Well logging while drilling risk prediction method, device and system based on digital twinning

Abstract

The invention relates to a logging-while-drilling risk prediction method, device and system based on digital twinning, and belongs to the technical field of logging risk prediction.The method comprises the steps that simulation is conducted through a virtual digital twinning body of a physical drilling system, and simulation state information of the physical drilling system is obtained; inputting the simulation state information and / or actual measurement state information of the physical drilling system into a trained risk prediction model to obtain first risk information; obtaining second risk information according to the simulation state information in combination with a physical principle followed by the physical drilling system; and comprehensively determining the risk of the physical drilling system according to the first risk information and the second risk information. A physical mechanism and an artificial intelligence algorithm are combined to cooperatively predict the drilling risk, and the reliability and scientificity of a risk prediction result can be improved.

Description

Technical Field

[0001] This invention relates to the field of well logging risk prediction technology, and in particular to a method, device and system for predicting well logging risk based on digital twins. Background Technology

[0002] Logging while drilling (LOD) is a key technology for acquiring downhole information and formation parameters in real time during modern oil and gas drilling operations. Through downhole sensors and real-time data transmission systems, it can continuously acquire and monitor various parameters such as resistivity, acoustic waves, gamma rays, and pressure, providing important information for drilling decisions.

[0003] However, existing logging-while-drilling methods still have significant limitations in terms of safety risk early warning. Current methods largely rely on empirical formulas for downhole condition assessment and risk identification, resulting in high false alarm and false negative rates. These methods cannot meet the demand for advanced and accurate early warning of safety risks in challenging operations such as deep wells and wells with complex structures. Summary of the Invention

[0004] In view of this, it is necessary to provide a method, device and system for predicting risks while drilling based on digital twins, so as to solve the problem of insufficient accuracy of existing logging-while-drilling safety risk early warning.

[0005] To address the aforementioned problems, in a first aspect, the present invention provides a logging-while-drilling risk prediction method based on digital twins, comprising: Simulation of the physical drilling system is performed using a virtual digital twin of the physical drilling system to obtain the simulation status information of the physical drilling system. The simulation state information and / or the measured state information of the physical drilling system are input into the trained risk prediction model to obtain the first risk information; Based on the simulation state information and the physical principles followed by the physical drilling system, the second risk information is obtained; The risks of the physical drilling system are determined comprehensively based on the first risk information and the second risk information.

[0006] In one possible implementation, the first risk information includes the overflow probability, and the second risk information includes the ratio of bottom hole pressure to formation pressure; the step of comprehensively determining the risk of the physical drilling system based on the first risk information and the second risk information includes: The overflow risk index is calculated using the following formula:

[0007] In the formula, These are the weighting coefficients. For the bottom hole pressure, For formation pressure, For overflow probability, Increase the volume of the mud pit. Threshold volume of mud pit In one possible implementation, the first risk information includes drill string failure types, and the risk of the physical drilling system is comprehensively determined based on the first risk information and the second risk information, including: The cumulative damage of the drill string is calculated using the following formula:

[0008] In the formula, For the drill bit at stress level The number of loops below, ( M ) for the drill bit at stress level Number of failure cycles under each drill string failure type; The risk of the physical drilling system is determined by combining the cumulative damage of the drill string and the second risk information.

[0009] In one possible implementation, the risk prediction model includes a drill string failure type identification model; the step of inputting the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model to obtain first risk information includes: Based on the simulation state information or the measured state information of the physical drilling system, a vibration spectrum image of the drill string of the physical drilling system is generated. The vibration spectrum image is input into the trained drill string fault type recognition model to obtain the drill string fault type.

[0010] In one possible implementation, obtaining the second risk information based on the simulation state information and the physical principles followed by the physical drilling system includes: Based on the simulation state information and the Mohr-Coulomb criterion, the wellbore stability assessment results are obtained; Based on the simulation state information and wellbore trajectory covariance analysis, the collision risk between adjacent wells is obtained.

