 This is an example diagram illustrating the determination of the next layer information in an embodiment of this application. See [link / reference]. Figure 5As shown, the well to be logged is the X well in operation, and the adjacent wells include well X-1 and well X-2. In the X well, the stratigraphic information at 3551m, 3702m, 4372m, and 4441m has been determined, which are: upper mudstone section, rock salt section, middle mudstone section, and middle mudstone section, respectively. When drilling to the next depth, geological stratigraphy of the salt-gypsum layer will continue. The known stratigraphic information of the adjacent well X-1 includes: 3531m - upper mudstone section, 3702m - rock salt section, 4569m - middle mudstone section, and 4923m - gypsum-rock salt section. When the X well reaches the next depth, the stratigraphic information will be calculated based on the model. If this stratigraphic information is still the middle mudstone section, then the next stratigraphic information determined from the adjacent well X-1 will be the gypsum-rock salt section. Similarly, the known stratigraphic information for the adjacent well X-2 includes: 4384m - upper mudstone section, 4542m - rock salt section, 4919m - middle mudstone section, and 5298m - gypsum-rock salt section. When the ongoing drilling well X reaches the next depth, the stratigraphic information is calculated based on the model. If this stratigraphic information is still the middle mudstone section, then the next stratigraphic information determined from the adjacent well X-2 will be the gypsum-rock salt section.

[0116] Determining the next stratigraphic level requires comprehensively considering candidate stratigraphic levels found from the known stratigraphic levels of multiple adjacent wells. Specifically, this can be done by considering the weights associated with the stratigraphic level information of each well. The weight of the stratigraphic level information of each adjacent well is negatively correlated with the distance between that adjacent well and the well to be measured. In other words, the closer the adjacent well is to the well to be measured, the greater the weight of its stratigraphic level information.

[0117] S210: Weight the candidate stratigraphic information corresponding to each adjacent well with the weight to obtain different candidate stratigraphic information and their corresponding coefficients.

[0118] Specifically, the weights of adjacent wells with the same candidate stratigraphic information can be added together to obtain different candidate stratigraphic information and their corresponding coefficients.

[0119] For example, suppose that neighboring well 1 (weight 0.5) has a candidate stratigraphic position of medium mudstone, neighboring well 2 (weight 0.3) has a candidate stratigraphic position of gypsum-salt, and neighboring well 3 (weight 0.3) has a candidate stratigraphic position of gypsum-salt. Since neighboring wells 2 and 3 have the same candidate stratigraphic position, the weight 0.3 of neighboring well 2 and the weight 0.3 of neighboring well 3 are added together to obtain the gypsum-salt stratigraphic position and its corresponding coefficient 0.6, and the medium mudstone stratigraphic position and its corresponding coefficient 0.5.

[0120] S211: Determine the candidate layer information corresponding to the maximum coefficient as the next layer information.

[0121] Continuing with the example above, the coefficient 0.6 is the largest at this point, so the gypsum-salt section corresponding to the coefficient 0.6 is taken as the next stratigraphic information for the well to be logged.

[0122] S212: Determine if the next stratigraphic level is a salt-gypsum layer. If yes, proceed to S213; otherwise, proceed to S214.

[0123] S213: Gradually shorten the time interval for acquiring logging data of well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic wave, gamma and element content in the well to be logged.

[0124] When drilling is progressing steadily, the time interval between two logging data acquisitions of well number, depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, and elemental content can be considered as the difference between two depths. Each time the next stratigraphic unit is determined to be a salt-gypsum layer, the depth of the next data acquisition will be higher than the previously determined location.

[0125] Figure 6 This is an example diagram showing the completion of the card layer operation in an embodiment of this application. See [link / reference] Figure 6 As shown, assuming the stratigraphic information at 3551m, 3702m, 4372m, 4441m, 4510m, and 4580m in the current well to be logged is already determined, they are: upper mudstone section, rock salt section, middle mudstone section, middle mudstone section, middle mudstone section, and middle mudstone section, respectively. The known stratigraphic information of the adjacent well on the left includes: 3531m - upper mudstone section, 3702m - rock salt section, 4569m - middle mudstone section, and 4923m - gypsum-rock salt section. When drilling to 4650m, geological stratigraphic analysis of the gypsum-rock salt layer continues, i.e., stratigraphic information is calculated based on the model. This stratigraphic information is a middle mudstone section. At this point, the next stratigraphic information is determined from the adjacent well on the left as a gypsum-rock salt section, and the depth interval for geological stratigraphic analysis of the gypsum-rock salt layer continues to shorten. That is, when drilling to 4710m, stratigraphic information is calculated based on the model. The current stratigraphic information indicates a medium mudstone section. At this point, the next stratigraphic section, identified as a gypsum-salt section, is determined from the adjacent well on the left. This allows for further shortening of the depth interval for geological analysis of the gypsum-salt layer. Specifically, at a depth of 4760m, the stratigraphic information is calculated based on the model. This stratigraphic section is identified as a gypsum-salt section.

