A high-precision drilling rotary steerable trajectory intelligent control method and device

Abstract

The embodiment of the application provides a high-precision drilling rotary steering trajectory intelligent control method and device, and belongs to the technical field of drilling control. The method comprises the following steps: constructing an observation sequence for representing a drilling state based on multi-source observation data; performing state estimation based on the observation sequence to generate a state estimation result containing a wellbore attitude parameter and a dynamic index; constructing a constraint optimization model based on the state estimation result and a preset trajectory target, and performing model predictive control in the constraint optimization model to generate a control amount sequence for driving a rotary steering tool; and applying the control amount sequence to the rotary steering tool to perform trajectory adjustment, and performing residual error calculation and model parameter updating based on execution feedback information and corresponding state estimation results in real time until drilling is completed. The application scheme realizes high-precision closed-loop intelligent control of a drilling trajectory, significantly reduces trajectory deviation and improves drilling stability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of drilling control, in particular to a high-precision drilling rotary steerable trajectory intelligent control method and a high-precision drilling rotary steerable trajectory intelligent control device. BACKGROUND

[0002] Rotary steerable drilling technology is a key means for current directional well and horizontal well construction. Its basic principle is to continuously adjust the bit attitude by using downhole rotary steerable tools during drilling to realize the extension of the wellbore trajectory according to the preset target. Compared with the traditional build-up tool, the rotary steerable technology can continuously correct the trajectory during drilling, significantly improving the construction efficiency. However, the existing technology still has outstanding problems in high-precision trajectory control.

[0003] The existing rotary steerable trajectory control method usually relies on a distributed measurement system on the ground and downhole, calculates the trajectory deviation by measuring the wellbore attitude and feeding back a small amount of operating parameters, and then generates a steering instruction by a control program. However, under complex formation conditions, the downhole environment presents obvious nonlinear characteristics, such as friction effect between the drill string and the well wall, stick-slip and impact vibration, local formation heterogeneity, etc. These factors will cause fluctuations and even distortions in wellbore attitude data and operating parameters, and the existing method mostly uses linear approximation model or empirical formula for processing, which is difficult to fully reflect the actual trajectory state.

[0004] Due to the error between state estimation and the real wellbore trajectory, the control instruction often does not match the actual drilling effect after being executed, causing the trajectory to gradually deviate from the preset target, and relying on manual monitoring and frequent correction to maintain the construction requirements. This deviation accumulation not only reduces the precision of trajectory control, but also brings the risk of wellbore collision and uneven well pattern deployment, increasing the construction cycle and cost.

[0005] Therefore, the core deficiency of the existing rotary steerable trajectory control method is the lack of accurate modeling and estimation of complex downhole nonlinear characteristics, which cannot achieve high-precision determination of the wellbore trajectory state, thereby directly limiting the improvement of trajectory control precision. This problem has become a bottleneck for the further development of rotary steerable drilling technology. SUMMARY

[0006] The purpose of the embodiments of the present application is to provide a high-precision drilling rotary steerable trajectory intelligent control method and device to at least solve the problem of insufficient trajectory state estimation precision in the existing rotary steerable drilling process, which leads to trajectory control deviation accumulation.

[0007] In order to achieve the above object, the present application provides a high-precision drilling rotary steerable trajectory intelligent control method, which comprises the following steps: acquiring multi-source observation data including wellbore attitude data and bottomhole assembly operation data in a target drilling process, and constructing an observation sequence for representing a drilling state based on the multi-source observation data; performing state estimation based on the observation sequence to generate a state estimation result containing wellbore attitude parameters and dynamic indicators; constructing a constraint optimization model based on the state estimation result and a preset trajectory target, and performing model predictive control in the constraint optimization model to generate a control quantity sequence for driving a rotary steerable tool; and applying the control quantity sequence to the rotary steerable tool to perform trajectory adjustment, and performing residual error calculation and model parameter updating based on execution feedback information and corresponding state estimation results in real time until the drilling is completed.

[0008] Optionally, the wellbore attitude data includes a hole inclination angle, an azimuth angle and a spatial curvature parameter; the bottomhole assembly operation data includes a drilling pressure, a rotation speed and a torque; the step of constructing an observation sequence for representing a drilling state based on the multi-source observation data comprises the following steps: uniformly aligning time stamps of the multi-source observation data, and performing interpolation resampling based on the aligned data to obtain a synchronous sequence; performing outlier rejection and filtering processing on the synchronous sequence, and obtaining a corresponding attitude denoising sequence and an operation denoising sequence based on the processed wellbore attitude data and the processed bottomhole assembly operation data respectively; extracting a change rate of the hole inclination angle, the azimuth angle and the spatial curvature based on the attitude denoising sequence as a first feature; extracting the rotation speed, the torque and a mechanical power proxy based on the operation denoising sequence as a second feature; and connecting the first feature and the second feature in time sequence after normalization to form the observation sequence for representing the drilling state.

[0009] Optionally, the step of performing state estimation based on the observation sequence to generate a state estimation result containing wellbore attitude parameters and dynamic indicators comprises the following steps: inputting the observation sequence into a nonlinear state estimation model, extracting a sequence component related to a wellbore attitude and solving a hole inclination angle, an azimuth angle and a spatial curvature to obtain wellbore attitude parameters; simultaneously extracting a sequence component related to a bottomhole assembly operation and performing frequency domain analysis to identify a stick-slip frequency band and an impact vibration frequency band and calculate corresponding strength values to form dynamic indicators; and storing the wellbore attitude parameters and the dynamic indicators in combination as the state estimation result.

[0010] Optionally, the nonlinear state estimation model is constructed by establishing nonlinear observation equations and state transition equations between the observation sequence and state variables, and introducing an unscented Kalman filter algorithm or an extended Kalman filter algorithm into an equation solving framework to obtain the nonlinear state estimation model, which is used to output the state estimation result containing the wellbore attitude parameters and the dynamic indicators in the state estimation process.

