Underground geosteering closed-loop control system and control method

Abstract

The invention provides an underground geosteering closed-loop control system and method.The underground geosteering closed-loop control system and method can complete real-time design track correction and borehole track and control instruction generation underground so as to improve the accuracy and response speed of a control instruction, and the system comprises an information collection unit arranged underground and including a dynamic attitude measurement sensor LWD short section; the central control unit is in communication connection with the information acquisition unit; a real drilling track calculation module, a design track correction module, a self-closed loop control program module and an artificial intelligence prediction module are arranged in the central control unit; a track correction module is designed to obtain geological parameter data of the LWD short section, and the deviation value delta 0 between the geological parameter data and the geological requirement is calculated; when the geology deviation value delta0 is larger than a threshold value epsilon0, a second design trajectory is regenerated according to the geosteering trajectory adjustment principle; and the artificial intelligence prediction module inputs the first to-be-drilled track or the second to-be-drilled track into the control instruction artificial intelligence prediction model, and outputs a new rotary steering tool control instruction.

Description

Technical Field

[0001] The present invention relates to the field of oil and gas exploration and development drilling, and in particular to an underground geological steering closed-loop control system and a control method. Background Art

[0002] Patent application number 202311781609.2 is a method for implementing downhole closed-loop control of wellbore trajectory, which can complete the calculation of actual drilling trajectory underground, and its operation method is as follows: S1, sending well depth information and control instructions to the downhole central control; S2, recording the current time t0 corresponding to the current well depth information d0; S3, reading the measurement data of each attitude sensor; S4, data matching analysis, when the matching degree is the highest, marking the time as t1, and obtaining the well depth information d1 at the position of attitude sensor 1 at time t1 through trajectory calculation to complete the calculation of the wellbore trajectory; S5, comparing the trajectory requirements in the control instruction with the calculated wellbore trajectory, adjusting the rotary steering execution force parameters, so that the actual drilling trajectory is consistent with the control instruction.

[0003] Although it can complete the actual drilling trajectory calculation underground, reducing the time of uploading the measured data to the ground to complete the actual drilling trajectory calculation, and using PID self-feedback control to complete data measurement, comparison and execution underground, it can complete the underground self-feedback process, but in actual use, there are at least the following problems: 1. The distance interval of downhole trajectory calculation is the distance between two attitude sensors, so the adjustment interval of control instructions is also constrained by this distance, thus affecting the accuracy of control instructions; 2. PID self-feedback control is adopted. During the drilling process, the control command changes continuously with the progress of drilling. It is a dynamic change process, but the dynamic quality margin of PID is not large. The closed-loop dynamic quality is sensitive to the change of PID gain. Therefore, it is necessary to adjust the PID gain frequently according to environmental changes, resulting in low accuracy of command output; 3. It is not entirely reasonable for the PID controller to directly take the error between the target and the actual behavior, which causes overshoot in the output process and slow command output; 4. The integral link in PID control has a significant effect on suppressing constant disturbances. However, when there is no disturbance, the integral link makes the dynamic characteristics of the system worse, resulting in a sluggish response of the downhole closed-loop control, which is prone to side effects such as oscillation and control quantity saturation. As for disturbances that change over time, the integral link's suppression ability is not significant.

[0004] At the same time, since the uncertainty of geological factors has a great impact on the safety of drilling projects, a detailed geological survey must be conducted before drilling construction to ensure the scientificity and rationality of safe drilling construction, provide important reference value for exploration and development, and form a design trajectory. However, during the drilling process, when the geological conditions of the actual drilling trajectory deviate from the geological conditions at the point corresponding to the design trajectory, drilling based on the control instructions made by the preset design trajectory may cause stuck drill, well leakage, overflow and other phenomena, which will directly affect drilling safety. Summary of the invention

[0005] The purpose of the present invention is to provide an underground geological steering closed-loop control system and a control method thereof, which can complete the designed trajectory correction based on the real-time drilling geological conditions underground to ensure drilling safety; at the same time, new artificial intelligence prediction technology is applied to the wellbore trajectory and control instruction prediction to improve the accuracy and response speed of the control instructions, thereby improving the accuracy, quality and efficiency of trajectory control.

[0006] The embodiment of the present invention is achieved as follows: A downhole geosteering closed-loop control system, the system comprising: An information acquisition unit, the information acquisition unit includes a plurality of dynamic attitude measurement sensors arranged at intervals and a plurality of LWD short sections arranged at intervals from the dynamic attitude measurement sensors, for respectively acquiring drilling tool attitude data and geological parameter data; The central control unit is connected to the information acquisition unit in communication; the central control unit is provided with a real drilling trajectory calculation module, a design trajectory correction module, a self-closed loop control program module, an artificial intelligence prediction module and a communication module. The real drilling trajectory calculation module obtains the drilling tool posture data of the dynamic posture measurement sensor to generate the real drilling trajectory by fitting; the design trajectory correction module obtains the geological parameter data of the LWD short section and calculates the deviation Δ0 from the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, the artificial intelligence prediction model is used to automatically design the first trajectory to be drilled m meters ahead of the drill bit according to the geological guidance trajectory adjustment principle, and the first trajectory to be drilled is automatically designed according to the first trajectory to be drilled. The drilling trajectory updates the first design trajectory to obtain the second design trajectory; the self-closed loop control program module compares the continuous actual drilling trajectory with the first / second design trajectory to calculate the geometric deviation Δ1 when the geological deviation Δ0 is less than or equal to the threshold ε0; when the geometric deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters in front of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely connected; the artificial intelligence prediction module is configured to: input the first trajectory to be drilled or the second trajectory to be drilled into the control instruction artificial intelligence prediction model, and output a new rotary steering tool control instruction; The execution unit is connected to the central control unit for communication; the execution unit obtains the control instructions of the rotary steering tool to control the rib-guided drilling; the information acquisition unit continuously obtains the measurement data of the execution unit to form an underground geological steering closed loop.