[0126] Meanwhile, the known stratigraphic information of the adjacent well on the right includes: 4384m - upper mudstone section, 4542m - rock salt section, 4919m - middle mudstone section, and 5298m - gypsum-rock salt section. When drilling to 4650m, geological stratigraphic analysis of the gypsum-salt layer continued, i.e., stratigraphic information was calculated based on the model. This stratigraphic information is a middle mudstone section. At this point, the next stratigraphic information from the adjacent well on the right was determined to be a gypsum-rock salt section, and the depth interval for geological stratigraphic analysis of the gypsum-salt layer was further shortened. That is, when drilling to 4710m, stratigraphic information was calculated based on the model. This stratigraphic information is a middle mudstone section. At this point, the next stratigraphic information from the adjacent well on the right was determined to be a gypsum-rock salt section, and the depth interval for geological stratigraphic analysis of the gypsum-salt layer was further shortened. That is, when drilling to 4760m, stratigraphic information was calculated based on the model. This stratigraphic information is a gypsum-rock salt section.

[0127] When drilling reached 4760m using two adjacent wells, the model confirmed that the current stratigraphic information was a gypsum-salt section, indicating that drilling had reached the gypsum-salt layer and the geological stratification of the gypsum-salt layer had ended.

[0128] S214: Extend the time interval for acquiring logging data of well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, sonic wave, gamma and element content of the well to be logged.

[0129] If the next identified stratum is not a salt-gypsum layer, it indicates that other strata remain between the current point and the salt-gypsum layer, suggesting a considerable drilling distance is still required. In this case, the time interval for acquiring the next logging data (well number, depth, inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, standpipe pressure, casing pressure, resistivity, acoustic waves, gamma rays, and elemental content) can be extended. This means acquiring these logging data only after reaching a deeper depth, and then performing layer identification. This approach ensures accurate layer identification while reducing the number of model calculations, thereby lowering the model's resource consumption. Once the next layer of information is determined to be a salt-gypsum layer, logging data such as well number, well depth, well inclination angle, hook load, torque, drilling speed, drilling pressure, rotation speed, stand pressure, casing pressure, resistivity, acoustic wave, gamma ray and element content are obtained at shorter intervals to perform layer identification until the layer identification is completed.

[0130] This concludes the description of the salt-gypsum layer geological stratification method provided in the embodiments of this application.

[0131] Based on the same inventive concept, this application also provides a salt-gypsum layer geological carding device.

[0132] Figure 7 This is a schematic diagram of the structure of the salt-gypsum layer geological carding device in the embodiments of this application. Figure 1 See Figure 7As shown, the device may include:

[0133] Module 71 is used to acquire the current drilling information of the well to be logged;

[0134] The layer identification module 72 is used to input drilling information into the salt-gypsum layer geological layer identification model and obtain the identification result output by the salt-gypsum layer geological layer identification model. The identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer. The salt-gypsum layer geological layer identification model is an error backpropagation neural network model trained using drilling information and cuttings logging data corresponding to the well to be logged and the drilled part of the well in the area where the well to be logged is located.

[0135] The search module 73 is used to search for the next stratigraphic information corresponding to the stratigraphic information of the identification result in the known stratigraphic information of the adjacent wells of the well to be logged if the identification result does not represent the geology of the salt-gypsum layer.

[0136] The processing module 74 is used to shorten the time interval for obtaining the next drilling information of the well to be logged if the next layer information is a salt-gypsum layer.

[0137] Furthermore, as Figure 7 In a refinement and extension of the device shown, this application embodiment also provides a salt-gypsum geological carding device.

[0138] Figure 8 This is a schematic diagram of the structure of the salt-gypsum layer geological carding device in the embodiments of this application. Figure 2 See Figure 8 As shown, the device may include:

[0139] The training module 81 includes: a data acquisition unit 811, a data partitioning unit 812, a data deletion and supplementation unit 813, and a training unit 814.

[0140] The acquisition unit 811 is used to acquire drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well is located.

[0141] The acquisition unit 811 is specifically used to acquire drilling information and cuttings logging data from the drilling log, geological daily report, drilling engineering parameters and logging data of the well to be logged and the wells in the area where the well is located.

[0142] The acquisition unit 811 is also used to acquire one or more of the following data from the drilling log, geological daily report, drilling engineering parameters and logging data of the well to be logged and the well in the area where the well is located: stand pressure, casing pressure, resistivity, sonic logging, gamma logging, elemental content logging data and sidewall tack data. The stand pressure, casing pressure, resistivity, sonic logging, gamma logging, elemental content logging data and sidewall tack data are used for model training.

