Pump pressure prediction methods, devices, equipment, storage media, and program products

Abstract

This application provides a pump pressure prediction method, apparatus, device, storage medium, and program product. The method includes: acquiring data from multiple adjacent wells during the fracturing stage; inputting the data from multiple adjacent wells during the fracturing stage into a physical model of hydraulic fracturing, and outputting predicted wellhead pressure data; inputting the predicted wellhead pressure data and the data from multiple adjacent wells during the fracturing stage into a pre-trained correction model, and outputting corrected wellhead pressure data; and generating target pump pressure data based on the corrected wellhead pressure data and the predicted wellhead pressure data. Combining the physical model and the correction model to predict pump pressure improves the accuracy of the prediction.

Description

Technical Field

[0001] This application relates to the field of hydraulic fracturing technology, and in particular to a pump pressure prediction method, apparatus, equipment, storage medium, and program product. Background Technology

[0002] Hydraulic fracturing is a crucial technology in unconventional oil and gas resource extraction. It involves injecting high-pressure fluid into shale formations to fracture the rock and extract oil and gas. In actual fracturing operations, real-time pump pressure stability is critical. A continuous drop in pump pressure may lead to insufficient fracture propagation, affecting fracturing effectiveness; conversely, a continuous increase in pump pressure may cause operational accidents.

[0003] In the prior art, pump pressure values ​​are predicted through pump pressure prediction models to ensure pump pressure stability.

[0004] However, the pump pressure models used in existing technologies are mostly single models, and single models have the problem of inaccurate prediction of pump pressure parameters. Summary of the Invention

[0005] This application provides a pump pressure prediction method, apparatus, device, storage medium, and program product to solve the problem of inaccurate prediction of pump pressure parameters by a single model in the prior art.

[0006] In a first aspect, embodiments of this application provide a pump pressure prediction method, including:

[0007] Acquire data from multiple adjacent wells during the fracturing stage;

[0008] Input data from multiple adjacent wells during the fracturing stage into the physical model of hydraulic fracturing, and output predicted wellhead pressure data;

[0009] The predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage are input into a pre-trained correction model, which outputs corrected wellhead pressure data.

[0010] Based on the corrected wellhead pressure data and the predicted wellhead pressure data, target pump pressure data is generated.

[0011] In one possible implementation, before inputting the predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage into the pre-trained correction model, the method further includes: generating a training dataset based on historical datasets of multiple wellheads during the fracturing stage and historical predicted wellhead pressure data output by the physical model of hydraulic fracturing; defining a prior distribution of the correction model and determining the initial parameters of the correction model based on the prior distribution; training the initial parameters of the correction model based on the training dataset to generate optimized model parameters; and determining the correction model based on the optimized model parameters.

[0012] In one possible implementation, after training the initial parameters of the modified model based on the training dataset and generating the optimized model parameters, the method further includes: acquiring monitoring data of the fracturing and sand-addition stage according to a preset time window length; calculating and generating multiple sets of marginal likelihood functions of the modified model based on the monitoring data of the fracturing and sand-addition stage; and determining the corrected model parameters based on the multiple sets of marginal likelihood functions of the modified model.

[0013] In one possible implementation, the physical model of the hydraulic fracturing is:

[0014]

[0015] In the formula, Indicates wellhead pressure; Indicates the pressure at the bottom of the well; Indicates hydrostatic pressure; This indicates the pressure drop due to pipe friction; Indicates the perforation pressure; This indicates near-wellbore bending friction.

[0016] In one possible implementation, after generating the target pump pressure data based on the corrected wellhead pressure data and the predicted wellhead pressure data, the method further includes: acquiring the interface information of the pump pressure prediction software; generating a packaged file of the pump pressure prediction model based on the interface information of the pump pressure prediction software, the physical model of hydraulic fracturing, and the corrected model; and sending the packaged file of the pump pressure prediction model to the pump pressure prediction software, so that the pump pressure prediction software generates a model calling instruction based on the packaged file of the pump pressure prediction model to complete the integration of the pump pressure prediction model.