[0011] In one possible implementation, the simulation via a virtual digital twin of the physical drilling system includes: Construct a virtual digital twin of the physical drilling system and a multiphysics model; the digital twin includes a physical entity model of the physical drilling system and a simulation engine; The simulation engine performs simulations based on the multiphysics model and the physical entity model.

[0012] In one possible implementation, the physical entity model includes: a geological model, a wellbore trajectory model, a drill string system model, and a fluid system model; the multiphysics model includes a wellbore multiphase flow transient model, an electromagnetic field propagation and inversion model, and a drill string dynamics model.

[0013] In one possible implementation, the method further includes: The parameters of the virtual digital twin are updated using extended Kalman filtering.

[0014] Secondly, the present invention also provides a logging-while-drilling risk prediction device based on digital twins, comprising: The simulation module is used to perform simulation through a virtual digital twin of the physical drilling system to obtain the simulation status information of the physical drilling system. The first fault prediction module is used to input the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model to obtain the first risk information. The second fault prediction module is used to obtain second risk information based on the simulation state information and the physical principles followed by the physical drilling system. The drilling risk determination module is used to comprehensively determine the risk of the physical drilling system based on the first risk information and the second risk information.

[0015] Thirdly, the present invention also provides a logging-while-drilling risk prediction system based on digital twins, comprising: a data acquisition device, a well site server, a cloud platform, and a terminal device that are connected in sequence via communication. The data acquisition device is deployed at the well site to collect multi-source data from the well site. The well site server is used to preprocess the multi-source data and then send it to the cloud platform; The cloud platform is used to execute any of the steps in the above-described digital twin-based logging-while-drilling risk prediction method based on the preprocessed multi-source data. The terminal device is used to present the interactive interface of the cloud platform.

[0016] The beneficial effects of this invention are: This invention uses a virtual digital twin of the physical drilling system to simulate the system and obtain its simulation state information. Then, based on this simulation state information, combined with a trained drilling system fault diagnosis model and the physical principles followed by the physical drilling system, the risks of the physical drilling system are predicted. By combining physical mechanisms and artificial intelligence algorithms for collaborative prediction, the reliability and scientific validity of the risk prediction results can be improved. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating an embodiment of the logging-while-drilling risk prediction method based on digital twins provided by the present invention. Figure 2 A multi-dimensional risk assessment framework provided by the present invention; Figure 3 A schematic diagram of a real-time data processing process provided by the present invention; Figure 4 A schematic diagram of the structure of an embodiment of the logging-while-drilling risk prediction device based on digital twin provided by the present invention; Figure 5 A schematic diagram of an embodiment of the logging-while-drilling risk prediction system based on digital twins provided by the present invention; Figure 6 A framework diagram of a logging-while-drilling risk prediction system provided by the present invention; Figure 7 This is another framework diagram of a logging-while-drilling risk prediction system provided by the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] In the description of the embodiments of this invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0021] In the embodiments of this invention, the terms "first," "second," etc., are used to distinguish similar objects, and are not used to describe a specific order or sequence, nor to indicate or imply their relative importance or implicitly specify the number of technical features indicated. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, and the number of objects is not limited; for example, the first object can be one or more.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] Reference Figure 1 The diagram illustrates a flowchart of an embodiment of the logging-while-drilling risk prediction method based on digital twins provided by the present invention. The method includes: S101, through simulation using a virtual digital twin of the physical drilling system, obtains the simulation status information of the physical drilling system.

[0024] A physical drilling system (PDS) consists of formation, well, drill string, and multiphase fluids. A fully corresponding virtual digital twin can be constructed based on the measured state information of the PDS, enabling collaborative sharing of data from all components, visualization of complex downhole environments and dynamic processes, and simulation prediction. The measured state information can be acquired in real-time through data interfaces, including: downhole measurement-while-drilling data (gamma, resistivity, acoustic waves, caliper diameter, vibration, temperature, pressure, etc.), surface monitoring data (standby pressure, casing pressure, displacement, hook load, torque, rotational speed, etc.), adjacent well data (trajectory data, geological data, engineering data, etc.), and geological steering data (formation prediction models, actual drilling comparison data, etc.).