[0011] Optionally, sequence components related to the bottomhole assembly operation are extracted and frequency domain analysis is performed to identify stick-slip frequency bands and impact vibration frequency bands and calculate corresponding strength values to form dynamic indicators, including: extracting sequence components related to the bottomhole assembly operation from the observation sequence; performing fast Fourier transform based on the extracted sequence components to obtain a frequency spectrum distribution; identifying stick-slip frequency bands and impact vibration frequency bands in the frequency spectrum distribution and calculating amplitude and energy strength within the identified frequency band range to form dynamic indicators for state estimation results.

[0012] Optionally, a constraint optimization model is constructed based on the state estimation results and a preset trajectory target, including: comparing the wellbore attitude parameters and dynamic indicators in the state estimation results with the preset trajectory target to determine a current trajectory deviation amount and a stability constraint amount; establishing a nonlinear constraint set including a spatial curvature constraint, a build-up rate constraint and an anti-collision constraint based on the trajectory deviation amount and the stability constraint amount; converting the trajectory target into an objective function and combining the objective function with the nonlinear constraint set to form an optimization solving structure; calling a model predictive control algorithm in the rolling time domain in the optimization solving structure to obtain a constraint optimization model.

[0013] Optionally, model predictive control is performed in the constraint optimization model to generate a control amount sequence for driving the rotary steerable tool, including: predicting a future wellbore attitude change trend based on the state estimation results in the rolling time domain; performing deviation calculation on the future wellbore attitude change trend and the preset trajectory target to obtain a trajectory deviation amount for optimization solving; iteratively solving the objective function in the constraint optimization model under the trajectory deviation amount constraint to obtain a control variable combination in a plurality of candidate solution spaces; selecting a solution that simultaneously satisfies the spatial curvature constraint, the build-up rate constraint and the anti-collision constraint from the candidate solution space, and outputting a control amount sequence composed of a deflector angle, a phase parameter and a drilling pressure and rotation speed setting value.

[0014] Optionally, the control amount sequence is applied to the rotary steerable tool to perform trajectory adjustment, and residual error calculation and model parameter updating are performed in real time based on execution feedback information and corresponding state estimation results, including: issuing the deflector angle and the phase parameter in the control amount sequence to the deflection driving assembly of the rotary steerable tool, and simultaneously inputting the drilling pressure and rotation speed setting value to the ground control system to adjust the drilling pressure execution device and the rotation driving device; after the control amount sequence is applied, wellbore attitude data and operation data returned by the rotary steerable tool are collected in real time and compared with the state estimation results to calculate residual errors; when the residual error is below a preset residual error threshold, the original control period is maintained, and when the residual error exceeds the preset residual error threshold, model parameter updating and control period adjustment are triggered, and a new control amount sequence is generated to act on the rotary steerable tool again.

[0015] Optionally, when the residual exceeds the preset residual threshold, triggering the model parameter update and the control period adjustment, and generating a new control quantity sequence to act on the rotary steering tool again, comprising: comparing the residual with the preset residual threshold, if the residual exceeds the preset residual threshold, calling the proxy model for constrained optimization and correcting the friction coefficient and the formation stiffness parameter in the proxy model based on the residual change to form the updated model parameter; adjusting the control period according to the amplitude of the residual exceeding the preset residual threshold, the greater the exceeding amplitude, the shorter the control period; re-executing the model predictive control under the updated model parameter and the new control period configuration to generate a new control quantity sequence and act on the rotary steering tool again.

[0016] The second aspect of the present application provides a high-precision drilling rotary steering trajectory intelligent control device, the device comprising: an acquisition unit for acquiring multi-source observation data including wellbore attitude data and bottomhole assembly operation data in a target drilling process, and constructing an observation sequence for representing the drilling state based on the multi-source observation data; an estimation unit for performing state estimation based on the observation sequence to generate a state estimation result containing wellbore attitude parameters and dynamic indicators; a scheme generation unit for constructing a constrained optimization model based on the state estimation result and a preset trajectory target, and performing model predictive control in the constrained optimization model to generate a control quantity sequence for driving the rotary steering tool; an execution unit for applying the control quantity sequence to the rotary steering tool to perform trajectory adjustment, and performing residual calculation and model parameter update based on execution feedback information and corresponding state estimation results in real time until the drilling is completed.

[0017] Through the above technical solution, the present application scheme can form a more comprehensive description of the downhole state by collecting wellbore attitude data and bottomhole assembly operation data during drilling and fusing the multi-source observation data. On this basis, the state estimation result containing attitude parameters and dynamic indicators is obtained through nonlinear state estimation, so that the trajectory calculation is closer to the real wellbore state. The constrained optimization model is established in combination with the preset trajectory target, and the control quantity sequence is solved under the model predictive control framework, which can generate reasonable steering instructions while considering the spatial curvature, build-up rate and stability conditions. Further, by applying the control quantity to the rotary steering tool and using the feedback information to determine the residual and update the parameters in real time, a closed-loop correction mechanism can be formed to continuously correct the trajectory deviation. Overall, this method can effectively improve the control precision and stability of the rotary steering drilling trajectory, reduce the accumulation of trajectory deviation, and improve the construction efficiency and wellbore quality.

[0018] Other features and advantages of the present application will be described in detail in the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0019] The accompanying drawings are included to provide a further understanding of the embodiments of the application, and constitute a part of the specification, and are used to explain the embodiments of the application together with the specific embodiments described below in the specific embodiments section of the specification, but are not used to limit the embodiments of the application. In the drawings:

[0020] Figure 1 is a step flow chart of the high-precision drilling rotary steerable trajectory intelligent control method provided by an embodiment of the application;

[0021] Figure 2 is a device structure diagram of the high-precision drilling rotary steerable trajectory intelligent control device provided by an embodiment of the application. DETAILED DESCRIPTION

[0022] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. It should be understood that the specific embodiments described herein are only used to explain and illustrate the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present application.

[0023] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative positional relationship, movement condition, etc. between components in a certain specific posture (as shown in the drawings), and if the specific posture changes, the directional indications also change accordingly.

