CN119102579A - A closed-loop control method and system for geosteering between wells and ground - Google Patents

Abstract

The invention provides a method and a system for controlling geosteering closed loop between wells, which can adaptively adjust a design track according to geological information of real-time drilling and ensure drilling safety; the artificial intelligent prediction technology is applied to well track and instruction prediction, accuracy and response speed of a control instruction are improved, the control method comprises the steps of obtaining underground continuous attitude measurement data and LWD measurement parameter data, continuously comparing geological deviation between the LWD measurement parameter data and geological requirements along a real drilling track, automatically designing to obtain a second designed track when the geological deviation is larger than a threshold epsilon ₀, comparing the continuous real drilling track with the first/second designed track when the geological deviation is smaller than or equal to the threshold epsilon ₀, calculating geometric deviation, calculating a second track to be drilled according to the first/second designed track when the geometric deviation is larger than the threshold epsilon ₁, and inputting the track to be drilled and drilling operation parameters corresponding to the track to be drilled into a control instruction prediction model to obtain a new control instruction.

Description

Inter-well geosteering closed-loop control method and system

Technical Field

The invention relates to the field of petroleum and natural gas exploration and development drilling, in particular to a method and a system for inter-well geosteering closed-loop control.

Background

In rotary steerable drilling, the real wellbore trajectory is typically calculated by methods well known in the art based on downhole measurement while drilling data in combination with the depth of penetration of the surface drill string. And then taking the deviation of the real well drilling track on the ground and the expected well drilling track as the basis of track adjustment decision.

Because uncertainty of geological factors has a great influence on the safety of drilling engineering, detailed geological condition investigation is needed before drilling construction, so that scientificity and rationality of drilling safety construction are ensured, and important reference value is provided for exploration and development to form a design track. However, during the drilling process, when the geological condition of the real drilling track and the geological condition on the point position corresponding to the design track deviate, the drilling is performed based on the control instruction made by the preset design track, and the phenomena of drilling sticking, lost circulation, overflow and the like may occur, so that the drilling safety is directly affected.

Meanwhile, during the self-feedback control of the control command, how to timely feed back the accurate control command to the underground is important to improve the control precision and the track quality of the track, and some common self-feedback control methods in industry are difficult to meet the requirements.

The method for predicting the guiding instruction of the pushing rotary guiding tool in real time in Chinese patent application No. CN202310223456.3 can predict the guiding instruction in the next step by feeding back inclinometry data in real time, so that the feedback process of the instruction is optimized. The method comprises the steps of calculating the well deviation change rate and the azimuth change rate of a well depth interval implemented by each guide instruction by utilizing the inclinometry data fed back by the pushing type rotary guide tool in real time and the guide instruction implemented in the guide tool operation process, and then providing the guide force percentage and the guide tool face angle required by the rotary guide tool based on a guide instruction model and a target point drilled in the next step.

The rate of change of well inclination and the rate of change of azimuth are important reference data in the inclinometry calculation of the well trajectory. In current calculation methods, the well deviation angle change and the azimuth angle change of the wellbore section are used to calculate the well deviation change rate and the azimuth change rate. However, on a wellbore interval, if the well inclination or azimuth does not change linearly with the well depth, the well inclination or azimuth rate of change is not constant. And when the well section is longer, the true well deviation change rate and azimuth change rate have non-negligible deviation from the constant well deviation change rate and azimuth change rate.

Therefore, by adopting the method for predicting the guiding instruction of the pushing rotary guiding tool in real time, the force and angle adjustment from the drill bit to the target point can be directly obtained through the well deviation change rate and the azimuth change rate, the defect of low accuracy of the intelligently acquired control instruction can be directly caused, and the control accuracy and the track quality of a new track in the adjustment process are lower.

Disclosure of Invention

The invention aims to provide a method and a system for controlling geosteering closed loop between wells, which can adaptively adjust and update a design track according to real-time drilling conditions so as to ensure the safety of the drilling process, and simultaneously apply a novel artificial intelligent prediction technology to the well track and control instruction prediction so as to improve the accuracy and response speed of the control instruction, thereby improving the accuracy, quality and efficiency of track control.