[0025] Simulation state information can correspond to measured state information, such as simulation values ​​of downhole drilling measurement data.

[0026] S102, input the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model to obtain the first risk information.

[0027] Risk prediction models can include multiple models, such as drill string failure type identification models, well logging blowout risk identification models, and anomaly data detection models. The types of risk prediction models can be selected and designed according to requirements. Training data for risk prediction models can be obtained from historical simulation state information and / or historical measured state information. For example, drill string data from historical simulation state information and / or historical measured state information can be used to train the drill string failure type identification model.

[0028] S103. Based on the simulation status information and the physical principles followed by the physical drilling system, the second risk information is obtained.

[0029] The second risk information of the physical drilling system can be obtained by calculating the simulation state information combined with the expression of physical principles.

[0030] S104. Based on the first risk information and the second risk information, the risks of the physical drilling system are comprehensively determined.

[0031] For assessments of the same risk in both the first and second risk information sets, a weighted summation can be performed to synthesize the model predictions and physical mechanism assessments. For assessments of different risks in the first and second risk information sets, each assessment can be retained separately.

[0032] Based on the first risk information and the second risk information, the risk information of the physical drilling system in various dimensions can be comprehensively determined.

[0033] The logging-while-drilling risk prediction method based on digital twins provided in this embodiment can be applied to a logging-while-drilling risk prediction software system based on digital twins, which can run on a cloud platform.

[0034] In summary, this embodiment uses a virtual digital twin of the physical drilling system for simulation to obtain the simulation state information of the physical drilling system. Then, based on the simulation state information, combined with the trained drilling system fault diagnosis model and the physical principles followed by the physical drilling system, the risks of the physical drilling system are predicted. By combining physical mechanisms and artificial intelligence algorithms for collaborative prediction, the reliability and scientific nature of the risk prediction results can be improved.

[0035] In some embodiments of the present invention, the first risk information includes the overflow probability, and the second risk information includes the ratio of bottom hole pressure to formation pressure; S104 includes: The overflow risk index is calculated using the following formula:

[0036] In the formula, These are the weighting coefficients. For the bottom hole pressure, For formation pressure, For overflow probability, Increase the volume of the mud pit. This represents the threshold volume of the mud pit.

[0037] In this embodiment, based on the principle of bottom hole pressure balance:

[0038] In the formula, For the bottom hole pressure, = This refers to the hydrostatic pressure. For annular circulation friction, For casing pressure (if the wellhead is closed). This refers to formation pressure.

[0039] It is known that when the bottom hole pressure is less than or equal to the formation pressure, that is, when the ratio of bottom hole pressure to formation pressure is less than 1, the bottom hole pressure is balanced and the risk of overflow is low. Therefore, this embodiment obtains a parameter characterizing the overflow risk based on the bottom hole pressure balance principle. Based on this parameter and the risk prediction model, the overflow probability is predicted and a comprehensive calculation is performed to obtain the overflow risk index, which can improve the accuracy of overflow risk diagnosis.

[0040] In some embodiments of the present invention, the first risk information includes the drill string failure type, and S104 includes: The cumulative damage of the drill string is calculated using the following formula:

[0041] In the formula, For the drill bit at stress level The number of loops below, ( M ) for the drill bit at stress level Number of failure cycles under each drill string failure type; The risk of the physical drilling system is determined by combining the cumulative damage of the drill string and the secondary risk information.

[0042] In some embodiments of the present invention, the remaining life of the drill string can be further calculated based on the cumulative damage degree of the drill string, and the calculation formula is as follows:

[0043] In the formula, For remaining lifespan, This represents the total number of failure cycles. The value is the rotational speed.

