Well killing decision real-time generation method, device and equipment, medium and product

Abstract

The embodiment of the invention provides a real-time generation method and device for a well killing decision, equipment, a medium and a product. According to the method, real-time observation data of the flow state of a shaft is obtained. And obtaining prediction state parameters based on the shaft hydraulic model and the real-time observation data. And correcting and fusing the real-time observation data and the predicted state parameters to generate a shaft state estimation result. And determining a well killing control strategy based on the shaft state estimation result and a preset safety constraint condition. According to the method, the problems that in the prior art, shaft state sensing is insufficient, decision response is lagged, and the safety risk control capability is weak are solved, and the intelligent level and safety of well killing operation are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of well control safety, and in particular to a method, apparatus, equipment, medium, and product for real-time generation of well control decisions. Background Technology

[0002] In oil drilling operations under deep, ultra-deep, and complex geological conditions, well control accidents such as overflows and blowouts occur frequently, seriously threatening operational safety and the ecological environment. Well control operations, as the core component of well control technology, aim to balance formation pressure by controlling wellhead casing pressure and standpipe pressure, thereby preventing well leakage or blowouts caused by bottomhole pressure imbalance.

[0003] However, traditional well control operations rely heavily on the experience and judgment of field engineers and manual operation. When faced with complex and ever-changing dynamic conditions downhole (such as sudden changes in formation pressure, abnormal fluid properties, equipment failures, etc.), manual decision-making suffers from problems such as delayed response, insufficient accuracy, and high safety risks. Summary of the Invention

[0004] This application provides a method, apparatus, equipment, medium, and product for real-time generation of well control decisions, in order to achieve intelligent, real-time, and safe well control operations.

[0005] In a first aspect, embodiments of this application provide a real-time generation method for well control decisions, including:

[0006] To acquire real-time observation data on the flow state in the wellbore;

[0007] Based on the wellbore hydraulic model and real-time observation data, predicted state parameters are obtained. The wellbore hydraulic model is used to generate the predicted results of the wellbore flow state.

[0008] The real-time observation data and the predicted state parameters are corrected and fused to generate wellbore state estimation results;

[0009] Based on the wellbore state estimation results and preset safety constraints, a well control strategy is determined, which is used to adjust the parameters of the well control operation in real time.

[0010] Optionally, the process of obtaining predicted state parameters based on the wellbore hydraulic model and real-time observation data includes:

[0011] Obtain basic information about the well control site, including: detailed structural parameters of the wellbore, physical data of the fluid, and process parameters of the well control process;

[0012] A theoretical model for wellbore hydraulics is constructed based on multiphase flow theory and drift flow model.

[0013] Based on the theoretical model and the basic information, the real-time observation data is discretely solved to determine the predicted state parameters of the wellbore at a preset time.

[0014] Optionally, the step of correcting and fusing the real-time observation data and the predicted state parameters to generate a wellbore state estimation result includes:

[0015] The real-time observation data and the predicted state parameters are fused based on the minimax criterion to determine the filter gain;

[0016] The predicted state parameters are corrected based on the filter gain to generate wellbore state estimation results.

[0017] Optionally, before determining the filter gain, the method further includes:

[0018] The parameter error and observation noise level of the wellbore hydraulic model are determined based on the real-time observation data. The observation noise level is used to indicate the error when the sensor measures the real-time observation data.

[0019] The preset performance index value is adjusted based on the parameter error and the observed noise level.

[0020] Optionally, the method further includes:

[0021] If the wellbore condition estimation results meet the well control strategy update conditions, a well control strategy is generated based on the wellbore condition estimation results.

[0022] If the wellbore condition estimation results do not meet the well control strategy update conditions, well control is carried out based on the historical well control strategy.

[0023] Optionally, the preset safety constraints include at least one safety constraint, and determining the well control strategy based on the wellbore state estimation results and the preset safety constraints includes:

[0024] The weighting coefficients of the safety constraints are adjusted based on the real-time observation data, and the adjustment of the weighting coefficients is based on the evaluation results of bottom hole pressure stability, equipment load status, and operational stability.

[0025] The safety constraints are weighted based on the weighting coefficients to generate a constraint function;

[0026] Based on the constraint function and the wellbore state estimation results, a well control strategy is generated, which consists of control parameters that minimize the constraint function.

[0027] Optionally, the method further includes:

[0028] The well control strategy is fed back to the well control actuator.

[0029] Obtain the execution result of the well control actuator;

[0030] The parameters of the wellbore hydraulic model are adjusted based on the execution results.

[0031] Secondly, embodiments of this application provide a real-time well control decision generation device, comprising:

[0032] The acquisition module is used to acquire real-time observation data of the flow state in the wellbore;

[0033] The determination module is used to obtain predicted state parameters based on the wellbore hydraulic model and real-time observation data. The wellbore hydraulic model is used to generate the predicted results of the wellbore flow state.

[0034] The processing module is used to correct and fuse the real-time observation data and the predicted state parameters to generate wellbore state estimation results;

[0035] The determination module is used to determine the well control strategy based on the wellbore state estimation results and preset safety constraints. The well control strategy is used to adjust the parameters of the well control operation in real time.

[0036] Optionally, the acquisition module is specifically used to acquire basic information about the well control site, including: detailed structural parameters of the wellbore, physical data of the fluid, and process parameters of the well control process.

[0037] The module is specifically used to construct the theoretical model of wellbore hydraulics based on multiphase flow theory and drift flow model;

[0038] The determination module is specifically used to perform discrete solution on the real-time observation data based on the theoretical model and the basic information to determine the predicted state parameters of the wellbore at a preset time.

[0039] Optionally, the processing module is specifically used to fuse the real-time observation data with the predicted state parameters based on the minimax criterion to determine the filtering gain;

[0040] The processing module is specifically used to correct the predicted state parameters based on the filter gain and generate wellbore state estimation results.

[0041] Optionally, the determining module is further configured to determine the parameter error and observation noise level of the wellbore hydraulic model based on the real-time observation data, wherein the observation noise level is used to indicate the error when the sensor measures the real-time observation data;

[0042] The processing module is also used to adjust the preset performance index value according to the parameter error and the observation noise level.