[0017] In one possible implementation, after sending the encapsulated file of the pump pressure prediction model to the pump pressure prediction software, the method further includes: sending an integration test instruction to the pump pressure prediction software, so that the pump pressure prediction software generates integrated test data of pump pressure according to the integration test instruction; receiving the integrated test data sent by the pump pressure prediction software, and determining whether the prediction result in the integrated test data exceeds the test deviation range; if the prediction result in the integrated test data exceeds the test deviation range, generating a rollback instruction, and sending the rollback instruction to the pump pressure prediction software, so that the pump pressure prediction software resets the pump pressure prediction model according to the rollback instruction.

[0018] Secondly, embodiments of this application provide a pump pressure prediction device, comprising:

[0019] The first acquisition module is used to acquire data from multiple adjacent wells during the fracturing stage;

[0020] The first output module is used to input data from multiple adjacent wells during the fracturing stage into the physical model of hydraulic fracturing and output predicted wellhead pressure data.

[0021] The second output module is used to input the predicted wellhead pressure data and the data from multiple adjacent wells during the fracturing stage into the pre-trained correction model, and output the corrected wellhead pressure data.

[0022] The first generation module is used to generate target pump pressure data based on the corrected wellhead pressure data and the predicted wellhead pressure data.

[0023] Thirdly, embodiments of this application provide a pump pressure prediction device, including: a memory and a processor;

[0024] The memory stores computer-executed instructions;

[0025] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0026] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0027] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0028] The pump pressure prediction method, apparatus, equipment, storage medium, and program product provided in this application acquire data from multiple adjacent wells during the fracturing stage, generate predicted wellhead pressure data using a physical model of hydraulic fracturing, correct the predicted data using the predicted wellhead pressure data and the data from multiple adjacent wells during the fracturing stage through a correction model, output corrected wellhead pressure data, and generate target pump pressure data based on the corrected wellhead pressure data and the predicted wellhead pressure data. By combining the physical model and the corrected model to predict pump pressure, the accuracy of the prediction is improved. Attached Figure Description

[0029] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0030] Figure 1 This is a schematic diagram of the system structure of a computer device provided in an embodiment of this application;

[0031] Figure 2A schematic flowchart of the pump pressure prediction method provided in this application;

[0032] Figure 3 A schematic diagram of the pump pressure prediction curve provided in the embodiments of this application;

[0033] Figure 4 A schematic diagram illustrating the training and correction model provided in an embodiment of this application;

[0034] Figure 5 A schematic diagram of online calibration provided for embodiments of this application;

[0035] Figure 6 A schematic diagram of the pump pressure prediction device provided in this application;

[0036] Figure 7 A schematic diagram of the pump pressure prediction device provided in this application.

[0037] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0038] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0039] Hydraulic fracturing is a crucial technology in unconventional oil and gas resource extraction. It involves injecting high-pressure fluid into shale formations to fracture the rock and extract oil and gas. In actual fracturing operations, maintaining stable pump pressure in real time is critical. A continuous drop in pump pressure can lead to insufficient fracture propagation, affecting fracturing effectiveness; conversely, a continuous rise in pump pressure can cause operational accidents. Current technologies use pump pressure prediction models to forecast pump pressure values ​​to ensure stability. However, existing pump pressure models are often single models, which suffer from inaccurate predictions of pump pressure parameters.

[0040] To address the aforementioned technical problems, this application proposes the following technical concept: The inventors considered designing a physical model for hydraulic fracturing and a pre-trained correction model. By acquiring data from multiple adjacent wells during the fracturing stage, the physical model for hydraulic fracturing is used to generate predicted wellhead pressure data. Using the predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage, the correction model is used to correct the predicted data, outputting corrected wellhead pressure data. Based on the corrected data and the predicted data, target pump pressure data is generated. Detailed embodiments are described below.