[0044] In summary, this embodiment, through the deep integration of AI models such as LSTM-based time series prediction and CNN-based fault diagnosis with physical models, can improve the early warning accuracy to over 95%.

[0045] In some embodiments of the present invention, S102 includes: Based on the simulation state information or the measured state information of the physical drilling system, a vibration spectrum image of the drill string of the physical drilling system is generated. The vibration spectrum image is input into the trained drill string fault type recognition model to obtain the drill string fault type.

[0046] In some embodiments of the present invention, S103 includes: Based on the simulation state information and the Mohr-Coulomb criterion, the wellbore stability assessment results are obtained; Based on the simulation status information and well trajectory covariance analysis, the collision risk between adjacent wells is obtained.

[0047] Specifically, the expression for the Moore-Coulomb criterion is as follows:

[0048] in, c is the shear strength, and c is the cohesive force. For normal stress, It is the internal friction angle.

[0049] The covariance expression is as follows:

[0050] In the formula, For northbound variance, For eastward variance, The covariance is in the northeast direction.

[0051] The simulation state information includes the distance to neighboring wells. Based on the distance to neighboring wells, the well trajectory can be determined. Uncertainty analysis of the well trajectory can yield the collision probability with neighboring wells. The uncertainty is described by the covariance expression.

[0052] Furthermore, the risk of collision with adjacent wells can be characterized by a safety factor, the expression of which is as follows:

[0053] In the formula, To measure the distance between adjacent wells, For the safety radius, denoted as the standard deviation of the relative position uncertainty.

[0054] In summary, this embodiment establishes a complete risk prevention and control system by creating multiple safety barriers, including overflow early warning, wellbore stability, adjacent well collision prevention, and drill string health.

[0055] Reference Figure 2 This diagram illustrates a multi-dimensional risk assessment framework provided by the present invention. The assessment dimensions include well control safety risk, drilling efficiency risk, and equipment health risk. After determining the multi-dimensional risks, risk levels can be further calculated, and corresponding prevention and control decisions can be generated based on different risk levels.

[0056] In some embodiments of the present invention, different types of risk prediction models can be built on different frameworks. For example, a drill string failure type identification model can be built based on a convolutional neural network (CNN), a well logging overflow risk identification model / drilling parameter prediction model can be built based on a long short-term memory network (LSTM), and a data anomaly detection model can be built based on isolated forests.

[0057] The CNN convolutional layer structure is as follows:

[0058] In the formula, To output feature map at position The value, For convolution kernel weights, For input data, For bias terms, This is the activation function.

[0059] The structure of LSTM is as follows:

[0060] In the formula, For the Gate of Oblivion For input gate, For output gate, In cellular state, In hidden state, This is the weight matrix. For bias terms, for sigmoid function, This indicates element-wise multiplication.

[0061] The anomaly score for data in the isolated forest is as follows:

[0062] In the formula, For the sample Abnormal scores, For the expected path length, This represents the average path length. This represents the number of samples.

[0063] Referring to Table 1, the inputs and outputs of the above risk prediction models are shown.

[0064] Table 1. Schematic diagram of risk prediction model

[0065] In some embodiments of the present invention, the step of simulating a physical drilling system using a virtual digital twin includes: Construct a virtual digital twin of the physical drilling system and a multiphysics model; the digital twin includes the physical entity model of the physical drilling system and a simulation engine; Simulations are performed using a simulation engine based on multiphysics models and physical entity models.

[0066] In this embodiment, the physical entity model includes: a geological model, a wellbore trajectory model, a drill string system model, and a fluid system model; the multiphysics model includes a wellbore multiphase flow transient model, an electromagnetic field propagation and inversion model, and a drill string dynamics model.

[0067] The transient model of multiphase flow in a well includes: The governing equations for multiphase flow are expressed as follows:

[0068] In the formula, For the sake of the prime minister k volume fraction ( ), For the sake of the prime minister k density, For the sake of the prime minister k velocity vector This refers to the mass source term (the rate of inflow from the formation).