[0024] In addition, if the embodiments of the present application involve descriptions such as “first”, “second”, etc., the descriptions of “first”, “second”, etc. are only for description purposes, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features limited by “first” and “second” can explicitly or implicitly include at least one of the features. In addition, the technical solutions of the various embodiments can be combined with each other, but it must be based on the fact that a person of ordinary skill in the art can realize it, and when the combination of technical solutions contradicts each other or cannot be realized, it should be considered that the combination of technical solutions does not exist, and is also not within the scope of protection claimed by the present application.

[0025] As shown in FIG. 1, the embodiments of the present application provide a high-precision drilling rotary steerable trajectory intelligent control method, which comprises the following steps: Figure 1

[0026] ​Step S1: acquiring multi-source observation data including wellbore attitude data and bottomhole assembly operation data in a target drilling process, and constructing an observation sequence for representing a drilling state based on the multi-source observation data.

[0027] Specifically, the wellbore attitude data includes inclination, azimuth and spatial curvature parameters; the bottomhole assembly operation data includes weight on bit, rotary speed and torque; the observation sequence for representing the drilling state is constructed based on the multi-source observation data, including: uniformly aligning time stamps of the multi-source observation data, and performing interpolation resampling based on the aligned data to obtain a synchronous sequence; performing outlier rejection and filtering processing on the synchronous sequence, and obtaining corresponding attitude denoising sequence and operation denoising sequence based on the processed wellbore attitude data and bottomhole assembly operation data respectively; extracting rates of change of inclination, azimuth and spatial curvature based on the attitude denoising sequence as first features; extracting rotary speed, torque and mechanical power proxy based on the operation denoising sequence as second features; connecting the first features and the second features after normalization in chronological order to form the observation sequence for representing the drilling state.

[0028] In the specific implementation process, first, multi-source observation data needs to be continuously collected in the target drilling construction process to completely depict the downhole trajectory and operation state. The multi-source observation data mainly includes two categories: one is wellbore attitude data, and the other is bottomhole assembly operation data. The wellbore attitude data is usually provided by the measurement-while-drilling unit in real time, and its key indicators include inclination, azimuth and spatial curvature parameters calculated by combining inclination and azimuth. These parameters directly reflect the spatial position and trajectory bending degree of the wellbore, and are basic information for trajectory accuracy control. The bottomhole assembly operation data is mainly provided by the measurement elements installed on each node of the bottomhole assembly, including weight on bit, rotary speed and torque, which are used to represent the force state of the drill bit, the rotation speed and the torsion of the drill string, and are important basis for reflecting mechanical loading characteristics.

[0029] After the above two types of data are acquired, the data of different sources need to be processed and unified. Since the wellbore attitude data and the bottomhole assembly operation data come from different acquisition channels, their sampling frequencies and time stamps often differ, and if they are directly spliced and used, it will cause time sequence misalignment and feature distortion. Therefore, before data fusion, the time stamps of the multi-source data need to be uniformly aligned. The specific method is to perform interpolation and resampling operations on various data based on a unified time reference, so that they are mapped to the same time step, thereby generating a synchronous sequence. After such processing, the attitude data and the operation data at each time can be corresponded one by one, avoiding trajectory estimation errors caused by different sampling.

[0030] After obtaining the synchronization sequence, further improvement of data quality is needed. The drilling site environment is complex, and sensor signals are easily affected by factors such as high temperature and high pressure downhole, drill string vibration, and electromagnetic interference. The original data often contains noise or outliers. If not processed directly, it will significantly reduce the reliability of subsequent trajectory estimation and optimization models. Therefore, outlier rejection and filtering processing need to be performed on the synchronization sequence. Outlier rejection can be based on threshold rules or statistical distribution judgment to remove abnormal points far from the normal range; filtering processing can use Kalman filtering or low-pass filtering methods to weaken high-frequency noise. After this step, smoother sequence data is obtained, and it can be further divided into attitude denoising sequence and running denoising sequence. The former mainly retains the change law of inclination angle, azimuth angle, and spatial curvature, and the latter corresponds to the smooth trend of weight on bit, rotary speed, and torque.

[0031] After obtaining the denoised attitude and running sequences, further feature extraction is needed to make the data reflect the dynamic changes of the wellbore state. For the attitude denoising sequence, first calculate the change rates of inclination angle, azimuth angle, and spatial curvature. These three quantities quantify the speed and direction of the change of the wellbore space attitude with time, and can reflect the deviation trend of the wellbore trajectory. Inputting these change rates as the first feature can enhance the sensitivity of the state estimation model to trajectory changes. For the running denoising sequence, calculate the change rates of rotary speed and torque, as well as the mechanical power proxy quantity formed by the combination of rotary speed and torque. The mechanical power proxy quantity can reflect the power consumption level of the drill bit at the bottom of the well, and is directly related to the trajectory stability and wellbore shape. Through this step, the running data is not only the original signal, but also the second feature that has direct guiding value for the trajectory.

[0032] Finally, the first feature and the second feature are processed together. In order to avoid the imbalance caused by different dimensions and numerical ranges, normalization operation needs to be performed on all features to express them on the same scale. The normalized features are stacked in time sequence to form a continuous observation vector sequence, which is called an observation sequence for representing the drilling state. It contains feature information from both the wellbore attitude and the combined running at the bottom of the well, reflecting not only the change trend of the geometric position, but also the loading condition of the dynamics running.

[0033] Step S2: performing state estimation based on the observation sequence to generate a state estimation result containing wellbore attitude parameters and dynamics indicators.

[0034] Specifically, the observation sequence is input into a nonlinear state estimation model, a sequence component related to the wellbore attitude is extracted, and a hole angle, an azimuth angle, and a spatial curvature are solved to obtain a wellbore attitude parameter; meanwhile, a sequence component related to the bottom hole assembly operation is extracted and frequency domain analysis is performed to identify a stick-slip frequency band and an impact vibration frequency band and calculate corresponding strength values to form a dynamic index; and the wellbore attitude parameter and the dynamic index are combined and stored as a state estimation result.

[0035] Further, the nonlinear state estimation model is constructed by establishing nonlinear observation equations and state transition equations between the observation sequence and the state variable, and introducing a unscented Kalman filter algorithm or an extended Kalman filter algorithm into an equation solving framework, and is used to output a state estimation result containing the wellbore attitude parameter and the dynamic index in the state estimation process.