[0143] Division unit 812 is used to divide the drilling information and cuttings logging data corresponding to the drilled part according to the formation lithology and well body, so as to obtain the well number, well body parameters, drilling parameters and cuttings logging data corresponding to different formation lithologies and different well bodies.

[0144] The deletion and supplementation unit 813 is used to delete outliers in drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies, and to supplement missing values ​​in drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies to obtain training data.

[0145] Training unit 814 is used to train the error backpropagation neural network model by taking drilling information in the training data as input data and cuttings logging data in the training data as output data, and to identify the model whose performance parameters after training are greater than the preset parameters as the salt-gypsum layer geological card layer identification model.

[0146] The acquisition module 82 is used to acquire the current drilling information of the well to be logged.

[0147] In practical applications, drilling information includes wellbore parameters and drilling parameters; wellbore parameters include: well depth and well inclination angle; drilling parameters include: hook load, torque, drilling speed, drilling pressure and rotational speed.

[0148] The layer identification module 83 is used to input the acquired drilling information into the salt-gypsum layer geological layer identification model to obtain the identification result output by the salt-gypsum layer geological layer identification model. The identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer. The salt-gypsum layer geological layer identification model is an error backpropagation neural network model trained using drilling information and cuttings logging data corresponding to the well to be logged and the drilled part of the well in the area where the well to be logged is located.

[0149] When there are multiple adjacent wells, and the weight of the stratigraphic information of each adjacent well is negatively correlated with the distance between the adjacent well and the well to be logged, the search module 84 is used to search for the next candidate stratigraphic information corresponding to the stratigraphic information of the identification result in the known stratigraphic information of each adjacent well of the well to be logged if the identification result does not represent a salt-gypsum geology; the candidate stratigraphic information corresponding to each adjacent well is weighted with the weight to obtain different candidate stratigraphic information and their corresponding coefficients; the candidate stratigraphic information corresponding to the largest coefficient is determined as the next stratigraphic information.

[0150] The processing module 85 is used to shorten the time interval for obtaining the next drilling information of the well to be logged if the next layer information is a salt-gypsum layer.

[0151] The processing module 85 is also used to extend the time interval for obtaining the next drilling information of the well to be logged if the next layer information is not salt-gypsum geology.

[0152] It should be noted that the description of the above device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects. For technical details not disclosed in the device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0153] Based on the same inventive concept, this application also provides a computer device.

[0154] Figure 9 This is a schematic diagram of the structure of the computer device in an embodiment of this application. See also... Figure 9 As shown, the computer device may include: a memory 91, a processor 92, and a computer program stored on the memory 91, wherein the processor 92 executes the computer program to implement the methods described in the foregoing embodiments.

[0155] It should be noted that the description of the above computer device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects. For technical details not disclosed in the computer device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0156] Based on the same inventive concept, embodiments of this application also provide a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the methods described in the foregoing embodiments.

[0157] It should be noted that the description of the above computer-readable storage medium embodiments is similar to the description of the above method embodiments, and has similar beneficial effects. For technical details not disclosed in the computer-readable storage medium embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0158] Based on the same inventive concept, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the methods described in the foregoing embodiments.

[0159] It should be noted that the descriptions of the above computer program product embodiments are similar to those of the above method embodiments, and have similar beneficial effects. For technical details not disclosed in the computer program product embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0160] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for geologically identifying salt-gypsum layers, characterized in that, The method includes: Obtain the current drilling information of the well to be logged; The drilling information is input into the salt-gypsum layer geological card layer identification model to obtain the identification result output by the salt-gypsum layer geological card layer identification model. The identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer. The salt-gypsum layer geological card layer identification model is an error backpropagation neural network model trained using the drilling information and cuttings logging data corresponding to the well to be logged and the drilled part of the well in the area where the well to be logged is located. If the identification result does not represent the geology of the salt-gypsum layer, then search for the next layer information corresponding to the identification result in the known layer information of the adjacent wells of the well to be logged; If the next layer information is a salt-gypsum layer geology, then the time interval for obtaining the next drilling information of the well to be logged is gradually shortened; The error backpropagation neural network model has a network structure with three hidden layers. Before inputting the drilling information into the salt-gypsum layer geological card identification model and obtaining the identification result output by the salt-gypsum layer geological card identification model, the method further includes: Obtain drilling information and cuttings logging data corresponding to the well to be logged and the portion of the well already drilled in the area where the well is located; The drilling information and cuttings logging data corresponding to the drilled portion are divided according to the formation lithology and well body to obtain drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies; The outliers in the drilling information and cuttings logging data corresponding to the different formation lithologies and different well bodies are deleted, and the missing values ​​in the drilling information and cuttings logging data corresponding to the different formation lithologies and different well bodies are supplemented to obtain training data. The drilling information in the training data is used as input data, and the cuttings logging data in the training data is used as output data to train the error backpropagation neural network model. The model whose performance parameters after training are greater than the preset parameters is determined as the salt-gypsum layer geological card layer identification model. The number of adjacent wells is multiple, and the weight corresponding to the stratigraphic information of each adjacent well is negatively correlated with the distance between the corresponding adjacent well and the well to be measured; the step of searching for the next stratigraphic information corresponding to the identified stratigraphic information in the known stratigraphic information of the adjacent wells of the well to be measured includes: In the known stratigraphic information of each adjacent well to the well to be logged, find the next candidate stratigraphic information following the stratigraphic information corresponding to the identification result. The candidate stratigraphic information corresponding to each adjacent well is weighted with a weight to obtain different candidate stratigraphic information and their corresponding coefficients. The candidate layer information corresponding to the maximum coefficient is determined as the next layer information.