[0069] The momentum conservation equation for a mixed fluid is expressed as follows:

[0070]

[0071] In the formula, p m =∑ α k ρ k For mixed density, For mixing speed, P For pressure, For the shear stress tensor, It is the acceleration due to gravity. This represents the frictional resistance of the wellbore.

[0072] Electromagnetic field propagation and inversion models include: The electromagnetic field forward model based on Maxwell's equations is expressed as follows:

[0073] In the formula, For electric field strength, The relative permeability, The relative permittivity, σ For electrical conductivity, ω Angular frequency, The permeability of free space, The vacuum permittivity, j It is the imaginary unit.

[0074] The electromagnetic field inversion model obtained by real-time inversion of the formation resistivity distribution using a regularized inversion algorithm is expressed as follows:

[0075] In the formula, For observation data, For forward modeling data, Reg ( R t ) represents the regularization term.

[0076] The drill string dynamics model includes: The equation for the torsional vibration of the drill string is expressed as follows:

[0077] In the formula, I Let be the polar moment of inertia of the drill string element. θ For the twist angle, GJ For torsional stiffness, T fric The friction torque of the well wall. T bit This represents the drill bit torque.

[0078] The equation for axial vibration is as follows:

[0079] In the formula, u This is axial displacement. ρA Mass per unit length EA For tensile stiffness, F fric For frictional resistance, F bit This is the force exerted by the drill bit.

[0080] The equation for transverse vibration is as follows:

[0081] In the formula, w This is a lateral displacement. EI For bending stiffness, It is a lateral force.

[0082] Among them, the bottom hole pressure can be simulated based on the transient model of multiphase flow in the wellbore, and the distance between adjacent wells can be simulated based on the electromagnetic field propagation and inversion model.

[0083] This embodiment achieves a technological leap from simple "data display" to "mechanism simulation + intelligent prediction" through the synergistic effect of multiphysics coupling modeling and artificial intelligence algorithms, significantly improving the scientific nature and reliability of the early warning system.

[0084] In some embodiments of the present invention, the method further includes: The parameters of the virtual digital twin are updated using extended Kalman filtering.

[0085] Specifically, extended Kalman filtering is used for state estimation and state prediction.

[0086] In the formula, This is the predicted state value. This is the state transition function. To control the input.

[0087] Covariance prediction:

[0088] In the formula, To predict the covariance matrix, Let Jacobian be the state transition matrix. Let be the process noise covariance.

[0089] Kalman gain:

[0090] In the formula, For Kalman gain, For the observation matrix, To observe the noise covariance.

[0091] Status Update:

[0092] In the formula, The updated state estimate, These are actual observed values. For observation functions.

[0093] Covariance update:

[0094] In the formula, This is the updated covariance matrix.

[0095] Model confidence assessment and adaptive model weight adjustment mechanism:

[0096] In the formula, For the first i The weights of each model, For the first i The mean squared error of each model To adjust the parameters, This represents the total number of models.

[0097] In this embodiment, a real-time data processing framework based on extended Kalman filtering is employed, which solves the problem of disconnect between the virtual digital twin and the measured data in traditional methods. This framework can dynamically correct the state parameters of the virtual digital twin, ensuring that the virtual system and the physical drilling system remain highly synchronized, greatly improving the system's adaptability and accuracy under complex drilling conditions.

[0098] Reference Figure 3 This diagram illustrates a real-time data processing procedure provided by the present invention. First, real-time data, such as pressure, flow rate, and vibration, is collected from the well site using sensors deployed at the site. Then, edge nodes, such as a well site server, perform preprocessing on the real-time data, including filtering, compression, and verification. Next, a digital twin combines the preprocessed data with a physical model and an AI model to predict the future state of the well site. Based on the difference between the predicted future state and the actual measured state, an extended Kalman filter is used to correct and update the parameters of the digital twin. Finally, the corrected and updated digital twin is used to assess well site risks, and the assessment results, such as risk values ​​and prevention recommendations, are displayed and fed back to the user through a user interface.