[0043] Optionally, the apparatus further includes: a generation module;

[0044] The generation module is used to generate a well control strategy based on the wellbore state estimation result when the wellbore state estimation result meets the well control strategy update conditions.

[0045] The processing module is also used to perform well control based on historical well control strategies when the wellbore state estimation result does not meet the well control strategy update conditions.

[0046] Optionally, the preset safety constraints include at least one safety constraint. The determining module is specifically used to adjust the weighting coefficient of the safety constraint based on the real-time observation data. The adjustment of the weighting coefficient is based on the evaluation results of bottom hole pressure stability, equipment load status, and operational stability.

[0047] The processing module is specifically used to weight the security constraints based on the weighting coefficients to generate a constraint function;

[0048] The generation module is specifically used to generate a well control strategy based on the constraint function and the wellbore state estimation result. The well control strategy consists of control parameters that minimize the constraint function.

[0049] Optionally, the device further includes: a feedback module;

[0050] The feedback module is also used to feed back the well control strategy to the well control actuator;

[0051] The acquisition module is also used to acquire the execution result of the well control actuator;

[0052] The processing module is also used to adjust the parameters of the wellbore hydraulic model based on the execution results.

[0053] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0054] The memory stores computer-executed instructions;

[0055] 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.

[0056] 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.

[0057] 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.

[0058] The real-time generation method, apparatus, equipment, medium, and product for well control decision-making provided in this application deeply integrates real-time observation data (casing pressure, standpipe pressure, mud pit increment, etc.) during the well control process with the wellbore hydraulic model. It obtains accurate wellbore flow state through robust state estimation, and on this basis, combined with well control safety constraints (bottom hole pressure additional pressure limit, wellhead equipment pressure limit, formation fracture pressure, etc.), it calculates and optimizes well control operation parameters in real time, and automatically generates well control control strategies, namely safe and efficient casing pressure and standpipe pressure control curves. This solves the problems of traditional well control decision-making relying on human experience, slow response, and high safety risks. Attached Figure Description

[0059] 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.

[0060] Figure 1 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 1 ;

[0061] Figure 2 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 2 ;

[0062] Figure 3 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 3 ;

[0063] Figure 4 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 4 ;

[0064] Figure 5 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 5 ;

[0065] Figure 6 A schematic diagram of the decision-making effect of the real-time generation method for well control decisions provided in this application. Figure 1 ;

[0066] Figure 7 Schematic diagram of the decision-making effect of the real-time generation method for kill well decision-making provided by this application Figure 2 ;

[0067] Figure 8 Schematic diagram of the decision-making effect of the real-time generation method for kill well decision-making provided by this application Figure 3 ;

[0068] Figure 9 Schematic structural diagram of the real-time generation device for kill well decision-making provided by this application;

[0069] Figure 10 Schematic structural diagram of the electronic device provided by this application. [[ID=2,0]]

[0070] Through the above-mentioned drawings, specific embodiments of this application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments. Detailed Description of the Embodiments

[0071] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. On the contrary, they are merely examples of devices and methods consistent with some aspects of this application as detailed in the appended claims.

[0072] In oil drilling operations, the complex geological conditions and dynamically changing downhole environments of deep wells and ultra-deep wells pose extremely high requirements for well control safety. During the drilling process, formation fluids (such as natural gas, crude oil, etc.) may invade the wellbore due to pressure imbalance, leading to overflow and even blowout accidents. If the overflow is not controlled in time, it may cause serious consequences such as failure of wellhead equipment, rupture of the bottom formation, or leakage of hydrogen sulfide, directly threatening the lives of operating personnel, causing huge economic losses, and damaging the ecological environment.

[0073] The kill well operation is the core link of well control technology. Its core goal is to adjust the casing pressure and standpipe pressure at the wellhead to maintain the bottom hole pressure slightly higher than the formation pore pressure to prevent the overflow from further developing, while avoiding the bottom hole pressure exceeding the formation fracture pressure and causing well leakage.

[0074] Traditional well control operations rely on the experience and judgment of field engineers and manual operation. However, the complexity of gas-liquid two-phase flow in deep wells, the uncertainty of formation pressure, and the frequency of unexpected situations (such as equipment failure and operational errors) lead to problems such as delayed response, insufficient accuracy, and high safety risks in manual decision-making. In addition, the dynamic changes in wellbore flow under complex well conditions (such as gas expansion and mud density fluctuations) further exacerbate the difficulty of well control.

[0075] Therefore, there is an urgent need for an intelligent decision-making system that can perceive the flow state of the wellbore in real time, dynamically optimize the well control strategy, and strictly meet multiple safety constraints in order to cope with the changing downhole environment in deep well control operations and ensure operational safety and efficiency.

[0076] To address the aforementioned issues, this application proposes a real-time generation method for well control decisions. Through the deep integration of data assimilation technology and robust filtering algorithms, a closed-loop decision-making method is constructed that can perceive the wellbore flow state in real time, dynamically optimize well control strategies, and strictly meet multiple safety constraints. This method uses a wellbore hydraulic model as its physical basis and robustly fuses real-time observation data with model predictions using a robust filtering algorithm. This overcomes the sensitivity of traditional Kalman filtering to model uncertainties and noise, thereby achieving accurate estimation of the wellbore state. Furthermore, this method introduces a multi-safety-constraint optimization framework, comprehensively considering objectives such as bottomhole pressure, equipment pressure bearing capacity, and operational stability, to dynamically generate optimal casing pressure and standpipe pressure control curves, solving the problems of poor dynamic adaptability, lag response, and high safety risks in well control operations.

[0077] This application is applicable to real-time decision support in well control operations of deep and ultra-deep wells, especially for complex geological conditions and sudden operating conditions (such as sudden increases in casing pressure, secondary overflows, equipment failures, etc.). The closed-loop decision system is deployed on the oil drilling platform. By collecting real-time data such as casing pressure, standpipe pressure, and mud pit increments, it dynamically optimizes well control strategies using a wellbore hydraulic model and robust filtering algorithms. Its network architecture includes a data acquisition module, a physical model module, a state estimation module, and an optimization decision module. These modules work collaboratively through a closed-loop feedback mechanism to ensure that the bottom hole pressure remains within a safe range during well control, while preventing overpressure of wellhead equipment or formation fracturing.