[0036] Further, the sequence component related to the bottom hole assembly operation is extracted and frequency domain analysis is performed to identify the stick-slip frequency band and the impact vibration frequency band and calculate corresponding strength values to form the dynamic index, including: extracting the sequence component related to the bottom hole assembly operation from the observation sequence; performing a fast Fourier transform based on the extracted sequence component to obtain a frequency spectrum distribution; identifying the stick-slip frequency band and the impact vibration frequency band in the frequency spectrum distribution, and calculating the amplitude and the energy strength in the identified frequency band range to form the dynamic index for the state estimation result.

[0037] In the specific implementation process, it is necessary to perform state estimation based on the observation sequence to obtain more accurate trajectory information and operating state information, thereby providing reliable input for subsequent constraint optimization and control decision. The core goal of the state estimation described herein is to extract information related to the wellbore geometric attitude and information related to the bottom hole assembly operating state at the same time on the basis of the multi-source data having been pre-processed in the early stage, and to fuse them through a reasonable mathematical model to obtain a state estimation result that can reflect the wellbore spatial position and the dynamic working condition at the same time. The result contains at least two types of information, one is a wellbore attitude parameter, mainly including a hole angle, an azimuth angle, and a spatial curvature, and the other is a dynamic index, mainly including strength indexes of stick-slip vibration and impact vibration.

[0038] Specifically, first of all, the observation sequence needs to be input into the nonlinear state estimation model. The observation sequence is constructed after the wellbore attitude data and the bottom hole assembly operating data are aligned, filtered and feature extracted in the foregoing steps, and it already has a unified time reference and a high signal-to-noise ratio. After inputting the model, the first step is to extract the sequence components related to the wellbore attitude from it, such as the sequence of the change of the inclination angle with time, the sequence of the change of the azimuth angle with time, and the spatial curvature sequence calculated based on the combination of the two. In the nonlinear state estimation model, these components are substituted into the observation equation, and through joint solving with the state transition equation, more smooth and accurate wellbore attitude parameters are iteratively obtained. Unlike traditional linear models, the introduction of nonlinear characteristics here can better reflect the real attitude change of the wellbore under complex formation conditions.

[0039] At the same time, sequence components related to the bottom hole assembly operation also need to be extracted from the observation sequence. These components mainly include parameters such as the drilling pressure, the rotation speed and the torque. Simply relying on these original quantities often makes it difficult to directly depict the dynamics of the downhole, so in this embodiment, frequency domain analysis is performed on these sequence components. Specifically, based on the fast Fourier transform (FFT) method, the time domain signal is converted into a frequency domain spectrum distribution. In this way, it is possible to identify the stick-slip frequency band and the impact vibration frequency band in the frequency domain. These two frequency bands are extremely critical indicators in the downhole dynamics state, the stick-slip frequency band corresponds to the phenomenon of intermittent sticking and slipping between the drill string and the well wall, and the impact vibration frequency band corresponds to the periodic impact of the drill bit or the drilling tool with the well wall. By identifying the range of these two frequency bands in the spectrum distribution and calculating their amplitude and energy intensity, quantitative dynamics indicators can be formed. These indicators not only reflect the stability of the operating conditions, but also indirectly infer whether the trajectory control is disturbed by abnormal operating conditions.

[0040] Further, in order to ensure the accuracy and robustness of state estimation, the construction of the nonlinear state estimation model also needs to be described. The model is realized by establishing nonlinear observation equations and state transition equations between the observation sequence and the state variable. Among them, the observation equation is used to describe the nonlinear mapping relationship between the observation (such as the inclination angle, torque, etc.) and the internal state variable, and the state transition equation is used to describe the dynamic law of the state variable evolution with time. In actual calculation, the traditional linear filtering method cannot effectively deal with nonlinear relationship, therefore, the unscented Kalman filter (UKF) algorithm or the extended Kalman filter (EKF) algorithm is introduced in the solving framework in the embodiment. The unscented Kalman filter avoids linearization error through sampling point propagation, and can provide high estimation accuracy in nonlinear scenarios; the extended Kalman filter realizes nonlinear solution through first-order Taylor expansion approximation, and has relatively small calculation amount, and is suitable for working conditions with high real-time requirement. Both methods can be embedded into the nonlinear state estimation model, and the appropriate algorithm can be selected according to the actual demand. Through the above way, the nonlinear state estimation model can output the state estimation result containing the wellbore attitude parameter and the dynamic index after inputting the observation sequence and through multiple rounds of iteration calculation.

[0041] Specifically, let the state variable vector be , wherein, is the inclination angle; is the azimuth angle; is the spatial curvature; is the change rate of the drilling pressure; is the change rate of the torque; T is the matrix transpose. The state transition equation is defined as:

[0042]

[0043] , wherein, is the control input (such as the speed setting); is the process noise.

[0044]

[0045] , wherein, is a nonlinear function defined by the drill string mechanics empirical formula; is the sampling interval; is the process noise.

[0046] Further, the observation variable vector is defined as:

[0047]

[0048] , wherein the superscript obs represents the attitude and running data collected by the sensor. The observation equation is defined as:

[0049]

[0050] wherein, ; is the observation noise.

[0051] Taking the extended Kalman filter solving process as an example, the state estimation step includes:

[0052] 1. Prediction, the corresponding prediction rule is:

[0053]

[0054] wherein, ; Q is the process noise covariance matrix.

[0055] 2. Update, the corresponding update rule is:

[0056]

[0057] wherein, ; R is the observation noise covariance matrix.

[0058] It should be pointed out that in the frequency domain analysis, in addition to the fast Fourier transform, a sliding window mechanism can also be introduced, that is, a segmented FFT analysis is performed within a certain time window, so that the dynamic changes of stick-slip or impact vibration can be better captured, rather than just obtaining a global frequency spectrum distribution. By calculating the amplitude and energy intensity in the identified stick-slip frequency band and impact vibration frequency band range, and continuously tracking in the time dimension, more stable and dynamic sensitive dynamic indicators can be formed. These indicators are combined with the wellbore attitude parameters to form the final state estimation result.