[0099] In some embodiments of the present invention, the logging-while-drilling risk prediction method further includes: providing visualization and decision support: (1) displaying the digital twin status in real time in a three-dimensional visualization interface; (2) displaying the risk level distribution using color coding; (3) providing risk source analysis, displaying the main risk factors and their contribution; (4) generating optimization suggestions and disposal plans to support decision-making. Providing closed-loop control and continuous optimization: (1) establishing an early warning response mechanism to support manual confirmation or automatic control; (2) recording the early warning accuracy and response effect for model optimization; (3) continuously optimizing the early warning threshold and model parameters through reinforcement learning algorithms; (4) establishing a knowledge base to accumulate typical cases and disposal experience.

[0100] Reference Figure 4 The diagram shows a structural schematic of an embodiment of the logging-while-drilling risk prediction device based on digital twins provided by the present invention. The device 400 includes: The simulation module 401 is used to perform simulation through a virtual digital twin of the physical drilling system to obtain the simulation status information of the physical drilling system. The first fault prediction module 402 is used to input the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model to obtain the first risk information. The second fault prediction module 403 is used to obtain second risk information based on the simulation status information and the physical principles followed by the physical drilling system. The drilling risk determination module 404 is used to comprehensively determine the risk of the physical drilling system based on the first risk information and the second risk information.

[0101] It should be noted that the implementation principles or processes of the above modules can be referred to the aforementioned implementation examples of the logging-while-drilling risk prediction method based on digital twins, and will not be elaborated here.

[0102] Reference Figure 5 The diagram shows a structural schematic of an embodiment of the logging-while-drilling risk prediction system based on digital twin provided by the present invention. The system 500 includes: a data acquisition device 501, a well site server 502, a cloud platform 503, and a terminal device 504 that are connected in sequence. The data acquisition device 501 is deployed at the well site to collect multi-source data from the well site. Well site server 502 is used to preprocess multi-source data before sending it to the cloud platform; The cloud platform 503 is used to execute any of the steps in the above-mentioned logging-while-drilling risk prediction method based on digital twins, based on preprocessed multi-source data. Terminal device 504 is used to present the interactive interface of the cloud platform.

[0103] This embodiment achieves optimized allocation of computing resources through a cloud-edge-device (cloud platform-well site server-data acquisition device) collaborative architecture. Edge nodes are responsible for millisecond-level rapid response, while the cloud engine completes second-level fine simulation, effectively solving the bottleneck of real-time performance in traditional systems.

[0104] Reference Figure 6 This diagram illustrates a framework of a logging-while-drilling risk prediction system provided by the present invention. The data acquisition device includes downhole sensors, surface sensors, and control actuators. The well site server is used to perform real-time data processing, data caching, and emergency control logic. The cloud platform is used to build digital twins, train AI models, and perform big data analysis. Terminal devices include PCs and mobile apps, used to enable user interaction with the cloud platform.

[0105] Reference Figure 7 This diagram illustrates another framework of a logging-while-drilling risk prediction system provided by the present invention. The system includes: a data source layer for acquiring and storing data; a data processing layer for preprocessing data; a core engine layer for performing simulation calculations based on digital twins; a model calculation layer for performing calculations based on physical models and AI models; an application function layer for risk warning; and a decision and feedback layer for providing feedback.

[0106] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0107] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting logging-while-drilling risks based on digital twins, characterized in that, include: Simulation of the physical drilling system is performed using a virtual digital twin of the physical drilling system to obtain the simulation status information of the physical drilling system. The simulation state information and / or the measured state information of the physical drilling system are input into the trained risk prediction model to obtain the first risk information; Based on the simulation state information and the physical principles followed by the physical drilling system, the second risk information is obtained; The risks of the physical drilling system are determined comprehensively based on the first risk information and the second risk information.