[0078] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0079] Figure 1 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 1 ,like Figure 1 As shown, applied to a closed-loop decision-making system, the method includes:

[0080] S101. Obtain real-time observation data of the flow state in the wellbore.

[0081] Understandably, wellbore flow state refers to the set of dynamic parameters of gas-liquid two-phase flow within the wellbore, including pressure distribution, gas cut, and gas-liquid interface position. For example, real-time observation data of casing pressure and standpipe pressure can be used to infer wellbore flow state.

[0082] Real-time observation data can be acquired through sensors and may include: oil drilling site pressure sensor data, throttle valve opening displacement sensor data, pump discharge data, mud pit increment data, equipment operating status data, etc.

[0083] After acquiring real-time observation data, the data can be preprocessed, such as through error calculation, filtering and noise reduction, and data verification, to provide reliable observation data input for subsequent decision-making.

[0084] S102. Based on the wellbore hydraulic model and real-time observation data, the predicted state parameters are obtained.

[0085] Among them, the wellbore hydraulic model is used to generate prediction results of the wellbore flow state.

[0086] Understandably, the wellbore hydraulic model is a traditional well control mathematical model. It uses the three conservation equations of mass conservation, momentum conservation, and energy conservation in multiphase flow, along with auxiliary equations such as the friction coefficient equation, drift flow model, and formation gas production model, to iteratively solve the dynamic changes of parameters such as pressure, velocity, and gas content in the drill string and annulus during the well control process using the finite difference method.

[0087] Wellbore hydraulic models can be based on multiphase flow theory and drift flow models, taking into account factors such as transient temperature field and gas-liquid phase slip relationship, to calculate state parameters such as pressure, temperature, gas-liquid two-phase volume fraction and flow velocity at each node in the wellbore, providing a physical model basis for state estimation.

[0088] S103. Correct and fuse the real-time observation data and the predicted state parameters to generate wellbore state estimation results.

[0089] Understandably, a data assimilation framework can be used when fusing data. A data assimilation framework refers to a technical system that dynamically fuses real-time observation data with physical model predictions, using algorithms to correct model biases and optimize state estimates. For example, the combination of robust filtering algorithms and wellbore hydraulic models constitutes a data assimilation framework.

[0090] Preferably, the robust filtering algorithm can be the H∞ filtering algorithm. As a robust filtering method, H∞ filtering has stronger anti-interference ability and robustness compared with traditional Kalman filtering. It can maintain good estimation performance in environments with uncertain model parameters and large external noise, and is more suitable for complex and variable drilling environments.

[0091] By utilizing real-time observation data such as casing pressure, standpipe pressure, mud pit increment, and equipment operating status collected during the well control process, the prediction results of the wellbore hydraulic model can be constrained and corrected in real time, resulting in robust wellbore state estimation results.

[0092] S104. Based on the wellbore state estimation results and preset safety constraints, determine the well control strategy.

[0093] Understandably, well control strategies refer to well control operation parameter sequences generated through optimized calculations, including control plan curves showing how casing pressure and standpipe pressure change over time. For example, adjusting pump displacement or choke valve opening based on real-time wellbore conditions to maintain safe bottomhole pressure.

[0094] By establishing a well control safety constraint framework, the optimal well control strategy can be calculated in real time after obtaining an accurate wellbore state estimate. This framework includes multiple safety constraints and optimization objectives.

[0095] The real-time generation method for well control decisions provided in this application acquires real-time observation data of the wellbore flow state. Based on the wellbore hydraulic model and the real-time observation data, predicted state parameters are obtained. The real-time observation data and predicted state parameters are corrected and fused to generate a wellbore state estimation result. Based on the wellbore state estimation result and preset safety constraints, a well control strategy is determined. This method solves the problems of insufficient wellbore state perception, delayed decision response, and weak safety risk control capabilities in existing technologies, significantly improving the intelligence level and safety of well control operations.

[0096] Figure 2 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 2 ,like Figure 2 As shown, in this embodiment... Figure 1 Based on the examples, the process of generating the predicted state parameters is described in detail. The method includes:

[0097] S201. Obtain basic information about the well control site.

[0098] The basic information includes: detailed structural parameters of the wellbore, physical data of the fluid, and process parameters of the well control process.

[0099] Understandably, obtaining basic information is fundamental to the entire well control operation and analysis at the site. Detailed structural parameters of the wellbore can include: wellbore diameter, diameter variation at different depths, depth, and inclination angle. This data can help analyze fluid flow channel characteristics, fluid flow paths, fluid velocity, and fluid flow properties.

[0100] The physical properties of fluids can include compressibility and density, for both gases and liquids. Compressibility represents the volume change of a fluid within a wellbore, while density is used to calculate flow rate.

[0101] The process parameters for well control can include: well control fluid density, well control method, etc. A reasonable well control fluid density can balance formation pressure. Different well control methods differ in operation and pressure control, collectively determining the operation process and its effectiveness.

[0102] S202. A theoretical model for constructing a wellbore hydraulic model based on multiphase flow theory and drift flow model.

[0103] Understandably, multiphase flow theory is a physical theory that studies the conservation of mass, momentum, and energy in two-phase or multiphase flows of gas and liquid. For example, in well control, the kill fluid (liquid) and the gas produced in the formation flow together within the wellbore. The drift flow model is a commonly used model in multiphase flow theory.

[0104] When constructing a wellbore hydraulic model based on multiphase flow theory and drift flow model, factors such as the wellbore's geometry, the fluid's physical properties, and flow boundary conditions can be comprehensively considered. The fundamental flow equations for the gas-liquid two phases in the wellbore, such as the continuity equation, momentum equation, and energy equation, are determined according to multiphase flow theory. Simultaneously, the velocity distribution of the gas-liquid two phases is described using the drift flow model, enabling the model to more accurately reflect the actual flow state of the gas-liquid two phases within the wellbore.

[0105] A wellbore hydraulic model is a mathematical model that describes the dynamic changes of fluids (gas and liquid) within a wellbore, including equations for mass conservation, momentum conservation, and energy conservation. This wellbore hydraulic model can describe the complex gas-liquid two-phase flow characteristics within the wellbore throughout the entire process from overflow occurrence, well shut-in, to well control completion, providing an accurate physical basis for well control decisions.