[0007] In a preferred embodiment of the present invention, the above-mentioned geological parameter data at least include azimuthal gamma, azimuthal resistivity, while-drilling sonic wave, while-drilling neutron density, and while-drilling nuclear magnetic parameters.

[0008] In a preferred embodiment of the present invention, the artificial intelligence prediction module is integrated into the central control unit and is further configured as follows: The drilling operation data and rotary steering operation instruction data of the drilled wells are obtained, and abnormal data cleaning and data enhancement are performed; the drilling operation data are the well deviation a at the drill bit, the drill bit position b, the drill tool shape data c within n meters behind the drill bit, the average drilling pressure d within m meters in front of the drill bit, the average speed e, the average torque f, the drilling time g, the average displacement h, the well deviation change rate i and the change direction j within m meters in front of the drill bit; the rotary steering operation instruction data are the average value k and the average direction l of the resultant force of the steering tool within m meters in front of the drill bit; Constructing a sample set, using the cleaning and enhancement completed data to construct a sample set, using the drilling operation data as input, and using the rotary steering operation instruction data as output; Select a machine learning model with a regression algorithm, divide the constructed sample set into a training set, a validation set, and a test set, and perform training, validation, and testing on the machine learning model to obtain a mature control command artificial intelligence prediction model; The data related to the sample set is extracted from the actual drilling data and input into a mature artificial intelligence prediction model for control instructions to obtain new rotary steerable tool control instructions in real time.

[0009] In a preferred embodiment of the present invention, the abnormal data cleaning method of the above-mentioned well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit is further configured as follows: Sorting the dynamic well deviation and dynamic azimuth data during the drilling operation of a single well in chronological order; A first sliding window with a width of W is formed with a drill rod length W; Assuming that the deflection capability of the steering tool is X° / Ym, calculate the median M1 of all the measured values ​​in the first sliding window, and obtain the determination intervals [M1-X*W / Y, M1+X*W / Y] of the well inclination a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit; The first sliding window is used to slide and judge and clean the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit in the entire well section, and the data outside the judgment interval is deleted.

[0010] In a preferred embodiment of the present invention, the abnormal data cleaning method of the well deviation change rate i and change direction j within m meters before the drill bit is further configured as follows: According to different instructions, the magnitude and direction data of the deflection force corresponding to the instructions are filtered out; Arrange the deflection force magnitude and direction data according to the commanded start and end depths; Calculate the median M2 of all deflection force magnitude and direction data under the instruction condition; According to the fluctuation amplitude of the deflecting force and direction data under the command conditions, the threshold adjustment proportional coefficient r1 is determined, and the judgment interval of the well inclination change rate h and the change direction j is [M2- M2*r1, M2+ M2*r1], and the data outside the judgment interval is deleted.

[0011] In a preferred embodiment of the present invention, the abnormal data cleaning method of the average drilling pressure d, the average speed e, the average torque f, the drilling time g, and the average displacement h within the first m meters of the drill bit is further configured as follows: Arrange the drilling pressure, speed, torque, drilling time and displacement data in chronological order; A second sliding window with a width of W is formed with a drill rod length W; Calculate the median M3 of the corresponding data type in the second sliding window; According to the fluctuation range of different types of data, the threshold adjustment proportional coefficient r2 is determined. The judgment interval of drilling pressure, torque, speed, displacement and drilling time data is [M3- M3*r2, M3+ M3*r2], and the data outside the judgment interval is deleted.

[0012] In a preferred embodiment of the present invention, the above data enhancement method is further configured to: fill in missing values ​​and perform data encryption interpolation on the data cleaned along the well depth to make the data complete and smooth.

[0013] In a preferred embodiment of the present invention, the above-mentioned method of dividing the constructed sample set into a training set, a validation set and a test set for training, validation and testing on a machine learning model is further configured as follows: the test set is samples from one or several wells; the training set and the validation set are samples obtained from other wells different from the test set; 80% of the data are extracted from samples from different wells in a random order as the training set, and the remaining 20% ​​of the data are used as the validation set.

[0014] In a preferred embodiment of the present invention, the method for generating the above-mentioned second designed trajectory is further configured as follows: when the geological deviation Δ0 is greater than the threshold ε0, the artificial intelligence prediction model is used to automatically generate the first trajectory to be drilled m meters ahead of the drill bit, and the first trajectory to be drilled is spliced ​​onto the first designed trajectory to ensure that the well inclination and azimuth of the junction point or junction section of the first designed trajectory and the first trajectory to be drilled are exactly the same.

[0015] A downhole geosteering closed-loop control method, the control method comprising: The information acquisition unit is used to obtain the drilling tool posture data and the geological parameter data respectively; The actual drilling trajectory calculation module of the central control unit obtains the drilling tool posture data to generate the actual drilling trajectory; the design trajectory correction module obtains the geological parameter data and calculates the deviation Δ0 between the geological parameter data and the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to automatically design the first trajectory to be drilled m meters ahead of the drill bit, and the first design trajectory is updated according to the first trajectory to be drilled to obtain the second design trajectory; when the geological deviation Δ0 is less than or equal to the threshold ε0, the self-closed-loop control program module compares the continuous actual drilling trajectory with the first / second design trajectory to calculate the geometric deviation Δ1; when the deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters ahead of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely connected; the first trajectory to be drilled or the second trajectory to be drilled is input into the control instruction artificial intelligence prediction model, and a new rotary steering tool control instruction is output; The execution unit performs drilling operations according to the control instructions of the rotary steerable tool.