2. The method according to claim 1, characterized in that, The acquisition of drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well to be logged is located includes: Drilling information and cuttings logging data are obtained from the drilling logs, geological reports, drilling engineering parameters, and logging data of the well to be logged and the wells in the area where the well is located; The method further includes: One or more of the following data are obtained from the drilling logs, geological reports, drilling engineering parameters, and logging data of the well to be logged and wells in the area where the well to be logged is located: stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, elemental content logging data, and sidewall retrieval data. The stand pressure, casing pressure, resistivity, acoustic wave, gamma ray, elemental content logging data, and sidewall retrieval data are used for model training.

3. The method according to claim 1 or 2, characterized in that, After searching for the next stratigraphic information corresponding to the identified stratigraphic information in the known stratigraphic information of the adjacent wells of the well to be logged, the method further includes: If the next layer information is not a salt-gypsum layer geology, then the time interval for obtaining the next drilling information of the well to be logged is extended.

4. The method according to claim 1 or 2, characterized in that, The drilling information includes wellbore parameters and drilling parameters; The wellbore parameters include: well depth and well inclination angle; the drilling parameters include: hook load, torque, drilling speed, drilling pressure and rotational speed.

5. A geological stratification device for salt-gypsum layers, characterized in that, The device includes: The acquisition module is used to acquire the current drilling information of the well to be logged; The layer identification module is used to input the drilling information into the salt-gypsum layer geological layer identification model and obtain the identification result output by the salt-gypsum layer geological layer identification model. The identification result is used to characterize whether the current layer of the well to be logged is a salt-gypsum layer. The salt-gypsum layer geological layer identification model is an error backpropagation neural network model trained using drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well to be logged is located. The error backpropagation neural network model has a network structure with three hidden layers. The search module is used to search for the next stratigraphic information corresponding to the identification result in the known stratigraphic information of the adjacent wells of the well to be logged if the identification result does not represent a salt-gypsum layer geology. The processing module is used to shorten the time interval for obtaining the next drilling information of the well to be logged if the next layer information is a salt-gypsum layer. The device further includes a training module, which includes a data acquisition unit, a partitioning unit, a deletion and supplementation unit, and a training unit. The acquisition unit is used to acquire drilling information and cuttings logging data corresponding to the well to be logged and the drilled portion of the well in the area where the well to be logged is located. The division unit is used to divide the drilling information and cuttings logging data corresponding to the drilled part according to the formation lithology and well body, so as to obtain drilling information and cuttings logging data corresponding to different formation lithologies and different well bodies; The deletion and supplementation unit is used to delete outliers in the drilling information and cuttings logging data corresponding to the different formation lithologies and different well bodies, and to supplement missing values ​​in the drilling information and cuttings logging data corresponding to the different formation lithologies and different well bodies to obtain training data. The training unit is used to train the error backpropagation neural network model by taking the drilling information in the training data as input data and the cuttings logging data in the training data as output data, and to determine the model whose performance parameters after training are greater than the preset parameters as the salt-gypsum layer geological card layer identification model. The number of adjacent wells is multiple, and the weight of the stratigraphic information of each adjacent well is negatively correlated with the distance between the corresponding adjacent well and the well to be logged; the search module is used to search for the next candidate stratigraphic information corresponding to the stratigraphic information of the adjacent wells of the well to be logged in the known stratigraphic information of each adjacent well if the identification result does not represent a salt-gypsum layer geology. The candidate stratigraphic information corresponding to each adjacent well is weighted with a weight to obtain different candidate stratigraphic information and their corresponding coefficients. The candidate layer information corresponding to the maximum coefficient is determined as the next layer information.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 4.

Images