[0106] S203. Based on the theoretical model and basic information, the real-time observation data is discretely solved to determine the predicted state parameters of the wellbore at the preset time.

[0107] Understandably, after obtaining the theoretical model of the wellbore hydraulic model and the basic information of the well control site, real-time observation data can be used to perform discrete solutions to determine the predicted state parameters inside the wellbore at a preset time.

[0108] Discrete solution involves discretizing the continuous space and time of the wellbore. This can be done by dividing the wellbore into multiple small grid cells along the depth direction and dividing the well control operation time into multiple time steps. For each grid cell and each time step, a corresponding set of discrete equations is established based on the theoretical model and fundamental information.

[0109] By iteratively solving these discrete equations, predicted state parameters such as pressure, temperature, gas-liquid two-phase volume fraction, and flow velocity at each grid cell within the wellbore at a preset time can be gradually determined. These predicted state parameters can be compared and analyzed with actual observation data to promptly identify problems arising during well control.

[0110] The real-time generation method for well control decisions provided in this application acquires various basic information from the well control site, constructs a wellbore hydraulic theoretical model based on multiphase flow theory and drift flow model, and then discretizes and solves the real-time observation data based on this model and basic information to determine predicted state parameters such as pressure, temperature, gas-liquid two-phase volume fraction, and flow velocity at various points in the wellbore at preset times. This method can simulate and predict the multiphase flow state in the wellbore during the well control process, identify potential problems in advance, and provide a scientific basis for real-time monitoring and precise adjustment of well control operations.

[0111] Figure 3 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 3 ,like Figure 3 As shown, in this embodiment... Figure 1 Based on the examples, the process of generating wellbore condition estimation results is described in detail. The method includes:

[0112] S301. Based on the minimax criterion, real-time observation data and predicted state parameters are fused to determine the filter gain.

[0113] Understandably, the minimax criterion refers to the optimization principle of minimizing the estimation error energy gain in the worst case (i.e., when there is maximum modeling error and observation noise). The filter gain is a correction coefficient used to balance the reliability of model predictions and observation data.

[0114] The filter gain is determined using the minimax criterion to minimize the estimation error energy gain of the system under model uncertainty or observation noise. Real-time observation data (such as casing pressure and standpipe pressure) and physical model prediction results are input into the robust filtering algorithm. After calculating the filter gain, the model prediction results are corrected to generate a wellbore state estimate.

[0115] Specifically, the design goal of H∞ filtering is to find a filter such that the H∞ norm of the transfer function from external disturbances to estimation errors is less than a preset performance index γ. For well-controlled well control processes, let the state equation and observation equation be as follows:

[0116]

[0117]

[0118] in, Let k be the state vector containing the pressure and gas content distribution at each node of the wellbore; The model operator for H∞ filtering, also known as the wellbore hydraulic model; The control input matrix, which describes how the control inputs (pump displacement, throttle valve opening) affect the state, is determined by the wellbore hydraulic model. The control input vector, which includes operating parameters such as pump displacement, comes from field equipment; The model error is determined by a preset mean of zero and a covariance matrix of... Randomly generated from a Gaussian distribution.

[0119] It is a real-time observation vector of state parameters at time k, including additional observation noise. It originates from a real-time sensor. For observation operators, which describe the relationship between state variables and observations, they are usually relatively fixed and depend on sensor configuration and observation method; To account for observation errors, a preset mean of zero and a covariance matrix of... Randomly generated from a Gaussian distribution.

[0120] The specific recursive loop process of H∞ filtering includes:

[0121] At time k:

[0122] Inversion step (prediction): Optimal estimate based on time k-1 Estimate the current state:

[0123]

[0124] Update the inversion error covariance:

[0125] in, This is the optimal estimate that has been corrected in the previous time step. This is the inversion of the current moment when no new observation data has been received; It is the inverted covariance; It is the uncertainty inherent in the wellbore hydraulic model itself.

[0126] Once the real-time observation data is received, the filter gain can be calculated. Specifically:

[0127]

[0128] in, It measures noise; This is a characteristic feature of H∞ filtering.

[0129] After calculating the filter gain, the H∞ filter can be updated and corrected so that it can be used as the initial condition at the next time step K+1.

[0130] Status updated to:

[0131] Covariance updated to:

[0132]

[0133] in, For additional robustness terms; This is the value actually measured by the sensor; It is the error between the measured value and the inverted value; It is an identity matrix.

[0134] For preset performance indicators, By adjusting The value can control the robustness of the filter. The smaller the value, the stronger the ability to suppress uncertainty.

[0135] Optionally, the preset performance index values ​​are set during initialization and are usually fixed during well control inversion and decision-making processes, but they can also be adjusted according to the actual situation.

[0136] The parameter errors and observation noise levels of the wellbore hydraulic model are determined based on real-time observation data. Preset performance index values ​​are then adjusted based on these parameters and the observation noise level. The observation noise level indicates the error of the sensors when measuring real-time observation data.

[0137] Understandable. These are performance parameters of the H∞ filter, controlling the filter's robustness: where, The smaller the value, the stronger the ability to suppress model uncertainty and external disturbances, and the better the robustness, but the computational complexity increases. A larger value results in filter performance close to that of a Kalman filter, which is simpler to calculate but reduces robustness.

[0138] This can be set during system initialization. The initial value is usually chosen scope.

[0139] The adjustment of performance index values ​​is based on the following factors: model uncertainty: accuracy of the wellbore hydraulic model and magnitude of parameter uncertainty; observation noise level: sensor accuracy and intensity of on-site interference; actual filtering effect: state estimation error is evaluated through historical data.

[0140] Adjustment methods may include: offline parameter tuning: using historical well control data for simulation testing; trying different... Values ​​are used to evaluate the accuracy and robustness of state estimation; the value with the best overall performance is selected. value.

[0141] Online adaptive: It can dynamically adjust based on the actual estimation error. When the model deviates significantly from reality, reduce... Enhance robustness; when the system is stable, appropriately increase... Improve computational efficiency.

[0142] Empirical guideline: For deep wells and complex formations (high uncertainty), choose smaller... (e.g., 1.5-2.5); shallow wells with clear geological conditions: larger ones can be selected. (e.g., 3-5). Alternatively, expert experience guidelines can be used, pre-setting a fixed [guideline] based on the structure of different wells. value.