[0016] The beneficial effects of the embodiments of the present invention are: 1. In this solution, information collection, generation, comparison, correction, and control command generation of the designed trajectory / trace to be drilled are all completed underground, which reduces the time for information transmission and reduces ground intervention; 2. Before the actual drilling trajectory matches the designed trajectory, the geological deviation Δ0 on the actual drilling trajectory is also compared; when the geological deviation is too large, the first trajectory to be drilled m meters before the drill bit of the actual drilling trajectory is redesigned according to the geological guidance trajectory adjustment principle. The first trajectory to be drilled m meters before is generated by the artificial intelligence prediction model, and the first trajectory to be drilled is spliced ​​on the first designed trajectory to form a second designed trajectory; until the geological deviation meets the requirements, the actual drilling trajectory and the designed trajectory are compared again, so that the later control instructions can be adjusted in real time according to geological factors, thereby improving the control accuracy of downhole drilling and ensuring the quality and efficiency of the drilling trajectory; 3. The second design trajectory and the first m meters of the trajectory to be drilled are determined by the artificial intelligence prediction model. Intelligent design shortens the trajectory design time and the decision-making output time of the control command, thereby improving work efficiency; 4. In the control instruction giving link, an artificial intelligence prediction model is used to output the control instructions of the rotary steering tool, which can improve the intelligence level of the rotary steering tool, improve the trajectory control accuracy, quality and efficiency, and reduce the occurrence of errors in control instructions. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 This is a functional block diagram of a downhole geosteering closed-loop control system according to an embodiment of the present invention; Figure 2 It is a schematic diagram of the placement structure of the downhole sensor and the LWD short sub according to an embodiment of the present invention; Figure 3 This is a flow chart of actual drilling trajectory fitting according to an embodiment of the present invention; Figure 4 This is a schematic diagram of measurement values ​​within a time period of Δt read from each sensor at time t0 in an embodiment of the present invention; Figure 5 Schematic diagram of a method for calculating well depth information at time t1 according to an embodiment of the present invention; Figure 6 A flow chart of a data processing method for a design trajectory correction module according to an embodiment of the present invention; Figure 7 This is a flow chart of a data processing method of an artificial intelligence prediction module according to an embodiment of the present invention; Figure 8 The process of the underground geosteering closed-loop control method according to the embodiment of the present invention is as follows Figure 1 ; Fig. 9 The process of the underground geosteering closed-loop control method according to the embodiment of the present invention is as follows Figure 2 .

[0019] Icons: information acquisition unit 100; dynamic posture measurement sensor 110; first posture measurement sensor 111; second posture measurement sensor 112; LWD short section 120; central control unit 200; actual drilling trajectory calculation module 210; designed trajectory correction module 220; self-closed loop control program module 230; artificial intelligence prediction module 240; communication module 250; execution unit 300; execution unit 300. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0023] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concept of the example embodiments to those skilled in the art.

[0024] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0025] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0026] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0027] First embodiment See also Figure 1 This embodiment provides a downhole geosteering closed-loop control system, which includes: An information acquisition unit 100, the information acquisition unit 100 comprises a plurality of dynamic attitude measurement sensors 110 arranged at intervals and a plurality of LWD short sections 120 arranged at intervals from the dynamic attitude measurement sensors 110, for respectively acquiring drilling tool attitude data and geological parameter data; The central control unit 200 is connected to the information acquisition unit 100 for communication; the central control unit 200 is provided with a real drilling trajectory calculation module 210, a design trajectory correction module 220, a self-closed loop control program module 230, an artificial intelligence prediction module 240 and a communication module 250; the real drilling trajectory calculation module 210 obtains the drilling tool posture data of the dynamic posture measurement sensor 110 to fit and generate the real drilling trajectory; the design trajectory correction module 220 obtains the geological parameter data of the LWD short section 120, and calculates the deviation Δ0 between the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to automatically design the drill bit front m The first trajectory to be drilled is m meters, and the first design trajectory is updated according to the first trajectory to be drilled to obtain the second design trajectory; the self-closed loop control program module 230, when the geological deviation Δ0 is less than or equal to the threshold ε0, compares the continuous actual drilling trajectory with the first / second design trajectory to calculate the geometric deviation Δ1; when the geometric deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters ahead of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely connected; the artificial intelligence prediction module 240 is configured to: input the first trajectory to be drilled or the second trajectory to be drilled into the control instruction artificial intelligence prediction model, and output a new rotary steering tool control instruction; The execution unit 300 is connected to the central control unit 200 for communication; the execution unit 300 obtains the control instruction of the rotary steering tool to control the rib 310 to guide the drilling; the information acquisition unit 100 continuously obtains the measurement data of the execution unit 300 to form an underground geological steering closed loop.

[0028] The information acquisition unit 100 collects attitude data and geological parameter data in real time as the drilling progresses, and transmits the above data to the central control unit 200 for further processing; the central control unit 200 completes the actual drilling trajectory generation, geological deviation calculation, geometric deviation calculation, control instruction output, etc. according to the real-time collected data, and transmits the output control instruction to the execution unit 300; the execution unit 300 controls the rib 310 to guide the drilling according to the control instruction, and the information acquisition unit 100 continuously collects two types of data to form an underground closed-loop control. Information acquisition, generation, comparison, correction of the design trajectory / trail to be drilled, generation of control instructions, etc. are all completed underground, which reduces the time for information transmission up and down, and reduces ground intervention.

[0029] For details, see Figure 2The dynamic attitude measurement sensors 110 arranged at intervals include a first attitude measurement sensor 111 arranged on the rotary steering tool and a plurality of second attitude measurement sensors 112 arranged at intervals on the drill rod, which are used to collect the attitude data of the drill tool at the corresponding position in real time. The first attitude measurement sensor 111 is as close to the drill bit as possible. The dynamic attitude measurement sensors 110 only need to be placed at intervals, and there is no restriction on whether they are integrated with other measurement or control tools, and there is no restriction on the interval distance, which is generally within the range of several meters to more than ten meters.