[0059] Step S3: constructing a constraint optimization model based on the state estimation result and a preset trajectory target, and performing model predictive control in the constraint optimization model to generate a control quantity sequence for driving the rotary steering tool.

[0060] Specifically, constructing a constraint optimization model based on the state estimation result and a preset trajectory target includes: comparing the wellbore attitude parameters and dynamic indicators in the state estimation result with the preset trajectory target to determine a current trajectory deviation amount and a stability constraint amount; establishing a nonlinear constraint set including a spatial curvature constraint, a build-up rate constraint and an anti-collision constraint based on the trajectory deviation amount and the stability constraint amount; converting the preset trajectory target into an objective function, and combining the objective function with the nonlinear constraint set to form an optimization solving structure; calling a model predictive control algorithm in the rolling time domain in the optimization solving structure to obtain a constraint optimization model.

[0061] Further, the model predictive control is performed in the constraint optimization model to generate a control quantity sequence for driving the rotary steerable tool, including: predicting a future wellbore posture change trend based on the state estimation result in a rolling time domain; calculating a trajectory deviation quantity for optimization solving by performing deviation calculation on the future wellbore posture change trend and the preset trajectory target; iteratively solving the objective function in the constraint optimization model under the trajectory deviation quantity constraint to obtain a control variable combination in a plurality of candidate solution spaces; and selecting a solution that satisfies the spatial curvature constraint, the build-up rate constraint and the anti-collision constraint simultaneously from the candidate solution space, and outputting a control quantity sequence composed of a deflection vane angle, a phase parameter and a weight on bit and rotary speed setting value.

[0062] In the embodiment of the present application, based on the state estimation result and the preset trajectory target, a constraint optimization model needs to be further constructed to realize closed-loop control of the drilling trajectory. The constraint optimization model is the core of the trajectory control problem, which is converted into a mathematical optimization problem. In the optimization problem, on the one hand, the trajectory deviation quantity is minimized by the objective function, and on the other hand, various physical constraint conditions are ensured to be met. In this way, the output result of the optimization can meet the accuracy requirements of the trajectory control and will not break the safety boundary of the downhole mechanics and geometric constraints.

[0063] Specifically, first, the state estimation result and the preset trajectory target need to be compared. The state estimation result includes wellbore posture parameters and dynamic indicators. The wellbore posture parameters include a hole inclination angle, an azimuth angle and a spatial curvature, which are used to reflect the geometric shape of the wellbore in the three-dimensional space. The dynamic indicators are obtained by frequency domain analysis, including the strength values of stick-slip vibration and impact vibration, which are used to represent the dynamic stability of the drill string in the drilling process. The preset trajectory target is an ideal wellbore path designed before construction, which usually defines a target hole inclination angle, a target azimuth angle and a preset trajectory curvature with well depth as the independent variable. By comparing the actual state with the target trajectory point by point, the trajectory deviation quantity can be obtained. The trajectory deviation quantity here not only has a geometric meaning, but also can be considered together with the dynamic indicators, for example, when the stick-slip index exceeds a certain threshold, the trajectory will be considered as an unstable state even if the geometric deviation is small. Therefore, the trajectory deviation quantity and the stability constraint quantity are obtained simultaneously in the comparison process.

[0064] On this basis, a set of nonlinear constraints needs to be established. This set includes at least three aspects: first, the spatial curvature constraint, that is, the curvature of the wellbore at any depth cannot exceed the preset limit value, otherwise it may cause excessive bending of the drill string or even sticking; second, the build rate constraint, that is, the change rate of the wellbore trajectory slope cannot be too large, so as to ensure that the steering tool can adjust the trajectory within a reasonable range, and not to produce a dramatic attitude jump; third, the anti-collision constraint, that is, in the multi-well group development scene, a certain safety distance must be maintained between wellbores to avoid trajectory intersection or wellbore collision. The above constraint conditions are all nonlinear constraints, and their mathematical forms can be written as inequalities, for example, the spatial curvature constraint can be written as , the build rate constraint can be written as , and the anti-collision constraint can be written as , wherein is the spatial curvature of the wellbore at a certain depth position; is the preset upper limit value of the spatial curvature; is the build rate, that is, the change rate of the wellbore trajectory inclination θ with respect to the depth L; is the preset upper limit value of the build rate is the spatial distance between any two wellbores i and j at position x in the multi-well group scene; is the preset safety distance threshold.

[0065] Further, the preset trajectory target is converted into an objective function. Generally, the objective function is defined as the sum of squares of the trajectory deviation, that is, , wherein are the deviation amounts of the inclination angle, the azimuth angle and the curvature, respectively. Of course, in some embodiments, the objective function can also add a weight term of dynamic stability, so that the optimization process takes into account both trajectory accuracy and running stability. Finally, the objective function and the aforementioned set of nonlinear constraints are combined to form a complete optimization solving structure, that is, the so-called constrained optimization model.

[0066] After the construction of the constrained optimization model is completed, the model predictive control (MPC) algorithm of the rolling time domain needs to be called to perform optimization solving. The so-called rolling time domain refers to considering only the trajectory and state evolution in a limited prediction time domain in the future within each control period, then optimizing the control variables within the prediction time domain, and finally only executing the first control action, and then rolling and updating the prediction window and repeating the optimization in the next control period. The advantage of this method is that it can correct the trajectory in real time in an uncertain environment, and has strong adaptability.

[0067] Specifically, the future wellbore attitude change trend is first predicted based on the state estimation result in the rolling time domain. The prediction process combines the state transition equation, and brings in the current inclination, azimuth and curvature to obtain the future wellbore trajectory evolution result. Subsequently, the future wellbore attitude change trend is compared with the preset trajectory target point by point to obtain the trajectory deviation in the prediction time domain. This trajectory deviation is one of the inputs of the subsequent optimization solution.