[0143] S302. Based on the filter gain, the predicted state parameters are corrected to generate wellbore state estimation results.

[0144] Understandably, after calculating the filter gain, the predicted state parameters can be corrected based on real-time observation data and the inversion results to ultimately determine the wellbore state estimation result. Specifically, the correction is as follows: when the accuracy of the inversion result (prior estimate) is high (…),… When the sensor observation accuracy is high (small), the filter tends to trust the inversion results; when the sensor observation accuracy is high (small), the filter tends to trust the inversion results. When the value is small, the filter is more dependent on the observations.

[0145] Optionally, if the wellbore condition estimation results meet the well control strategy update conditions, a well control strategy can be generated based on the wellbore condition estimation results.

[0146] Understandably, wellbore condition estimation results can accurately reflect the current true condition of the wellbore. When the wellbore condition estimation results meet the well control strategy update conditions, that is, when the current actual state of the wellbore has changed significantly from the initial conditions or predetermined state on which it was based, and this change requires the formulation of a new control strategy.

[0147] For example, if the casing pressure currently controlled by the engineer deviates from the planned casing pressure by more than 0.5 MPa, or if an abnormal increase in the mud pit increment is detected by more than 0.3 m...3 (cubic meters), or the on-site pump malfunctions and cannot provide sufficient pump displacement and vertical pressure.

[0148] Formulating a control strategy involves analyzing the inversion results, clarifying the changing trends and interrelationships of various parameters within the wellbore, and then, in conjunction with the goals and requirements of well control, determining the control parameters and formulating a control strategy.

[0149] If the wellbore condition estimation results do not meet the conditions for updating the well control strategy, well control is carried out based on the historical well control strategy.

[0150] Understandably, when the wellbore condition estimation results do not meet the conditions for updating the well control strategy, that is, although the actual state of the wellbore has changed, this change has not yet posed a substantial threat to the safety and effectiveness of the well control operation, or the change is small and still within the adjustable range of the original control strategy, the historical well control strategy can be used.

[0151] When implementing well control based on historical well control strategies, field operators need to closely monitor real-time changes in the wellbore status and record key parameters such as wellhead pressure and fluid flow rate. If abnormal changes in the wellbore status are detected, and these changes show a tendency to worsen or exceed the adjustable range of the historical control strategy, operators should immediately stop the current operation and generate a new well control strategy.

[0152] The real-time generation method for well control decisions provided in this application receives real-time observation data and model calculation results, and uses the H∞ filtering algorithm to invert the wellbore state, maintaining good estimation performance even under conditions of high model uncertainty and observation noise.

[0153] Figure 4 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 4 ,like Figure 4 As shown, in this embodiment... Figure 1 Based on the examples, the process of generating a well control strategy is described in detail. The method includes:

[0154] S401. Adjust the weighting coefficients of the safety constraints based on real-time observation data.

[0155] Among them, the preset security constraints include at least one security constraint.

[0156] Well control safety constraints mainly include: bottom hole pressure constraints, requiring the bottom hole pressure to be slightly greater than the formation pore pressure to prevent re-flow, while not exceeding the formation fracturing pressure to avoid well leakage; wellhead equipment pressure constraints, casing pressure and standpipe pressure must not exceed the rated pressure capacity of the wellhead equipment; pump discharge constraints, mud pump discharge must not exceed the equipment capacity limit, while ensuring sufficient annular return velocity to effectively carry cuttings; and operational stability constraints, the rate of change of pressure and flow should be controlled within a reasonable range to avoid drastic fluctuations that cause wellbore instability.

[0157] Understandably, weighting coefficients are numerical parameters used to balance the priorities of multiple safety constraints. The adjustment of these weighting coefficients is based on assessments of bottom hole pressure stability, equipment load status, and operational stability. For example, when bottom hole pressure fluctuations are significant in the initial stages of well control, the weight of the bottom hole pressure constraint is dynamically increased; conversely, when equipment load is low during the venting phase, the weight of the equipment pressure constraint is decreased. The adjusted weighting coefficients can then be used as input parameters for constraint optimization algorithms to generate control strategies.

[0158] S402. Weight the safety constraints based on the weighting coefficients to generate the constraint function.

[0159] Understandably, by weighting multiple safety constraints—that is, multiplying each safety constraint by its corresponding weight coefficient and then integrating these weighted results—a constraint function can be obtained. This constraint function is a comprehensive indicator that can fully and accurately reflect the combined impact of all safety constraints on well control operations.

[0160] S403. Based on the constraint function and wellbore state estimation results, generate a well control strategy.

[0161] Among them, the well control strategy consists of control parameters that minimize the constraint function.

[0162] Understandably, based on preset safety constraints and optimization objectives, as well as real-time on-site observation data, the constrained optimization problem can be solved, and the optimal casing pressure and standpipe pressure control plan curves can be dynamically generated to adapt to the well control process under different construction conditions and provide support for real-time well control decision-making.

[0163] Because the constraint function integrates multiple safety constraints, minimizing the constraint function, that is, determining an optimal pressure control scheme while satisfying all safety constraints, can maximize the safety and smooth progress of well control operations.

[0164] In practice, on-site personnel can use these control parameters to precisely control the operation of the well control equipment, adjust the stand pressure and casing pressure, and ensure that the well control operation is carried out according to the preset planned curve.

[0165] The optimization objective is:

[0166]

[0167] The constraints are:

[0168]

[0169] Where J is the comprehensive constraint function, Total well-killing time; and These represent the changes in throttle valve pressure and pump displacement, respectively. , , These are the weighting coefficients. , These are formation pore pressure and fracture pressure, respectively. Adding pressure to safety; , , These represent the bottom hole pressure, casing pressure, and riser pressure at time K, respectively.

[0170] The real-time generation method for well control decisions provided in this application embodiment is based on the wellbore state inversion results obtained by H∞ filtering, considers multiple safety constraints (bottom hole pressure constraints, equipment pressure constraints, operational stability constraints, etc.), solves the optimal control strategy, and generates casing pressure and standpipe pressure control plan curves in real time, thereby adapting to the well control process under different construction conditions and providing support for real-time well control decisions.