[0030] There is no restriction on the setting positions of the multiple LWD pup joints 120, and the more pup joint types, the better. The measurement pup joints such as gamma, resistivity, acoustic wave, and nuclear logging can be set on the rotary steering tool and the drill pipe, or only on the drill pipe. The geological parameter data at least include azimuthal gamma, azimuthal resistivity, while drilling acoustic wave, while drilling neutron density, and while drilling nuclear magnetic parameters.

[0031] The communication module 250 of the central control unit 200 is used to communicate with the information collection unit 100 and the execution unit 300, which can be wirelessly connected or electrically connected, and is used to obtain the collected data and issue control instructions to the execution unit 300. In fact, the functional modules of the drilling trajectory calculation module 210, the design trajectory correction module 220, the self-closed loop control program module 230, and the artificial intelligence prediction module 240 are used to complete calculations of different contents. Each functional module is described in turn below.

[0032] The actual drilling trajectory calculation module 210 is used to generate the actual drilling trajectory by fitting the drilling tool posture data obtained from the dynamic posture measurement sensor 110. Figure 3-5 , the calculation steps include: The time when the current well depth information d0 (including sounding, vertical depth and horizontal displacement), drilling operation parameters (bit pressure, rotation speed, torque, displacement, drilling speed, etc.), initial control instructions (force magnitude and force direction) of the rotary steering tool and steering decision criteria (including pre-collected geological parameters and preset thresholds ε0 and ε1 of geometric parameter changes, the first design trajectory, and the design principles of trajectory adjustment (including segmented design principles, well inclination change rate range, and the docking principle of the new estimated trajectory to be drilled with the original design trajectory, etc.)) are sent from the surface to the well, is recorded as t0; The actual drilling trajectory calculation module 210 reads the measurement data of each attitude sensor within the time period Δt before the current time t0 , , …, , including well inclination and azimuth, data length n=Fs*Δt, Fs is the sampling frequency, and then the measurement data of the first attitude measurement sensor 111 is As the reference value R; During the drilling process, the measurement data of the second attitude measurement sensor 112 is transmitted in real time. Correlation matching is performed with the reference value R, and the time of maximum correlation is marked as t1, at which the drilling distance is equal to the position distance ΔL of the two sensors; By using the measurement data of the first posture measurement sensor 111 at time t0 and t1 and combining the drilling distance ΔL, the position of the first posture measurement sensor 111 at time t1 is located, and the actual drilling wellbore trajectory is calculated.

[0033] In the correlation matching process of the actual drilling trajectory generation step, the specific operation method is: drilling according to the received initial control instruction of the rotary steering tool, and as time goes by, the well depth increases, and the data obtained from the second posture measurement sensor 112 Constantly changing, using a method similar to sliding the sensor with Δt as the time window to obtain data, each time a position is slid, the current and To perform data matching analysis, a cross-correlation method may be used, but is not limited to this method. When the second posture measurement sensor 112 moves forward to the position of the first posture measurement sensor 111, and The two groups of numbers should be basically the same, and the matching degree of the two groups of numbers is the highest at this time. This moment can be marked as time t1 based on this feature. Obviously, from time t0 to time t1, the length drilled by the drill bit is equal to the distance ΔL between the first posture measurement sensor 111 and the second posture measurement sensor 112.

[0034] In the step of generating the actual drilling trajectory, the specific operation method is as follows: based on the distance and the drilling tool posture measured by the first posture measurement sensor 111 at time t0 and time t1 and Combined with the well depth information d0 at time t0 received by the central control, according to the trajectory calculation method commonly used in the industry, such as the minimum curvature method, the well depth information d1 at the position of the first attitude measurement sensor 111 at time t1 can be easily obtained, thereby realizing the calculation of the wellbore trajectory.

[0035] Since geological factors have a great impact on the safety of drilling projects during drilling operations, inaccurate geological predictions will significantly increase the probability of drilling project accidents, which may cause significant losses to workers and oilfield companies.

[0036] The existing trajectory comparison method is: when designing a drilling project, the geological analysis of the exploration project should be carried out in depth according to the actual situation, and then the design trajectory preset before going down the well is compared with the actual drilling trajectory. According to the comparison, it is determined whether the trajectory needs to be adjusted. When there is an error between the geological conditions during the actual drilling process and the pre-collected geological data, the subsequent drilling trajectory will have a large error, thus affecting the entire drilling operation.

[0037] In this embodiment, a design trajectory correction module 220 is added, which is used to obtain a deviation Δ according to the difference between the real-time measured geological parameter data and the geological requirements of the design trajectory at the point / time. 0， Determine whether design trajectory correction is needed, see Figure 6 , the specific steps are as follows: Continuously subtract the geological parameter data from the geological requirement data along the actual drilling trajectory to obtain the geological deviation Δ0; When the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to design the first trajectory to be drilled m meters ahead of the drill bit, and the first designed trajectory is updated according to the first trajectory to be drilled to obtain the second designed trajectory; the second designed trajectory is a junction point or junction segment from the end point of the actual drilling trajectory to the starting point of the first trajectory to be drilled, and the end point of the first trajectory to be drilled is joined to the first designed trajectory, and the spatial position and posture of each junction point or junction segment are guaranteed to be exactly the same; that is, when the geological data collected in real time is different from the geological data before going down to the well, it is necessary to regenerate a new first trajectory to be drilled to adjust the current actual drilling point, and re-update the designed trajectory, and perform a dual adjustment process of the actual drilling trajectory and the designed trajectory; When the geological deviation Δ0 is less than or equal to the threshold ε0, further trajectory determination is performed, that is, the continuous actual drilling trajectory is compared with the first / second designed trajectory to calculate the geometric deviation Δ1.