[0068] After obtaining the trajectory deviation, the iterative solution process of the constraint optimization is entered. At this time, the objective function is composed of the trajectory deviation in the prediction time domain, and the constraint conditions are composed of the spatial curvature constraint, the build-up rate constraint and the anti-collision constraint. Through the iterative optimization method (such as the sequential quadratic programming SQP or the interior point method), the control variable combination that satisfies the constraint condition is searched in the candidate solution space. The control variables mainly include the deflection piece angle, the tool phase and the setting values related to the drilling pressure and the rotating speed. These variables are the key parameters that directly act on the rotary steering tool.

[0069] After the iterative solution, multiple candidate solution combinations can be obtained. At this time, screening is needed, and finally the solutions with the minimum objective function under the premise of satisfying all nonlinear constraints are selected as the optimization output. The final output control quantity sequence includes the deflection piece angle, the phase parameter, the drilling pressure setting value and the rotating speed setting value. The control quantity sequence is sent to the rotary steering tool, so that the tool can be driven to adjust the trajectory according to the optimized instructions.

[0070] Step S4: applying the control quantity sequence to the rotary steering tool to perform trajectory adjustment, and performing residual error calculation and model parameter updating in real time based on the execution feedback information and the corresponding state estimation result until the drilling is completed.

[0071] Specifically, the control quantity sequence is applied to the rotary steering tool to perform trajectory adjustment, and residual error calculation and model parameter updating are performed in real time based on execution feedback information and corresponding state estimation results, including: the deflection piece angle and the phase parameter in the control quantity sequence are sent to the deflection driving assembly of the rotary steering tool, and the drilling pressure and rotating speed setting values are input to the ground control system to adjust the drilling pressure execution device and the rotating driving device; after the control quantity sequence is applied, the wellbore attitude data and operation data returned by the rotary steering tool are collected in real time, and the residual error is calculated by comparing with the state estimation result; when the residual error is lower than the preset residual error threshold, the original control period is maintained, and when the residual error exceeds the preset residual error threshold, the model parameter updating and the control period adjustment are triggered, and a new control quantity sequence is generated to act on the rotary steering tool again.

[0072] Further, when the residual exceeds the preset residual threshold, triggering the model parameter update and control cycle adjustment, and generating a new control sequence to act on the rotary steerable tool again, including: comparing the residual with the preset residual threshold, if the residual exceeds the preset residual threshold, calling the surrogate model for constrained optimization and correcting the friction coefficient and formation stiffness parameters in it based on the residual change to form the updated model parameters; adjusting the control cycle according to the magnitude of the residual exceeding the preset residual threshold, the greater the exceeding magnitude, the shorter the control cycle; re-executing the model predictive control under the updated model parameters and the new control cycle configuration to generate a new control sequence and act on the rotary steerable tool again.

[0073] In the implementation process, it is necessary to actually act the control sequence output by the foregoing constrained optimization model on the rotary steerable tool to drive the drill bit attitude and drilling parameters to perform trajectory adjustment according to the optimization instruction. The control sequence here is obtained by model predictive control, which contains multiple key variables, including the deflector angle, the phase parameter, the drilling pressure set value and the rotation speed set value. These quantities correspond to the direct control means of the rotary steerable tool and the downhole dynamics respectively, and are the basis for realizing dynamic correction of the well trajectory.

[0074] Specifically, first, the deflector angle and the phase parameter in the control sequence are sent to the deflection driving assembly of the rotary steerable tool, so that the bottom hole assembly can adjust the deflection direction and deflection amplitude according to the instruction, thereby correcting the drilling direction of the drill bit. At the same time, the drilling pressure and rotation speed set values are input to the ground control end, and by adjusting the drilling pressure execution device and the rotation driving device, the drill bit can be maintained in the set loading state during drilling. Through the above manner, the rotary steerable tool downhole and the ground transmission device work cooperatively, so that the trajectory adjustment can be gradually implemented into the actual drilling process according to the calculation results of the optimization model.

[0075] After the control sequence is completed, real-time feedback information returned by the rotary steerable tool needs to be obtained in time. These feedback information includes wellbore attitude data and bottom hole assembly operation data, which are consistent with the parameters in the foregoing state estimation results. For example, the attitude data includes inclination, azimuth and curvature, and the operation data includes drilling pressure, torque and rotation speed, etc. These data will be returned to the upper processing link at the end of the control cycle, and used for comparison with the current state estimation results. By calculating the difference between the feedback data and the state estimation results, the residual can be obtained. The meaning of residual here is "the difference between prediction and actual", which reflects the accuracy of the model in describing the downhole state, and also represents the real-time deviation level of the trajectory control.

[0076] Next, the control strategy needs to be decided based on the comparison of the residual with a preset threshold. If the residual is below the preset residual threshold, it means that the model prediction is consistent with the actual feedback, at which point the original control cycle can be maintained, and the new control amount sequence can continue to be generated by the original logic without additional model updating. If the residual exceeds the preset residual threshold, it indicates that there is a significant deviation between the trajectory prediction and the actual situation, which may be caused by sudden changes in formation conditions, abnormal drilling string dynamics (such as stick-slip or impact), or tool parameter mismatch, etc. At this point, model parameter updating and control cycle adjustment must be triggered to avoid further accumulation of trajectory deviation.

[0077] In the scenario of triggering updating, a proxy model for constrained optimization needs to be called first. The proxy model can be understood as a fast approximation version of the constrained optimization model, which retains the spatial curvature constraint, build-up rate constraint, and anti-collision constraint in structure, but allows dynamic adjustment at the parameter level. The specific approach is to compare the residual with the threshold, and if the residual exceeds the threshold, the friction coefficient and formation stiffness parameters in the proxy model are corrected according to the residual change. The friction coefficient is an important factor affecting the drilling string dynamics, which directly affects the mechanical stability of the downhole; the formation stiffness parameter affects the relationship between the wellbore attitude change and the bit stress. By adjusting these two parameters, the proxy model can be closer to the real downhole working conditions, avoiding the continuous expansion of the prediction deviation.

[0078] After the model parameters are corrected, the control cycle also needs to be adjusted according to the magnitude of the residual exceeding the threshold. The logic here is that the larger the residual, the more serious the deviation, and the model needs to be updated more frequently, so the control cycle must be shortened. For example, when the residual slightly exceeds the threshold, the prediction step can be appropriately shortened; when the residual is significantly over the threshold, both the prediction step and the control step need to be shortened to generate a new control amount sequence more quickly. In this way, trajectory control can dynamically adapt to different degrees of working condition deviation, thereby maintaining stability.