[0041] Figure 1 This is a schematic diagram of the system architecture of the computer device provided in an embodiment of this application. Figure 1 As shown, the computer device includes: a receiving device 101, a processing device 102, and a display device 103.

[0042] It is understood that the structure illustrated in the embodiments of this application does not constitute a specific limitation on the pump pressure prediction method. In other feasible embodiments of this application, the above architecture may include more or fewer components than illustrated, or combine some components, or split some components, or arrange different components, which can be determined according to the actual application scenario and is not limited here. Figure 1 The components shown can be implemented in hardware, software, or a combination of both.

[0043] In the specific implementation process, the receiving device 101 can be an input / output interface or a communication interface, and can acquire data from multiple adjacent wells during the fracturing stage.

[0044] The processing device 102 can generate target data for pump pressure.

[0045] The display device 103 can be used to display target data such as pump pressure.

[0046] The display device can also be a touch screen, used to receive user commands while displaying the above content, so as to realize the operation interaction with the user.

[0047] It should be understood that the above-mentioned processing device can be implemented by a processor reading instructions from memory and executing those instructions, or it can be implemented by a chip circuit.

[0048] Furthermore, the network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0049] Figure 2A flowchart illustrating the pump pressure prediction method provided in this application is shown below. Figure 2 As shown, the method includes:

[0050] S201: Acquire data from multiple adjacent wells during the fracturing stage.

[0051] In this embodiment, the data from multiple adjacent wells includes static and dynamic data of multiple sections to be fractured in multiple adjacent oil and gas wells in the geological structure where the fracturing stage is located.

[0052] Static data includes, but is not limited to, geological stratification information, fracturing process data, wellbore dimensions, and bridge plug locations.

[0053] The dynamic data includes, but is not limited to, time, pumping rate, rate of change of pumping rate, rate of change of proppant concentration, proppant concentration, total liquid volume, and total proppant volume.

[0054] S202: Input data from multiple adjacent wells during the fracturing stage into the physical model of hydraulic fracturing, and output predicted wellhead pressure data.

[0055] In this embodiment, the physical model of hydraulic fracturing is:

[0056]

[0057] In the formula, Indicates wellhead pressure; Indicates the pressure at the bottom of the well; Indicates hydrostatic pressure; This indicates the pressure drop due to pipe friction; Indicates the perforation pressure; This indicates near-wellbore bending friction.

[0058] In this embodiment, the data input to the physical model includes, but is not limited to, hydrostatic pressure, tubing friction pressure drop, perforation pressure, and near-wellbore bending friction.

[0059] S203: Input the predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage into the pre-trained correction model, and output the corrected wellhead pressure data.

[0060] Specifically, the preliminary predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage are used as input features to input into the pre-trained correction model. The correction model predicts the error correction value of the physical model based on the input features and the error features learned during pre-training.

[0061] S204: Generate target pump pressure data based on the corrected wellhead pressure data and the predicted wellhead pressure data.

[0062] Specifically, the preliminary predicted value and the error correction value are added together to obtain the final predicted wellhead pressure value.

[0063] In this embodiment, the target data for pump pressure includes, but is not limited to, drilling fluid pressure, drilling fluid flow rate, and polymer consumption.

[0064] Figure 3 This is a schematic diagram of the pump pressure prediction curve provided in an embodiment of this application.

[0065] like Figure 3 As shown, Figure 3 Curve (1) is a graph showing the change of wellhead pressure over time; curve (2) is a graph showing the change of proppant flow rate over time; curve (3) is a graph showing the change of proppant concentration over time.

[0066] As can be seen from the above embodiments, by acquiring data from multiple adjacent wells during the fracturing stage, predictive data of wellhead pressure is generated using the physical model of hydraulic fracturing. The predicted data of wellhead pressure and data from multiple adjacent wells during the fracturing stage are then used to correct the predicted data through a correction model, and corrected data of wellhead pressure is output. Target data of pump pressure is generated based on the corrected data of wellhead pressure and the predicted data of wellhead pressure. By combining the physical model and the correction model to predict pump pressure, the accuracy of the prediction is improved.