[0171] Figure 5 A flowchart illustrating the real-time generation method for well control decisions provided in this application. Figure 5 ,like Figure 5 As shown, in this embodiment... Figure 1 Based on the examples, the closed-loop decision-making process of model updating and policy is described in detail. The method includes:

[0172] S501, Feedback the well control strategy to the well control actuator.

[0173] Understandably, the well control actuator is the operational unit for well control operations, which includes various key equipment such as the well control pump, choke valve, and blowout preventer. These devices work together to effectively control wellbore pressure by controlling the injection rate and pressure of the well control fluid, as well as adjusting the wellhead pressure.

[0174] Feedback of well control strategies to well control actuators can be accomplished using a communication system. This system needs to have the capability to transmit data at high speed to ensure the accurate transmission of control strategies, while also guaranteeing the reliability and integrity of the transmission.

[0175] S502. Obtain the execution results of the well control actuator.

[0176] Understandably, after the well control actuator operates according to the feedback control strategy, relevant data on its execution status can be collected in real time and accurately. This data can be obtained from the well control actuator and from sensors and monitoring equipment at the well control site.

[0177] S503. Adjust the parameters of the wellbore hydraulic model based on the execution results.

[0178] Understandably, once the execution results of the well control actuator are obtained, these actually measured data can be compared and analyzed with the prediction results of the wellbore hydraulic model. This comparison can determine the differences between the model predictions and the actual execution, and identify which parameters have a greater impact on these differences.

[0179] Based on the results of the comparative analysis, the parameters of the wellbore hydraulic model can be adjusted. These parameters can be adjusted using the least squares method or the Bayesian update algorithm. The adjustment must ensure that the adjusted model more accurately reflects the actual flow state of the wellbore, while avoiding over-adjustment that could cause the model to lose its universality.

[0180] For example, if it is found that the density of the kill fluid has a significant impact on the wellbore pressure distribution, and the actual kill fluid density deviates from the value set in the model, then the kill fluid density parameter in the model can be adjusted appropriately to improve the model's prediction accuracy.

[0181] By continuously adjusting the model parameters based on the execution results, the wellbore hydraulic model can be made closer to the actual situation, providing more reliable predictions and guidance for subsequent well control operations.

[0182] Simultaneously, the well control strategy can be adjusted based on the actual measurement results. For example, when a deviation between the actual casing pressure and the target value is detected, the closed-loop feedback mechanism can adjust model parameters (such as slip velocity) and control strategies (such as pump displacement) to form a continuously optimized closed-loop control loop.

[0183] The real-time well control decision generation method provided in this application accurately feeds back the well control strategy to the well control actuator, acquires the execution results with the help of sensors, and then precisely adjusts the key parameters in the wellbore hydraulic model based on these results. This process can form a closed-loop control, enabling real-time and precise control of the well control process, making the well control operation more closely match the actual well conditions, effectively improving the well control success rate, reducing the risk of accidents such as well blowouts and lost circulation, ensuring the safety of personnel and equipment, and improving the overall efficiency of drilling operations.

[0184] In some embodiments, the real-time generation method for well control decisions of this application is applied to a closed-loop decision-making system, and the decision-making process may specifically include:

[0185] First, initialize the parameters of the closed-loop decision system, including setting the wellbore geometry parameters, fluid property parameters, H∞ filter parameters (performance index γ, initial covariance matrix, etc.), constraints, and optimization objectives.

[0186] The system then enters a decision-making loop. At each time step, the closed-loop decision-making system first receives real-time observation data, including casing pressure, standpipe pressure, and mud pit increments. Then, the closed-loop decision-making system calls the wellbore hydraulic model and, based on the current state parameters and well control process parameters, performs preliminary state predictions, calculating the predicted state parameters and prediction error covariance for the next time step.

[0187] Next, the closed-loop decision system uses the H∞ filtering algorithm to calculate the H∞ gain, uses real-time observation data to perform robust correction on the predicted state, updates the state estimate and error covariance, and obtains a more accurate wellbore flow state estimate, that is, inverts a more accurate wellbore pressure and gas-liquid distribution.

[0188] Based on the inversion results, the closed-loop decision system calls the constraint optimization algorithm, considering multiple constraints such as bottom hole pressure constraint, wellhead equipment pressure constraint, formation fracture pressure constraint, pump discharge constraint, and operation stability constraint. With the shortest well control time and the most stable operation as the optimization objectives, the optimal casing pressure and standpipe pressure control plan curves are solved.

[0189] When abnormal operating conditions are detected (such as sudden increases or decreases in casing pressure), the closed-loop decision-making system can respond quickly and replan the well control curve. Through this iterative and rolling optimization, the closed-loop decision-making system can continuously monitor changes in wellbore status, dynamically optimize control strategies, and achieve real-time intelligent decision-making for well control operations.

[0190] The following specific example will be used to illustrate and verify the effectiveness of the well control decision in this application.

[0191] For example, in a case study of an offshore oilfield well, the initial shut-in casing pressure was approximately 7.5 MPa, and the shut-in standby pressure was approximately 1.0 MPa. During well control, the pump flow rate was set to 42 L / s, and the density of the kill fluid was 0.05-0.10 g / cm³ higher than that of the drilling fluid. Well control safety constraints included: an additional safety factor of 0.3-0.5 MPa for bottomhole pressure, and a pressure limit of 35 MPa for wellhead equipment.

[0192] Figure 6 A schematic diagram of the decision-making effect of the real-time generation method for well control decisions provided in this application. Figure 1 ,like Figure 6 As shown, Figure 6 It demonstrates the effectiveness of real-time decision-making in dealing with sudden increases in casing pressure during well control. Figure 6The left side shows the initial casing pressure control curve (solid line) and the adjusted casing pressure control plan curve (dashed line), while the right side shows the change in well gas content with well depth at the corresponding time.

[0193] In the initial stage of well control (t=57.5min), when the engineer was carrying out the well control plan, a sudden increase in casing pressure occurred due to operational errors or other circumstances. The closed-loop decision-making system of this application can quickly invert the current wellbore flow state based on real-time collected observation data such as casing pressure and standpipe pressure, using H∞ filtering. This includes key parameters such as pressure distribution, gas cut distribution, and gas-liquid interface position at various locations within the wellbore. Based on the accurate wellbore state obtained from the inversion, the closed-loop decision-making system can immediately initiate a constrained optimization decision-making process, adjusting the subsequent well control design curve in real time while ensuring that the bottomhole pressure does not exceed the formation fracture pressure.