[0079] Finally, under the updated model parameters and the new control cycle configuration, the model predictive control is re-executed. The new MPC optimization process will be based on the corrected parameter set to predict the future trajectory evolution and iteratively solve the objective function under nonlinear constraint conditions. The new control amount sequence obtained in this way will be more consistent with the actual working conditions than the original sequence, containing the corrected deflection pad angle, phase parameter, and weight on bit and rotary speed set value. The new control amount sequence again acts on the rotary steerable tool to drive trajectory adjustment. With the continuous repetition of this closed-loop process, trajectory control can continuously correct the deviation between the model and the actual situation in each cycle, ensuring that the wellbore trajectory closely follows the preset target.

[0080] As Figure 2As shown, the embodiment of the present application provides a high-precision drilling rotary steerable trajectory intelligent control device, which comprises: an acquisition unit configured to acquire multi-source observation data including wellbore attitude data and bottom-hole combined operation data in a target drilling process, and construct an observation sequence for representing a drilling state based on the multi-source observation data; an estimation unit configured to perform state estimation based on the observation sequence, and generate a state estimation result containing wellbore attitude parameters and kinetic indicators; a scheme generation unit configured to construct a constraint optimization model based on the state estimation result and a preset trajectory target, and perform model predictive control in the constraint optimization model to generate a control quantity sequence for driving a rotary steerable tool; and an execution unit configured to apply the control quantity sequence to the rotary steerable tool to perform trajectory adjustment, and perform residual error calculation and model parameter updating based on execution feedback information and a corresponding state estimation result in real time until the drilling is completed.

[0081] Those skilled in the art can understand that all or part of the steps of the method for implementing the above-mentioned embodiments can be completed by programs instructing related hardware, the programs are stored in a storage medium, and the programs include a plurality of instructions for causing a single-chip microcomputer, a chip or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various storage medium capable of storing program codes.

[0082] The above describes the optional embodiments of the present application in detail in combination with the drawings, but the embodiments of the present application are not limited to the specific details in the above-mentioned embodiments. Within the technical concept range of the embodiments of the present application, the technical solutions of the embodiments of the present application can be subjected to various simple modifications, and these simple modifications all belong to the protection range of the embodiments of the present application. In addition, it should be noted that each specific technical feature described in the above-mentioned specific embodiments can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, the embodiments of the present application will not further describe various possible combination manners.

[0083] In addition, various different embodiments of the present application can also be combined in any manner, as long as it does not deviate from the idea of the embodiments of the present application, and it should be considered as the disclosed content of the embodiments of the present application.

Claims

1. A high-precision drilling rotary steerable trajectory intelligent control method, characterized in that, The method comprises: acquiring multi-source observation data including wellbore posture data and bottom-hole combined operation data in a target drilling process, and constructing an observation sequence for representing a drilling state based on the multi-source observation data; performing state estimation based on the observation sequence to generate a state estimation result containing wellbore posture parameters and kinetic indicators; constructing a constraint optimization model based on the state estimation result and a preset trajectory target, and performing model predictive control in the constraint optimization model to generate a control quantity sequence for driving a rotary steerable tool; wherein constructing a constraint optimization model based on the state estimation result and a preset trajectory target comprises: comparing the wellbore posture parameters and kinetic indicators in the state estimation result with the preset trajectory target to determine a current trajectory deviation amount and a stability constraint amount; establishing a nonlinear constraint set including a spatial curvature constraint, a build-up rate constraint and an anti-collision constraint based on the trajectory deviation amount and the stability constraint amount; converting the preset trajectory target into an objective function, and combining the objective function with the nonlinear constraint set to form an optimization solving structure; and calling a model predictive control algorithm of a rolling time domain in the optimization solving structure to obtain a constraint optimization model; performing model predictive control in the constraint optimization model to generate a control quantity sequence for driving a rotary steerable tool comprises: predicting a future wellbore posture change trend based on the state estimation result in a rolling time domain; performing deviation calculation on the future wellbore posture change trend and the preset trajectory target to obtain a trajectory deviation amount for optimization solving; iteratively solving the objective function in the constraint optimization model under the trajectory deviation amount constraint to obtain a control variable combination in a plurality of candidate solution spaces; selecting a solution that simultaneously satisfies the spatial curvature constraint, the build-up rate constraint and the anti-collision constraint from the candidate solution spaces, and outputting a control quantity sequence composed of a deflector angle, a phase parameter and drilling pressure and rotation speed setting values; applying the control quantity sequence to the rotary steerable tool to perform trajectory adjustment, and performing residual error calculation and model parameter updating based on execution feedback information and corresponding state estimation results in real time until the drilling is completed; wherein applying the control quantity sequence to the rotary steerable tool to perform trajectory adjustment, and performing residual error calculation and model parameter updating based on execution feedback information and corresponding state estimation results in real time comprises: issuing the deflector angle and the phase parameter in the control quantity sequence to a deflection driving assembly of the rotary steerable tool, and simultaneously inputting the drilling pressure and rotation speed setting values to a ground control system to adjust a drilling pressure execution device and a rotation driving device; after the control quantity sequence is applied, real-time acquisition of wellbore posture data and operation data returned by the rotary steerable tool is performed, and residual error calculation is performed by comparing the wellbore posture data and the operation data with the state estimation result; when the residual error is lower than a preset residual error threshold, the original control period is maintained; when the residual error exceeds the preset residual error threshold, model parameter updating and control period adjustment are triggered, and a new control quantity sequence is generated to act on the rotary steerable tool again.