[0067] In one embodiment of this application, before step S203, the method further includes:

[0068] S301: Generate a training dataset based on historical datasets of multiple wellheads during the fracturing stage and historical prediction data of wellhead pressure output from the physical model of hydraulic fracturing.

[0069] In this embodiment, the training dataset includes input features and the corresponding error values ​​of the input features.

[0070] The input features include, but are not limited to, time, pumping rate, and proppant concentration.

[0071] S302: Define the prior distribution of the modified model and determine the initial parameters of the modified model based on the prior distribution.

[0072] In this embodiment, the prior distribution of the modified model includes the mean function and the covariance function.

[0073] The mean function is a linear function, and the covariance function is a rational quadratic covariance function.

[0074] In this embodiment, the modified model is a Gaussian process regression model.

[0075] S303: Train the initial parameters of the corrected model based on the training dataset, and generate the optimized model parameters.

[0076] In this embodiment, the hyperparameters of the Gaussian process regression model are optimized using the Bayesian method.

[0077] The hyperparameters are the initial parameters.

[0078] Specifically, the maximum likelihood estimation method is used to obtain the minimum value of the negative log-likelihood function, and the optimized model parameters are determined based on the hyperparameters corresponding to the minimum value.

[0079] S304: Determine the corrected model based on the optimized model parameters.

[0080] Specifically, based on the optimized hyperparameters, a trained Gaussian process regression model is established, and the error of the physical model is compensated and corrected through the trained Gaussian process regression model.

[0081] Figure 4 This is a schematic diagram of the training and correction model provided in the embodiments of this application.

[0082] like Figure 4 As shown, the input parameters of the model are determined, and the model is trained and corrected using Gaussian process regression. The corrected model trained by Gaussian process regression is used to make a preliminary prediction of the pump pressure and generate a preliminary prediction value.

[0083] In Gaussian process regression, the curves are visualizations of the probability distribution, including the prior curve, likelihood curve, and posterior curve, corresponding to the expected curve, actual curve, and predicted curve, respectively. The prior curve represents uncertainty, independent of observed data, and is an initial conjecture about the output. The likelihood curve is calculated after obtaining the observed input and output samples, and is affected by noise. The posterior curve combines the prior curve (initial conjecture) and the likelihood curve (constraints from real data), obtained through Bayes' theorem, representing the final conjecture after model training and closely resembling the real data.

[0084] Among them, the predicted value is within the prediction confidence interval.

[0085] As can be seen from the above embodiments, by acquiring historical datasets of multiple wellheads during the fracturing stage and historical datasets of wellhead pressure output by the physical model of hydraulic fracturing, a training dataset is constructed. The prior distribution of the correction model is defined using the Bayesian method, the initial parameters of the correction model are determined, the initial parameters are trained using the training dataset, and the correction model is optimized, thereby improving the accuracy of the correction model's predictions.

[0086] In one embodiment of this application, after step S303, the method further includes:

[0087] S401: Obtain monitoring data for the fracturing and sand-addition stage according to the preset time window length.

[0088] Specifically, during the fracturing and sand-addition stage, on-site monitoring data is collected in real time, with a data collection frequency of once per second.

[0089] The on-site monitoring data includes, but is not limited to, time, pumping rate, proppant concentration, and wellhead pressure.

[0090] Specifically, the latest field monitoring data within a fixed time window prior to the current moment is selected as the dataset for model calibration.

[0091] S402: Calculate and generate multiple sets of marginal likelihood functions for the modified model based on monitoring data from the fracturing and sand-addition stage.

[0092] Specifically, the hyperparameters of the modified model are re-optimized using Bayesian methods, and the marginal likelihood function of the model under the new dataset is calculated.