[0194] like Figure 6 As shown in the left and middle figures, the adjusted casing pressure control curve (dashed line in the middle figure) is generally higher than the original plan (dashed line in the left figure). This is because after a sudden increase in casing pressure, the system needs to re-plan the control strategy based on the new pressure. Since the adjusted casing pressure is higher than the original plan, the expansion of gas in the wellbore during its ascent is reduced, resulting in a significant reduction in the casing pressure change caused by the gas reaching the wellhead. The pressure decreases from the initial design ΔP=1.0MPa to ΔP=0.6MPa, a reduction of 40%.

[0195] This adjustment effectively avoids the impact of excessively high casing pressure peaks on wellhead equipment, ensuring the safety of well control operations. As can be seen from the wellbore gas content distribution map, the closed-loop decision-making system accurately inverts the location and concentration distribution of gas in the wellbore, providing a reliable basis for adjusting the control strategy.

[0196] Figure 7 A schematic diagram of the decision-making effect of the real-time generation method for well control decisions provided in this application. Figure 2 ,like Figure 7 As shown, Figure 7 It demonstrates the effectiveness of real-time decision-making in dealing with sudden drops in casing pressure and secondary overflow during well control operations. Figure 7 The left side shows a comparison of the curves before and after the sudden drop in casing pressure control, while the right side shows the dynamic change of well gas content with well depth and time.

[0197] During well control (t=91.5min), an engineer's operational error caused a sudden drop in casing pressure, resulting in excessively low bottomhole pressure below the formation pore pressure, triggering a secondary bottomhole overflow. From Figure 7 As can be seen, a violent pressure fluctuation occurred immediately after the sudden drop in casing pressure, which is a clear characteristic of the onset of secondary overflow. At this point, formation fluids re-entered the wellbore, adding a new gas section within the wellbore.

[0198] The closed-loop decision-making system of this application can quickly identify abnormal operating conditions. Utilizing real-time casing pressure observations and mud pit increment changes, it robustly estimates the current wellbore flow state using the H∞ filtering algorithm. Based on the inverted wellbore state, including information such as the additional gas intrusion caused by secondary overflow, the location and concentration of the newly added gas, the system re-plans the subsequent well control design curve in real time.

[0199] The adjusted well control curve fully considers the impact of secondary overflow, ensuring that the bottom hole pressure remains higher than the formation pore pressure throughout the entire venting process, preventing the overflow from continuing to develop. The wellbore gas cut distribution map clearly shows a new high gas cut zone at the bottom of the wellbore, direct evidence of secondary overflow. By dynamically adjusting the casing pressure control strategy, the system successfully prevented the overflow from continuing, keeping the risk of secondary overflow within an acceptable range and avoiding a more serious well control accident.

[0200] Figure 8 A schematic diagram of the decision-making effect of the real-time generation method for well control decisions provided in this application. Figure 3 ,like Figure 8 As shown, Figure 8 The real-time decision-making effect after casing pressure control returned to normal during well control is shown. After experiencing abnormal conditions such as sudden increases and decreases in casing pressure, casing pressure control gradually returned to normal (t=143.5min), the bottom hole pressure was once again greater than the formation pore pressure, and the secondary overflow stopped.

[0201] At this point, the flow state within the wellbore differs significantly from the initial design conditions: due to the secondary overflow, at least two gas segments exist within the wellbore—the first being the initial overflow gas, and the second being the gas intruded by the secondary overflow. The system needs to continuously monitor the migration and discharge processes of these gas segments within the wellbore and dynamically optimize the control strategy.

[0202] Depend on Figure 8 As can be seen from the initial well control design curve (dashed line in the left figure), due to the secondary overflow, when the first gas reaches the wellhead, the peak casing pressure increases from the originally planned 9.0 MPa to approximately 9.5 MPa, an increase of 5.6%. The closed-loop decision system of this application can accurately predict this change and adjust the casing pressure control strategy in advance to ensure that the peak casing pressure does not exceed the pressure limit of the wellhead equipment.

[0203] Subsequently, based on the real-time wellbore status data obtained through H∞ filtering inversion, including information such as the location, volume, and migration velocity of the second gas stage, the system dynamically adjusts the casing pressure control strategy during the discharge of the second gas stage. The wellbore gas cut distribution map clearly shows two high gas cut peak regions, corresponding to the two gas stages respectively. The system uses model predictive control to optimize the future casing pressure control sequence in the prediction time domain, ensuring that subsequent gas stages can be safely and smoothly discharged from the wellbore, avoiding drastic fluctuations in casing pressure.

[0204] Throughout the process, the system consistently meets multiple safety constraints, including bottom hole pressure constraints (always slightly greater than formation pore pressure, but not exceeding fracture pressure), wellhead equipment pressure constraints, and operational stability constraints.

[0205] Figure 9 A schematic diagram of the structure of the real-time generation device for well control decisions provided in this application is shown below. Figure 9 As shown, the real-time well control decision generation device 600 provided in this embodiment includes:

[0206] The acquisition module 601 is used to acquire real-time observation data of the flow state in the wellbore;

[0207] The determination module 602 is used to obtain predicted state parameters based on the wellbore hydraulic model and real-time observation data. The wellbore hydraulic model is used to generate the prediction results of the wellbore flow state.

[0208] Processing module 603 is used to correct and fuse real-time observation data and predicted state parameters to generate wellbore state estimation results;

[0209] The determination module 602 is used to determine the well control strategy based on the wellbore state estimation results and preset safety constraints. The well control strategy is used to adjust the parameters of the well control operation in real time.

[0210] Optionally, module 601 is used to acquire basic information about the well control site, including: detailed structural parameters of the wellbore, physical data of the fluid, and process parameters of the well control process.

[0211] Module 602 is specifically used to construct a theoretical model for wellbore hydraulics based on multiphase flow theory and drift flow model;

[0212] The determination module 602 is specifically used to perform discrete solution on real-time observation data based on theoretical models and basic information to determine the predicted state parameters of the wellbore at a preset time.