[0038] The self-closed loop control program module 230 is used to calculate the geometric deviation Δ1 when the geological deviation meets the conditions. The specific operation method is as follows: When the geometric deviation Δ1 exceeds the given threshold ε1, the actual drilling trajectory needs to be adjusted. First, the artificial intelligence prediction model redesigns the trajectory to be drilled m meters in front of the drill bit, requiring that the trajectory to be drilled can be completely connected with the designed trajectory, that is, the spatial position and posture of the junction point are completely consistent. In particular, before the time t for trajectory judgment, all geological deviations Δ0 in the time period t are less than or equal to the threshold ε0, then the first designed trajectory is used for comparison and judgment with the actual drilling trajectory. When the geological deviation Δ0 at any time in the time period t is greater than the threshold ε0, the second designed trajectory is used for comparison and judgment with the actual drilling trajectory.

[0039] The artificial intelligence prediction module 240 is used to intelligently output control instructions and determine the length of the trajectory to be drilled, the selection of the splicing point or segment. When there is a calculation of the trajectory to be drilled, the generated trajectory to be drilled and the drilling operation data need to be input into the artificial intelligence prediction model to generate new guidance control instructions. The trajectory to be drilled in this embodiment can be obtained using conventional methods in the industry or using an artificial intelligence prediction model. It should be noted that the artificial intelligence prediction model must be used to determine the length of the trajectory to be drilled and the selection of the splicing point to ensure the output speed of the decision-making instructions.

[0040] For details, see Figure 7 , the artificial intelligence prediction module 240 is integrated in the central control unit 200, and is further configured as follows: The drilling operation data and rotary steering operation instruction data of the drilled wells are obtained, and abnormal data cleaning and data enhancement are performed; the drilling operation data are the well deviation a at the drill bit, the drill bit position b, the drill tool shape data c within n meters behind the drill bit, the average drilling pressure d within m meters in front of the drill bit, the average speed e, the average torque f, the drilling time g, the average displacement h, the well deviation change rate i and the change direction j within m meters in front of the drill bit; the rotary steering operation instruction data are the average value k and the average direction l of the resultant force of the steering tool within m meters in front of the drill bit; Constructing a sample set, using the cleaning and enhancement completed data to construct a sample set, using the drilling operation data as input, and using the rotary steering operation instruction data as output; Select a machine learning model with a regression algorithm, divide the constructed sample set into a training set, a validation set, and a test set, and perform training, validation, and testing on the machine learning model to obtain a mature control command artificial intelligence prediction model; The data related to the sample set is extracted from the actual drilling data and input into a mature artificial intelligence prediction model for control instructions to obtain new rotary steerable tool control instructions in real time.

[0041] In the construction method of the rotary steerable tool control command prediction model, data selection, preprocessing method and sample set establishment are the most critical and core innovative elements for the output of control commands, while there is diversity in the choice of model.

[0042] In terms of data selection, the drilling operation data only needs to select the well inclination a at the drill bit, the drill bit position b, the drill tool shape data c within n meters behind the drill bit, the average drilling pressure d within m meters in front of the drill bit, the average speed e, the average torque f, the drilling time g, the average displacement h, the well inclination change rate i and the change direction j within m meters in front of the drill bit, so as to improve the output speed of the control instructions.

[0043] In order to improve the accuracy of the control instructions, this embodiment also uses different methods to clean and enhance different drilling operation data.

[0044] Among them, the abnormal data cleaning method of the first type of drilling operation data, namely, the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit, is further configured as follows: Sorting the dynamic well deviation and dynamic azimuth data during the drilling operation of a single well in chronological order; A first sliding window with a width of W is formed with a drill rod length W; Assuming that the deflection capability of the steering tool is X° / Ym, calculate the median M1 of all the measured values ​​in the first sliding window, and obtain the determination intervals [M1-X*W / Y, M1+X*W / Y] of the well inclination a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit; The first sliding window is used to slide and judge and clean the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit in the entire well section, and the data outside the judgment interval is deleted.

[0045] Sorting the dynamic well deviation and dynamic azimuth data during the drilling operation of a single well in chronological order; A first sliding window with a width of W is formed with a drill rod length W. For example, in this embodiment, a conventional drill rod length of 10 meters is selected to form a 10*10 square sliding window; Assume that the deflection capability of the steering tool is X° / Ym, for example, the maximum deflection rate of the current downhole tool is 18° / 30m, calculate the median M1 of all the measured values ​​in the first sliding window, and obtain the determination intervals [M1-X*W / Y, M1+X*W / Y] of the well inclination a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit; The first sliding window is used to slide and judge and clean the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit in the entire well section, and the data outside the judgment interval is deleted.

[0046] The first sliding window is used to slide and judge and clean all the dynamic well inclination and dynamic azimuth data of the entire well section, and the dynamic well inclination and dynamic azimuth data outside the judgment interval of the well section corresponding to the sliding window are deleted, and the normal dynamic well inclination and dynamic azimuth data are retained, thereby improving the prediction accuracy; for the dynamic well inclination and dynamic azimuth data, since the drilling process is continuous, the changes of these two types of data are also continuous. According to this feature, the technical method uses the sliding window method to clean abnormal values, which can quickly and effectively complete the cleaning of abnormal data.

[0047] Then, the dynamic well deviation and dynamic azimuth data after cleaning are enhanced, and the enhanced dynamic well deviation and dynamic azimuth data are smoothed. Data enhancement is to fill in the missing values ​​and encrypt and interpolate the data after cleaning along the well depth, wherein the interpolation algorithm must meet the curve smoothness and have the characteristics of interpolation while providing data points, such as the Akima interpolation method. In this embodiment, all data enhancements use the cubic spline interpolation method.