[0093] S403: Determine the corrected model parameters based on multiple sets of marginal likelihood functions of the corrected model.

[0094] Specifically, the hyperparameter combination with the largest likelihood function is obtained based on the calculated multiple sets of marginal likelihood functions, and the corrected model parameters are determined based on the hyperparameter with the largest likelihood function.

[0095] Specifically, as fracturing operations progress, new field monitoring data are continuously acquired, and the model is constantly updated and corrected to ensure the accuracy of the prediction results.

[0096] Figure 5 This is a schematic diagram illustrating online calibration provided in an embodiment of this application.

[0097] like Figure 5 As shown, the modeling data of adjacent test sections are stored, the target section is predicted by the correction model, the real-time data stream is processed according to the time node, and the data segment is extracted by sliding window at each time node to be used as the dataset for correction model calibration. The correction model is optimized and continuously updated as the fracturing operation progresses to improve the accuracy of the prediction data.

[0098] As can be seen from the above embodiments, by introducing online model calibration technology, monitoring data of the fracturing and sand-addition stage with a preset time window length is obtained. Multiple sets of marginal likelihood functions of the modified model are calculated and generated based on the monitoring data of the fracturing and sand-addition stage. The corrected model parameters are determined based on the maximum likelihood function. By introducing online model calibration technology, the modified model can adapt to the dynamic changes in the fracturing construction process, thereby improving the accuracy of prediction.

[0099] In one embodiment of this application, after step S204, the method further includes:

[0100] S205: Obtain interface information from the pump pressure prediction software.

[0101] In this embodiment, the interface information includes, but is not limited to, interface type, interface protocol, and data format.

[0102] S206: Generate a packaged file of the pump pressure prediction model based on the interface information of the pump pressure prediction software, the physical model of hydraulic fracturing, and the modified model.

[0103] Specifically, the output data formula and communication protocol of the pump pressure prediction model are determined based on the interface information, and the physical model and the modified model of hydraulic fracturing are encapsulated into a class to obtain the encapsulation file.

[0104] S207: Send the encapsulated file of the pump pressure prediction model to the pump pressure prediction software, so that the pump pressure prediction software can generate model call instructions based on the encapsulated file of the pump pressure prediction model to complete the integration of the pump pressure prediction model.

[0105] Specifically, the packaged file is sent to the pump pressure prediction software. The pump pressure prediction software generates a call instruction based on the packaged file. When the input features are received, the integrated pump pressure prediction model is invoked through the call instruction. The input features are then input into the pump pressure prediction model to generate the prediction result.

[0106] As can be seen from the above embodiments, by obtaining the interface information of the pump pressure prediction software, generating a packaged file based on the interface information, the physical model and the correction model of hydraulic fracturing, and sending the packaged file to the pump pressure prediction software, the pump pressure prediction software can be integrated, and the prediction model can be integrated into the software, thereby improving the accuracy of the pump pressure prediction software.

[0107] In one embodiment of this application, after step S207, the method further includes:

[0108] S208: Send an integrated test command to the pump pressure prediction software so that the pump pressure prediction software generates integrated test data of the pump pressure according to the integrated test command.

[0109] Specifically, the integration test dataset includes the input features of the integration test and the prediction results of the integration test.

[0110] S209: Receive integrated test data sent by the pump pressure prediction software and determine whether the prediction results in the integrated test data exceed the test deviation range.

[0111] Specifically, the input features of the integration test are input into the pump pressure prediction software, and the pump pressure prediction software outputs the model prediction results.

[0112] S210: If the prediction result in the integrated test data exceeds the test deviation range, a rollback command is generated and sent to the pump pressure prediction software so that the pump pressure prediction software resets the pump pressure prediction model according to the rollback command.

[0113] Specifically, the prediction deviation between the model prediction results and the integrated test results is compared. If the prediction deviation exceeds the test deviation range, a rollback command is generated, and the pump pressure prediction software resets the pump pressure prediction model according to the rollback command.