[0213] Optionally, the processing module 603 is specifically used to fuse real-time observation data with predicted state parameters based on the minimax criterion to determine the filtering gain;

[0214] The processing module 603 is specifically used to correct the predicted state parameters based on the filter gain and generate wellbore state estimation results.

[0215] Optionally, the determination module 602 is also used to determine the parameter error of the wellbore hydraulic model and the observation noise level based on the real-time observation data. The observation noise level is used to indicate the error when the sensor measures the real-time observation data.

[0216] The processing module 603 is also used to adjust the preset performance index value according to the parameter error and the observation noise level.

[0217] Optionally, the apparatus may also include: a generation module 604;

[0218] The generation module 604 is used to generate a well control strategy based on the wellbore state estimation results, provided that the wellbore state estimation results meet the well control strategy update conditions.

[0219] The processing module 603 is also used to perform well control based on historical well control strategies when the wellbore state estimation results do not meet the well control strategy update conditions.

[0220] Optionally, the preset safety constraints include at least one safety constraint. The determining module 602 is specifically used to adjust the weight coefficient of the safety constraint based on real-time observation data. The adjustment of the weight coefficient is based on the evaluation results of bottom hole pressure stability, equipment load status, and operational stability.

[0221] Processing module 603 is specifically used to perform weighted processing on safety constraints based on weight coefficients to generate constraint functions;

[0222] The generation module 604 is specifically used to generate a well control strategy based on the constraint function and the wellbore state estimation results. The well control strategy consists of control parameters that minimize the constraint function.

[0223] Optionally, the device may also include: a feedback module 605;

[0224] The feedback module 605 is also used to feed back the well control strategy to the well control actuator;

[0225] The acquisition module 601 is also used to acquire the execution results of the well control actuator;

[0226] The processing module 603 is also used to adjust the parameters of the wellbore hydraulic model based on the execution results.

[0227] The real-time well control decision generation 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.

[0228] Figure 10 A schematic diagram of the structure of the electronic device provided in this application. Figure 10 As shown, the electronic device 700 provided in this embodiment includes at least one processor 701 and a memory 702. Optionally, the device 700 further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0229] In a specific implementation, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to perform the above-described method.

[0230] 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.

[0231] 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.

[0232] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0233] 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.

[0234] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0235] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0236] 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.

[0237] 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.

[0238] 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.

[0239] 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.

[0240] 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.

[0241] 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.

[0242] 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.

[0243] 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 method for real-time generation of well kill decisions, characterized by, The method comprises the following steps: acquiring real-time observation data of a wellbore flow state; obtaining predicted state parameters based on a wellbore hydraulics model and the real-time observation data, wherein the wellbore hydraulics model is used to generate a prediction result of the wellbore flow state; correctly fusing the real-time observation data and the predicted state parameters to generate a wellbore state estimation result; determining a well killing control strategy based on the wellbore state estimation result and a preset safety constraint condition, wherein the well killing control strategy is used to adjust parameters of well killing operation in real time.

2. The method of claim 1, wherein, The method for obtaining the predicted state parameters based on the wellbore hydraulics model and the real-time observation data comprises the following steps: acquiring basic information of a well killing site, wherein the basic information comprises detailed structural parameters of a wellbore, physical data of a fluid, and process parameters of a well killing process; constructing a theoretical model of the wellbore hydraulics model based on a multiphase flow theory and a drift flow model; discretely solving the real-time observation data based on the theoretical model and the basic information to determine predicted state parameters of the wellbore at a preset time.

3. The method of claim 1, wherein, The method for correctly fusing the real-time observation data and the predicted state parameters to generate the wellbore state estimation result comprises the following steps: fusing the real-time observation data and the predicted state parameters based on a minimax criterion to determine a filter gain; correcting the predicted state parameters based on the filter gain to generate the wellbore state estimation result.

4. The method of claim 3, wherein, Before determining the filter gain, the method further comprises the following steps: determining a parameter error of the wellbore hydraulics model and an observation noise level based on the real-time observation data, wherein the observation noise level is used to indicate an error when a sensor measures the real-time observation data; adjusting a preset performance index value based on the parameter error and the observation noise level.

5. The method of claim 3, wherein, The method further comprises the following steps: generating a well killing control strategy based on the wellbore state estimation result in a case where the wellbore state estimation result meets a well killing control strategy updating condition; controlling well killing based on a historical well killing control strategy in a case where the wellbore state estimation result does not meet the well killing control strategy updating condition.

6. The method of claim 5, wherein, The preset safety constraint condition comprises at least one safety constraint condition, and the method for determining the well killing control strategy based on the wellbore state estimation result and the preset safety constraint condition comprises the following steps: adjusting a weight coefficient of the safety constraint condition based on the real-time observation data, wherein the adjustment of the weight coefficient is based on an evaluation result of bottom hole pressure stability, equipment load state, and operation smoothness; performing weighted processing on the safety constraint condition based on the weight coefficient to generate a constraint function; generating the well killing control strategy based on the constraint function and the wellbore state estimation result, wherein the well killing control strategy is composed of control parameters that minimize the constraint function.

7. The method of claim 5, wherein, The method further comprises the following steps: feeding back the well killing control strategy to a well killing execution mechanism; acquiring an execution result of the well killing execution mechanism; adjusting parameters of the wellbore hydraulics model based on the execution result.

8. A device for real-time generation of well control decisions, characterized in that The method comprises the following steps: an acquiring module, configured to acquire real-time observation data of a wellbore flow state; a determining module, configured to obtain predicted state parameters based on a wellbore hydraulics model and the real-time observation data, wherein the wellbore hydraulics model is used to generate a prediction result of the wellbore flow state; a processing module, configured to correct and fuse the real-time observation data and the predicted state parameters to generate a wellbore state estimation result; a determining module, configured to determine a kill control strategy based on the wellbore state estimation result and a preset safety constraint condition, the kill control strategy being used to adjust parameters of a kill operation.

9. An electronic device, comprising: comprising: a memory, a processor; the memory stores computer-executed instructions; the processor executes the computer-executed instructions stored in the memory, so that the processor executes the method in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, the computer-readable storage medium stores computer-executed instructions, and the computer-executed instructions are executed by the processor to implement the method in any one of claims 1-7.

Images