2. The high-precision drilling rotary steerable trajectory intelligent control method according to claim 1, characterized in that, The wellbore posture data includes a hole inclination angle, an azimuth angle and a spatial curvature parameter. The bottom hole assembly operating data includes weight on bit, rotary speed and torque; Based on the multi-source observation data, an observation sequence for representing the drilling state is constructed, including: Timestamps of the multi-source observation data are uniformly aligned, and interpolation resampling is performed based on the aligned data to obtain a synchronous sequence; outlier rejection and filtering processing are performed on the synchronous sequence, and corresponding attitude denoising sequence and operating denoising sequence are obtained based on the processed wellbore attitude data and bottom hole assembly operating data respectively; Based on the attitude denoising sequence, the variation rates of inclination, azimuth and spatial curvature are extracted as first features; based on the operating denoising sequence, the rotary speed, torque and mechanical power proxy are extracted as second features; After normalization, the first features and the second features are connected in time sequence to form the observation sequence for representing the drilling state.

3. The high-precision drilling rotary steerable trajectory intelligent control method of claim 1, wherein, Based on the observation sequence, state estimation is performed to generate state estimation results including wellbore attitude parameters and dynamic indicators, including: The observation sequence is input into a nonlinear state estimation model, sequence components related to wellbore attitude are extracted, and inclination, azimuth and spatial curvature are solved to obtain wellbore attitude parameters; At the same time, sequence components related to bottom hole assembly operation are extracted and frequency domain analysis is performed to identify stick-slip frequency band and impact vibration frequency band and calculate corresponding strength values to form dynamic indicators; The wellbore attitude parameters and the dynamic indicators are combined and stored as state estimation results.

4. The high-precision drilling rotary steerable trajectory intelligent control method according to claim 3, characterized in that, The nonlinear state estimation model is constructed by establishing nonlinear observation equations and state transition equations between the observation sequence and the state variable, and introducing unscented Kalman filter algorithm or extended Kalman filter algorithm into the equation solving framework, which is used to output the state estimation results including wellbore attitude parameters and dynamic indicators in the state estimation process.

5. The high-precision drilling rotary steerable trajectory intelligent control method according to claim 3, characterized in that, Extracting sequence components related to bottom hole assembly operation and performing frequency domain analysis to identify stick-slip frequency band and impact vibration frequency band and calculate corresponding strength values to form dynamic indicators, including: Sequence components related to bottom hole assembly operation are extracted from the observation sequence; Based on the extracted sequence components, fast Fourier transform is performed to obtain frequency spectrum distribution; In the frequency spectrum distribution, the stick-slip frequency band and the impact vibration frequency band are identified, and the amplitude and energy intensity are calculated in the identified frequency band range to form the dynamic indicators for the state estimation results.

6. The high-precision drilling rotary steerable trajectory intelligent control method of claim 1, wherein, When the residual exceeds the preset residual threshold, model parameter updating and control period adjustment are triggered, and a new control amount sequence is generated to act on the rotary steering tool again, including: The residual is compared with the preset residual threshold, if the residual exceeds the preset residual threshold, a proxy model for constrained optimization is called and the friction coefficient and the formation stiffness parameter are corrected based on the residual change to form updated model parameters; According to the amplitude of the residual exceeding the preset residual threshold, the control period is adjusted, the greater the exceeding amplitude, the shorter the control period; Under the configuration of the updated model parameters and the new control period, model predictive control is re-executed to generate a new control amount sequence and act on the rotary steering tool again.

7. A high-precision drilling rotary steerable trajectory intelligent control device, characterized in that, The device includes: An acquisition unit is configured to acquire multi-source observation data including wellbore posture data and bottom-hole combined operation data in a target drilling process, and construct an observation sequence for representing a drilling state based on the multi-source observation data; An estimation unit is configured to perform state estimation based on the observation sequence, and generate a state estimation result including wellbore posture parameters and dynamic indicators; A scheme generation unit is configured to construct a constraint optimization model based on the state estimation result and a preset trajectory target, and perform model predictive control in the constraint optimization model to generate a control quantity sequence for driving a rotary steerable tool; wherein The constraint optimization model is constructed based on the state estimation result and the preset trajectory target, including: comparing the wellbore posture parameters and the dynamic indicators in the state estimation result with the preset trajectory target to determine a current trajectory deviation amount and a stability constraint amount; establishing a nonlinear constraint set including a spatial curvature constraint, a build-up rate constraint and an anti-collision constraint based on the trajectory deviation amount and the stability constraint amount; converting the preset trajectory target into an objective function, and combining the objective function with the nonlinear constraint set to form an optimization solving structure; and calling a model predictive control algorithm of a rolling time domain in the optimization solving structure to obtain a constraint optimization model; The model predictive control is performed in the constraint optimization model to generate the control quantity sequence for driving the rotary steerable tool, including: predicting a future wellbore posture change trend based on the state estimation result in a rolling time domain; performing deviation calculation on the future wellbore posture change trend and the preset trajectory target to obtain a trajectory deviation amount for optimization solving; iteratively solving the objective function in the constraint optimization model under the trajectory deviation amount constraint to obtain a control variable combination in a plurality of candidate solution spaces; selecting a solution that simultaneously satisfies the spatial curvature constraint, the build-up rate constraint and the anti-collision constraint from the candidate solution spaces, and outputting a control quantity sequence composed of a deflector angle, a phase parameter and drilling pressure and rotation speed setting values; An execution unit is configured to apply the control quantity sequence to the rotary steerable tool to perform trajectory adjustment, and perform residual error calculation and model parameter updating based on execution feedback information and corresponding state estimation results in real time until the drilling is completed; wherein The control quantity sequence is applied to the rotary steerable tool to perform trajectory adjustment, and residual error calculation and model parameter updating are performed based on execution feedback information and corresponding state estimation results in real time, including: issuing the deflector angle and the phase parameter in the control quantity sequence to a deflection driving assembly of the rotary steerable tool, and simultaneously inputting the drilling pressure and rotation speed setting values to a ground control system to adjust a drilling pressure execution device and a rotation driving device; after the control quantity sequence is applied, wellbore posture data and operation data returned by the rotary steerable tool are collected in real time, and residual error is calculated by comparing the wellbore posture data and the operation data with the state estimation result; when the residual error is lower than a preset residual error threshold, the original control period is maintained; when the residual error exceeds the preset residual error threshold, model parameter updating and control period adjustment are triggered, and a new control quantity sequence is generated to be applied to the rotary steerable tool again.

Images