[0048] Among them, the selection of interpolation spacing is the key. If the spacing is too small, the amount of data will increase and redundant data will be easily generated. If the spacing is too large, the data continuity will be poor, affecting the stability of the subsequent prediction model. In this embodiment, the change of the trajectory itself is a slow process. Taking the current maximum inclination rate of the downhole tool of 18° / 30m as an example, the average change rate per meter is 0.6°. Calculated at a well inclination change rate of 0.1°, the spacing is about 0.17m. In order to accurately describe the trajectory change, the data interpolation spacing can be taken as 0.1m, which is in line with engineering practice.

[0049] Among them, the abnormal data cleaning method of the second type of drilling operation data, that is, the well inclination change rate i and change direction j within m meters before the drill bit, is further configured as follows: According to different instructions, the magnitude and direction data of the deflection force corresponding to the instructions are filtered out; Arrange the deflection force magnitude and direction data according to the commanded start and end depths; Calculate the median M2 of all deflection force magnitude and direction data under the instruction condition; According to the fluctuation amplitude of the deflecting force and direction data under the command conditions, the threshold adjustment proportional coefficient r1 is determined, and the judgment interval of the well inclination change rate h and the change direction j is [M2- M2*r1, M2+ M2*r1], and the data outside the judgment interval is deleted.

[0050] For the guide force size and direction data, due to the large differences between different control instructions, these two data show discontinuous characteristics during the instruction switching process, and it is difficult to clean multiple instruction data at the same time, but within an instruction interval, these two types of numbers basically fluctuate within a smaller range.

[0051] Based on this, this solution processes the data of each instruction interval separately, that is, the starting and ending depths of each instruction are taken as a processing unit. In this unit, the median M2 of all measured guiding force magnitudes or orientations is used as the benchmark, and a suitable threshold is selected according to the actual fluctuation range of the guiding force magnitude and orientation. For example, the threshold can be set as the ratio of the benchmark value, and the proportional coefficient r1 can be adjusted according to the actual effect, generally set to r=30%. Then, when the measured value q exceeds the interval [M2- M2*r1, M2+ M2*r1], it is judged as abnormal and deleted, and the rest is judged as normal value.

[0052] After performing the same processing on the units corresponding to each instruction interval, the cleaning of this type of data is completed. The thresholds of the guide force size and direction in each processing unit can be flexibly adjusted according to the actual effect to achieve the optimal cleaning effect.

[0053] After cleaning, the data is enhanced in the same way as the first type of drilling operation design to complete missing values ​​and perform data encryption interpolation.

[0054] Among them, the abnormal data cleaning method of the third type of drilling operation data, namely the average drilling pressure d, average speed e, average torque f, drilling time g, and average displacement h within m meters before the drill bit, is further configured as follows: Arrange the drilling pressure, speed, torque, drilling time and displacement data in chronological order; A second sliding window with a width of W is formed with a drill rod length W; Calculate the median M3 of the corresponding data type in the second sliding window; According to the fluctuation range of different types of data, the threshold adjustment proportional coefficient r2 is determined. The judgment interval of drilling pressure, torque, speed, displacement and drilling time data is [M3- M3*r2, M3+ M3*r2], and the data outside the judgment interval is deleted.

[0055] For the data of drilling pressure, torque, rotation speed, displacement, drilling time, etc., they are all obtained by ground measurement, with few data anomalies, and the actual changes also have certain continuity characteristics. This embodiment adopts the sliding window method to clean abnormal values. The window length and benchmark are the same as the cleaning method of dynamic well deviation, that is, the window length W=10m, and the median M3 of the data in the window is used as the benchmark value. The selection of the threshold is the same as the cleaning method of the guide force size, and is set as the proportion of the benchmark value. The proportional coefficient r can be adjusted according to the actual effect, and can generally be set to r2=30%. When the measured value s exceeds the interval [M3-M3*r2, M3+M3*r2], it is judged as abnormal and deleted, and the rest is judged as normal value. According to this principle, the data abnormality cleaning of the entire well section is completed by sliding the window.

[0056] After data preprocessing is completed, a sample set is further established. The drilling operation data is used as input, and the rotary steering operation instruction data is used as output, and is organized into the format shown in Table 1 to facilitate the rapid output of control instructions later:

[0057] Table 1 The sample set construction in this step is highly related to the command output speed of the rotary steerable drilling operation process, and is one of the key technical points of the present invention.

[0058] The input of each sample is: well inclination a at the drill bit, drill bit position b, drill tool shape data c within n meters behind the drill bit, average values ​​c, d, e, f, g of drilling pressure, rotation speed, torque, drilling time and displacement within m meters in front of the drill bit, and well inclination change rate i and change direction j within m meters in front of the drill bit; the output of each sample is: the average value of the resultant force k and direction l of the guide tool within m meters in front of the drill bit.

[0059] Among them, in terms of the drill tool shape within n meters behind the drill bit, n meters are subdivided at intervals of 1m to form n drill tool shape data, so as to describe the drill tool shape more carefully. It should be noted that all parameters of each sample only correspond to one rotary steering control instruction and cannot span two instructions.

[0060] The method of dividing the constructed sample set into a training set, a validation set and a test set for training, validation and testing on the machine learning model is further configured as follows: the test set is the samples of one or several wells; the training set and validation set are samples obtained from other wells different from the test set; 80% of the data are extracted from samples of different wells in a random order as the training set, and the remaining 20% ​​of the data are used as the validation set.

[0061] In this embodiment, the sample set is divided into a training set, a validation set and a test set. To test the generalization of the prediction model, the test set can be selected as samples from one or several wells, while the training set and validation set are obtained from samples from other wells.

[0062] In terms of the division of training sets and validation sets, due to the particularity of drilling projects, the input and output parameters in samples of different operating sections are quite different. For example, the well inclination change rate in samples of the build-up section is usually larger, while the well inclination change rate in samples of the horizontal section is usually smaller. In order to make the training set cover samples with a wider parameter range, samples can be randomly selected from samples of multiple wells in a certain proportion (such as 80%) in a disordered manner as the training set, and the remaining 20% ​​of the samples are used as the validation set.