2. The method for predicting logging-while-drilling risks based on digital twins according to claim 1, characterized in that, The first risk information includes the probability of a blowout, and the second risk information includes the ratio of bottom hole pressure to formation pressure; the step of comprehensively determining the risk of the physical drilling system based on the first risk information and the second risk information includes: The overflow risk index is calculated using the following formula: In the formula, These are the weighting coefficients. For the bottom hole pressure, For formation pressure, For overflow probability, Increase the volume of the mud pit. This represents the threshold volume of the mud pit.

3. The logging-while-drilling risk prediction method based on digital twins according to claim 1, characterized in that, The first risk information includes drill string failure types. Based on the first risk information and the second risk information, the risk of the physical drilling system is comprehensively determined, including: The cumulative damage of the drill string is calculated using the following formula: In the formula, For the drill bit at stress level The number of loops below, ( M ) for the drill bit at stress level Number of failure cycles under each drill string failure type; The risk of the physical drilling system is determined by combining the cumulative damage of the drill string and the second risk information.

4. The method for predicting logging-while-drilling risks based on digital twins according to claim 3, characterized in that, The risk prediction model includes a drill string failure type identification model; the first risk information is obtained by inputting the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model, including: Based on the simulation state information or the measured state information of the physical drilling system, a vibration spectrum image of the drill string of the physical drilling system is generated. The vibration spectrum image is input into the trained drill string fault type recognition model to obtain the drill string fault type.

5. The method for predicting logging-while-drilling risks based on digital twins according to claim 1, characterized in that, The second risk information, obtained by combining the simulation state information with the physical principles followed by the physical drilling system, includes: Based on the simulation state information and the Mohr-Coulomb criterion, the wellbore stability assessment results are obtained; Based on the simulation state information and wellbore trajectory covariance analysis, the collision risk between adjacent wells is obtained.

6. The logging-while-drilling risk prediction method based on digital twins according to claim 1, characterized in that, The simulation using a virtual digital twin of the physical drilling system includes: Construct a virtual digital twin of the physical drilling system and a multiphysics model; the digital twin includes a physical entity model of the physical drilling system and a simulation engine; The simulation engine performs simulations based on the multiphysics model and the physical entity model.

7. The method for predicting logging-while-drilling risks based on digital twins according to claim 6, characterized in that, The physical entity model includes: a geological model, a wellbore trajectory model, a drill string system model, and a fluid system model; the multiphysics model includes a wellbore multiphase flow transient model, an electromagnetic field propagation and inversion model, and a drill string dynamics model.

8. The method for predicting logging-while-drilling risks based on digital twins according to claim 1, characterized in that, The method further includes: The parameters of the virtual digital twin are updated using extended Kalman filtering.

9. A logging-while-drilling risk prediction device based on digital twins, characterized in that, include: The simulation module is used to perform simulation through a virtual digital twin of the physical drilling system to obtain the simulation status information of the physical drilling system. The first fault prediction module is used to input the simulation state information and / or the measured state information of the physical drilling system into the trained risk prediction model to obtain the first risk information. The second fault prediction module is used to obtain second risk information based on the simulation state information and the physical principles followed by the physical drilling system. The drilling risk determination module is used to comprehensively determine the risk of the physical drilling system based on the first risk information and the second risk information.

10. A logging-while-drilling risk prediction system based on digital twins, characterized in that, It includes a data acquisition device, a well site server, a cloud platform, and terminal equipment that are connected in sequence via communication. The data acquisition device is deployed at the well site to collect multi-source data from the well site. The well site server is used to preprocess the multi-source data and then send it to the cloud platform; The cloud platform is used to perform the steps in the logging-while-drilling risk prediction method based on digital twins as described in any one of claims 1 to 8, based on the preprocessed multi-source data. The terminal device is used to present the interactive interface of the cloud platform.

Images