[0114] As can be seen from the above embodiments, by performing integrated testing on the integrated pump pressure prediction software, it is determined whether the integrated test data exceeds the test deviation range. If it exceeds the test deviation, a rollback command is generated. The pump pressure prediction software resets the pump pressure prediction model according to the rollback command, thereby avoiding prediction errors caused by the integrated pump pressure prediction model.

[0115] Figure 6 A schematic diagram of the pump pressure prediction device provided in this application is shown below. Figure 6 As shown, the pump pressure prediction device 60 provided in this embodiment includes: a first acquisition module 601, a first output module 602, a second output module 603, and a first generation module 604.

[0116] The first acquisition module 601 is used to acquire data from multiple adjacent wells during the fracturing stage.

[0117] The first output module 602 is used to input data from multiple adjacent wells during the fracturing stage into the physical model of hydraulic fracturing and output predicted wellhead pressure data.

[0118] The second output module 603 is used to input the predicted wellhead pressure data and data from multiple adjacent wells during the fracturing stage into the pre-trained correction model, and output the corrected wellhead pressure data.

[0119] The first generation module 604 is used to generate target data for pump pressure based on the corrected data and predicted data of wellhead pressure.

[0120] In one possible implementation, the pump pressure prediction device 60 further includes:

[0121] The second generation module is used to generate a training dataset based on historical datasets of multiple wellheads during the fracturing stage and historical prediction data of wellhead pressure output from the physical model of hydraulic fracturing.

[0122] The definition module is used to define the prior distribution of the modified model and determine the initial parameters of the modified model based on the prior distribution.

[0123] The third generation module is used to train the initial parameters of the corrected model based on the training dataset and generate the optimized model parameters.

[0124] The first determination module is used to determine the corrected model based on the optimized model parameters.

[0125] In one possible implementation, the pump pressure prediction device 60 further includes:

[0126] The second acquisition module is used to acquire monitoring data of the fracturing and sand-addition stage according to a preset time window length.

[0127] The calculation module is used to calculate and generate multiple sets of marginal likelihood functions for the modified model based on the monitoring data of the fracturing and sand-addition stage.

[0128] The second determination module is used to determine the corrected model parameters based on multiple sets of marginal likelihood functions of the corrected model.

[0129] In one possible implementation, the pump pressure prediction device 60 further includes:

[0130] The third acquisition module is used to acquire interface information from the pump pressure prediction software.

[0131] The fourth generation module is used to generate a packaged file of the pump pressure prediction model based on the interface information of the pump pressure prediction software and the physical and modified models of hydraulic fracturing.

[0132] The first sending module is used to send the encapsulated file of the pump pressure prediction model to the pump pressure prediction software, so that the pump pressure prediction software can generate model calling instructions based on the encapsulated file of the pump pressure prediction model to complete the integration of the pump pressure prediction model.

[0133] In one possible implementation, the pump pressure prediction device 60 further includes:

[0134] The second sending module is used to send integrated test instructions to the pump pressure prediction software, so that the pump pressure prediction software can generate integrated test data of pump pressure according to the integrated test instructions.

[0135] The receiving module is used to receive the integrated test data sent by the pump pressure prediction software and determine whether the prediction results in the integrated test data exceed the test deviation range.

[0136] The fifth generation module is used to generate a rollback instruction if the prediction result in the integrated test data exceeds the test deviation range, and send the rollback instruction to the pump pressure prediction software so that the pump pressure prediction software can reset the pump pressure prediction model according to the rollback instruction.

[0137] The pump pressure prediction device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0138] Figure 7 This is a schematic diagram of the pump pressure prediction device provided in this application. Figure 7 As shown, the pump pressure prediction device 70 provided in this embodiment includes at least one processor 701 and a memory 702. Optionally, the pump pressure prediction device 70 further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0139] In the specific implementation process, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to execute the above-described pump pressure prediction method.