[0063] The construction principle of the training set, the validation set and the test set is closely related to the drilling operation process of the rotary steerable tool and is one of the key technical points of the present invention.

[0064] The machine learning regression model includes any one or more regression model combinations of SVM, BP, LSTM, and deep learning. For the training set and validation set constructed by the above method, a machine learning regression prediction method, such as SVM, BP, LSTM, deep learning, etc., is selected to train the training set composed of multiple samples. The parameter adjustment process in the specific training process needs to be carried out according to the characteristics of the specific algorithm. Through training, the control instruction prediction model of the rotary steering tool, that is, the combined force size and direction instruction prediction model of the steering tool, is obtained. The more wells that have been drilled for constructing the training set, the more samples there are, and the more accurate the prediction model is. In addition, the prediction model can be dynamically updated, that is, the drilled depth data of the well being drilled can be added to the training set to predict the control instructions in the subsequent drilling process of the well.

[0065] Specifically, the control instruction prediction model is in implicit form and can be represented by f, that is, (k, l) = f (a, b, c, d, e, f, g, h, i, j). Taking the LSTM machine learning prediction method as an example, the training set constructed by the drilling data is trained, and the training model is optimized using the validation set to obtain the optimal model f1 under the current sample. The prediction example is shown in Table 2 below:

[0066] Table 2 Therefore, in the actual drilling process, the actual drilling data is organized into input parameters as shown in Table 1 and input into the prediction model, so that the best control instructions can be obtained in real time, avoiding the problem of inaccurate control instructions caused by insufficient human experience. New control instructions can be adjusted according to the measured geological data, thereby improving drilling efficiency and wellbore trajectory quality.

[0067] In this embodiment, the downhole self-feedback method is improved, the original PID self-feedback control is abandoned, and a new artificial intelligence prediction module 240 is used to quickly and accurately complete measurement data acquisition and trajectory comparison in the downhole, and rely on artificial intelligence prediction to output high-precision control instructions, thereby improving trajectory control quality, drilling quality and efficiency.

[0068] At the same time, a design trajectory correction module 220 is introduced between the actual drilling trajectory and the designed trajectory. When the geological deviation is too large, the first trajectory to be drilled m meters before the drill bit of the actual drilling trajectory is redesigned according to the geological guidance trajectory adjustment principle. The first trajectory to be drilled m meters before is generated by an artificial intelligence prediction model, and the first trajectory to be drilled is spliced ​​on the first designed trajectory to form a second designed trajectory. When the geological deviation meets the requirements, the actual drilling trajectory and the designed trajectory are compared again, so that the later control instructions can adjust the guidance control instructions in real time according to geological factors, thereby improving the control accuracy of downhole drilling and ensuring the quality, efficiency and safety of the drilling trajectory.

[0069] Second embodiment A closed-loop control method for underground geosteering, see Figure 8 and Fig. 9 , the control method comprises: S1: using the information acquisition unit 100 to respectively acquire drilling tool posture data and geological parameter data; S2: The actual drilling trajectory calculation module 210 of the central control unit 200 obtains the drilling tool posture data to generate the actual drilling trajectory; the design trajectory correction module 220 obtains the geological parameter data and calculates the deviation Δ0 between the geological parameter data and the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to automatically design the first trajectory to be drilled m meters ahead of the drill bit, and the first design trajectory is updated according to the first trajectory to be drilled to obtain the second design trajectory; when the geological deviation Δ0 is less than or equal to the threshold ε0, the self-closed-loop control program module 230 compares the continuous actual drilling trajectory with the first / second design trajectory to calculate the geometric deviation Δ1; when the deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters ahead of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely connected; the first trajectory to be drilled or the second trajectory to be drilled is input into the control instruction artificial intelligence prediction model, and a new rotary steering tool control instruction is output; S3: The execution unit 300 performs the drilling operation according to the rotary steerable tool control instruction.

[0070] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0071] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0072] The units described as separate components may or may not be physically separated. As a unit, it can be appreciated by those skilled in the art that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0073] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0074] If the integrated unit is implemented in the form of 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 the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a grid device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0075] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A downhole geosteering closed-loop control system, characterized in that: The control system includes: An information acquisition unit, wherein the downhole information acquisition unit comprises a plurality of dynamic attitude measurement sensors arranged at intervals and a plurality of LWD short sections arranged at intervals from the dynamic attitude measurement sensors, and is used to respectively acquire drilling tool attitude data and geological parameter data; The central control unit is communicatively connected with the information acquisition unit; the central control unit is provided with an actual drilling trajectory calculation module, a design trajectory correction module, a self-closed-loop control program module, an artificial intelligence prediction module and a communication module, the actual drilling trajectory calculation module obtains the drilling tool posture data of the dynamic posture measurement sensor to fit and generate the actual drilling trajectory; the design trajectory correction module obtains the geological parameter data of the LWD short section, and calculates the deviation Δ0 between the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to automatically design the first trajectory to be drilled m meters ahead of the drill bit, and the first design trajectory is updated according to the first trajectory to be drilled to obtain the second design trajectory; the self-closed-loop control program module, when the geological deviation Δ0 is less than or equal to the threshold ε0, compares the continuous actual drilling trajectory with the first / second design trajectory, and calculates the geometric deviation Δ1; when the geometric deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters ahead of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely engaged; The artificial intelligence prediction module is configured to: input the first trajectory to be drilled or the second trajectory to be drilled into the control instruction artificial intelligence prediction model, and output a new rotary steerable tool control instruction; The execution unit is communicatively connected with the central control unit; the execution unit obtains the control instruction of the rotary steering tool to control the rib-guided drilling; the information acquisition unit continuously obtains the measurement data of the execution unit to form an underground geological steering closed loop.

2. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The geological parameter data at least include azimuthal gamma, azimuthal resistivity, while-drilling sonic wave, while-drilling neutron density, and while-drilling nuclear magnetic parameters.

3. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The artificial intelligence prediction module is integrated into the central control unit and is further configured as follows: Acquire the drilling operation data and rotary steering operation instruction data of the drilled well, and perform abnormal data cleaning and data enhancement; the drilling operation data are the well deviation a at the drill bit, the drill bit position b, the drill tool shape data c within n meters behind the drill bit, the average drilling pressure d within m meters before the drill bit, the average speed e, the average torque f, the drilling time g, the average displacement h, the well deviation change rate i and the change direction j within m meters before the drill bit; the rotary steering operation instruction data are the average value k and the average direction l of the resultant force of the guide tool within m meters before the drill bit; Constructing a sample set, constructing a sample set with the cleaning enhancement completed data, taking the drilling operation data as input, and taking the rotary steering operation instruction data as output; Select a machine learning model with a regression algorithm, divide the constructed sample set into a training set, a validation set, and a test set, and perform training, validation, and testing on the machine learning model to obtain a mature control command artificial intelligence prediction model; The data related to the sample set is extracted from the actual drilling data and input into the mature control instruction artificial intelligence prediction model to obtain new rotary steering tool control instructions in real time.

4. The downhole geosteering closed-loop control system according to claim 3, characterized in that: The abnormal data cleaning method of the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit is further configured as follows: Sorting the dynamic well inclination and dynamic azimuth data during the drilling operation of a single well in chronological order; A first sliding window with a width of W is formed with a drill rod length W; Assuming that the deflection capability of the steering tool is X° / Ym, calculate the median M1 of all the measured values ​​in the first sliding window, and obtain the determination intervals [M1-X*W / Y, M1+X*W / Y] of the well inclination a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit; The first sliding window is used to slide and judge and clean the well deviation a at the drill bit, the drill bit position b, and the drilling tool shape data c within n meters behind the drill bit in the entire well section, and the data outside the judgment interval is deleted.

5. The downhole geosteering closed-loop control system according to claim 3, characterized in that: The abnormal data cleaning method of the well deviation change rate i and change direction j within m meters before the drill bit is further configured as follows: According to different instructions, the magnitude and direction data of the deflecting force corresponding to the instructions are filtered out; Arranging the deflecting force magnitude and direction data according to the starting and ending depths of the instructions; Calculate the median M2 of all deflection force magnitude and direction data under the instruction condition; According to the fluctuation amplitude of the deflecting force size and direction data under the command conditions, the threshold adjustment proportional coefficient r1 is determined, and the determination interval of the well inclination change rate h and the change direction j is [M2- M2*r1, M2+ M2*r1], and the data outside the determination interval is deleted.

6. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The abnormal data cleaning method of the average drilling pressure d, the average rotation speed e, the average torque f, the drilling time g, and the average displacement h within the first m meters of the drill bit is further configured as follows: Arrange the drilling pressure, rotation speed, torque, drilling time and displacement data in chronological order; A second sliding window with a width of W is formed with a drill rod length W; Calculate the median M3 of the corresponding data type in the second sliding window; According to the fluctuation range of different types of data, the threshold adjustment ratio coefficient r2 is determined. The determination interval of the drilling pressure, torque, speed, displacement and drilling time data is [M3- M3*r2, M3+ M3*r2], and the data outside the determination interval is deleted.

7. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The data enhancement method is further configured to: perform missing value completion and data encryption interpolation on the data cleaned along the well depth to make the data complete and smooth.

8. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The method of dividing the constructed sample set into a training set, a validation set and a test set for training, validation and testing on a machine learning model is further configured as follows: the test set is samples from one or several wells; the training set and validation set are samples obtained from other wells different from the test set; 80% of the data are extracted from samples from different wells in a random order as the training set, and the remaining 20% ​​of the data are used as the validation set.

9. The downhole geosteering closed-loop control system according to claim 1, characterized in that: The method for generating the second designed trajectory is further configured as follows: when the geological deviation Δ0 is greater than the threshold ε0, the artificial intelligence prediction model is used to automatically generate the first trajectory to be drilled m meters ahead of the drill bit, and the first trajectory to be drilled is spliced ​​onto the first designed trajectory to ensure that the well inclination and azimuth of the junction point or junction section of the first designed trajectory and the first trajectory to be drilled are exactly the same.

10. A downhole geosteering closed-loop control method, characterized in that: The control method comprises: The information acquisition unit is used to obtain the drilling tool posture data and the geological parameter data respectively; The actual drilling trajectory calculation module of the central control unit obtains the drilling tool posture data to generate an actual drilling trajectory; the design trajectory correction module obtains the geological parameter data and calculates the deviation Δ0 between the geological parameter data and the geological requirements; when the geological deviation Δ0 is greater than the threshold ε0, according to the geological guidance trajectory adjustment principle, the artificial intelligence prediction model is used to automatically design the first trajectory to be drilled m meters ahead of the drill bit, and the first design trajectory is updated according to the first trajectory to be drilled to obtain the second design trajectory; when the geological deviation Δ0 is less than or equal to the threshold ε0, the self-closed-loop control program module compares the continuous actual drilling trajectory with the first / second design trajectory to calculate the geometric deviation Δ1; when the deviation Δ1 is greater than the threshold ε1, the second trajectory to be drilled m meters ahead of the drill bit is recalculated according to the first / second design trajectory, and the connection between the second trajectory to be drilled and the first / second design trajectory is completely engaged; the first trajectory to be drilled or the second trajectory to be drilled is input into the control instruction artificial intelligence prediction model, and a new rotary steering tool control instruction is output; The execution unit executes the drilling operation according to the control instruction of the rotary steering tool.

Images