[0140] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0141] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0142] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0143] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0144] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described pump pressure prediction method.

[0145] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described pump pressure prediction method.

[0146] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0147] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0148] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0149] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0150] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0151] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0153] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A pump pressure prediction method characterized by, The application is applied to a computer device, comprising: obtaining a plurality of adjacent well data of a fracturing stage; inputting the plurality of adjacent well data of the fracturing stage into a physical model of hydraulic fracturing to output predicted data of wellhead pressure; inputting the predicted data of wellhead pressure and the plurality of adjacent well data of the fracturing stage into a pre-trained correction model to output correction data of wellhead pressure; generating target data of pump pressure according to the correction data of wellhead pressure and the predicted data of wellhead pressure; before the inputting, further comprising: generating a training data set according to a historical data set of a plurality of wellheads of the fracturing stage and historical predicted data of wellhead pressure output by the physical model of hydraulic fracturing; defining a prior distribution of the correction model and determining initial parameters of the correction model according to the prior distribution; training the initial parameters of the correction model according to the training data set to generate optimized model parameters; determining the correction model according to the optimized model parameters; after the training, further comprising: obtaining monitoring data of a fracturing and sanding stage according to a preset time window length; calculating a plurality of marginal likelihood functions of the correction model according to the monitoring data of the fracturing and sanding stage; determining corrected model parameters according to the plurality of marginal likelihood functions of the correction model.

2. The method of claim 1, wherein, the physical model of hydraulic fracturing is: wherein Pw represents wellhead pressure; Pwf represents bottom hole pressure; Psl represents static liquid pressure; Ptf represents pipe friction pressure drop; Pp represents perforation pressure; Pmb represents near wellbore bending friction.

3. The method of claim 1, wherein, after the generating, further comprising: obtaining interface information of pump pressure prediction software; generating an encapsulation file of a pump pressure prediction model according to the interface information of the pump pressure prediction software, the physical model of hydraulic fracturing and the correction model; sending the encapsulation file of the pump pressure prediction model to the pump pressure prediction software to enable the pump pressure prediction software to generate model calling instructions according to the encapsulation file of the pump pressure prediction model to complete integration of the pump pressure prediction model.

4. The method of claim 3, wherein, after the sending, further comprising: sending integration test instructions to the pump pressure prediction software to enable the pump pressure prediction software to generate integration test data of pump pressure according to the integration test instructions; receiving the integration test data sent by the pump pressure prediction software and judging whether a predicted result in the integration test data is out of a test deviation range; if the predicted result in the integration test data is out of the test deviation range, generating a rollback instruction and sending the rollback instruction to the pump pressure prediction software to enable the pump pressure prediction software to reset the pump pressure prediction model according to the rollback instruction.

5. A pump pressure prediction device characterized by comprising: the pump pressure prediction device is used to implement the pump pressure prediction method in any one of claims 1-4, comprising: a first obtaining module for obtaining a plurality of adjacent well data of a fracturing stage; a first output module for inputting the plurality of adjacent well data of the fracturing stage into a physical model of hydraulic fracturing to output predicted data of wellhead pressure; The second output module is configured to input the predicted data of the wellhead pressure and the plurality of adjacent well data of the fracturing stage into a pre-trained correction model, and output correction data of the wellhead pressure. The first generation module is configured to generate target data of the pump pressure according to the correction data of the wellhead pressure and the predicted data of the wellhead pressure.

6. A pump pressure prediction device characterized by comprising: It comprises: a memory and a processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory, so that the processor executes the pump pressure prediction method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer execution instructions, and the computer execution instructions are executed by the processor to implement the pump pressure prediction method according to any one of claims 1 to 4.

8. A computer program product, characterised in that, It comprises a computer program, which is executed by the processor to implement the pump pressure prediction method according to any one of claims 1 to 4.

Images