Embodiments of the present invention are implemented as follows:

a method of inter-well geosteering closed loop control, the control method comprising:

acquiring underground continuous attitude measurement data and LWD measurement parameter data;

Fitting according to the continuous attitude measurement data to obtain a continuous real drilling track;

Continuously comparing the LWD measurement parameter data with the geological deviation delta ₀ along the real drilling track, when the geological deviation delta ₀ is larger than a threshold epsilon ₀, automatically designing a first track to be drilled of m meters in front of the drill bit by using an artificial intelligent prediction model according to a geosteering track adjustment principle, updating the first design track according to the first track to be drilled to obtain a second design track, and when the geological deviation delta ₀ is smaller than or equal to the threshold epsilon ₀, comparing the continuous real drilling track with the first/second design tracks to calculate the geometric deviation delta ₁;

When the geometric deviation delta ₁ is larger than the threshold epsilon ₁, recalculating a second track to be drilled of m meters in front of the drill bit according to the first/second design tracks, wherein the joint of the second track to be drilled and the first/second design tracks is completely jointed;

When the geometric deviation delta ₁ is smaller than or equal to the threshold epsilon ₁, the drilling parameters and the control instructions are kept unchanged, and the control instructions are issued to the rotary guiding tool;

Inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into a rotary guiding tool control instruction prediction model to obtain a new control instruction, issuing the control instruction to the rotary guiding tool, and continuously obtaining continuous attitude measurement data and LWD measurement parameter data.

In a preferred embodiment of the present invention, the downhole continuous attitude measurement data and the LWD measurement parameter data are acquired by a continuous attitude measurement sensor arranged on the rotary guiding tool and a plurality of LWD sub-sections arranged on the cabled drill pipe, and the LWD sub-sections are arranged at intervals.

In a preferred embodiment of the invention, the LWD nipple measures at least azimuthal gamma or azimuthal resistivity.

In a preferred embodiment of the present invention, the method for constructing a prediction model of a control command of a rotary steerable tool includes:

The method comprises the steps of acquiring drilling operation data and rotation guiding operation instruction data of a drilled well, and cleaning and enhancing abnormal data, wherein the drilling operation data comprise drilling deviation a at a drill bit, azimuth b at the drill bit, drilling tool form data c in n meters behind the drill bit, average value d of drilling pressure in m meters in front of the drill bit, average value e of rotation speed, average value f of torque, g when drilling, average value h of displacement, well deviation change rate i and change direction j in m meters in front of the drill bit, and the rotation guiding operation instruction data comprise resultant force magnitude average value k and direction average value l of guiding tools in m meters in front of the drill bit;

Constructing a sample set, namely constructing the sample set by using the data subjected to cleaning enhancement, taking drilling operation data as input, and taking rotation guiding operation instruction data as output;

selecting a machine learning model with a regression algorithm, dividing a constructed sample set into a training set, a verification set and a test set, and training, verifying and testing on the machine learning model to obtain a mature control instruction artificial intelligent prediction model.

In a preferred embodiment of the present invention, the method for cleaning abnormal data of drilling tool morphology data c in n meters behind the drill bit, including well deviation a at the drill bit, azimuth b at the drill bit, and the like, is further configured to:

sequencing dynamic well deviation and dynamic azimuth data in the single well drilling operation process according to time sequence;

forming a first sliding window with the width W by the length W of the drill rod;

Setting the deflecting capability of a guiding tool as X degrees/Ym, and calculating the median M ₁ of all measured values in a first sliding window to obtain a well deflection a at the drill bit, a direction b at the drill bit and a judging section [ M ₁-X*W/Y,M₁ +X W/Y ] of drilling tool form data c in n meters behind the drill bit;

And (3) using a first sliding window to slide, judge and clean drilling tool form data c in drill bit position deviation a, drill bit position b and drill bit rear n meters of the whole well section, and deleting data outside the judging section.

In a preferred embodiment of the present invention, the method for cleaning abnormal data of the well deviation change rate i and the change direction j in m meters in front of the drill bit is further configured to:

Screening out the data of the deflecting force and the direction corresponding to the instructions according to the different instructions;

The size and direction data of the deflecting force are arranged according to the starting depth and the ending depth of the instruction;

Calculating the median value M ₂ of all the oblique force magnitude and direction data under the instruction condition;

And determining a threshold adjustment proportionality coefficient r1 according to the magnitude of the deflecting force and the fluctuation amplitude of the direction data under the instruction condition, wherein a judgment section of the well deflection change rate h and the change direction j is [ M ₂- M₂*r1,M₂+ M₂ x r1], and deleting the data outside the judgment section.

In a preferred embodiment of the present invention, the method for cleaning abnormal data of the average value d of weight on bit, the average value e of rotation speed, the average value f of torque, the average value g of drilling time and the average value h of displacement in m meters before the drill bit is further configured to:

Arranging the data of the bit pressure, the rotating speed, the torque, the drilling time and the displacement according to the time sequence;

Forming a second sliding window with the width W by the length W of the drill rod;

Calculating a median value M ₃ of the corresponding data types in the second sliding window;

And determining a threshold adjustment proportionality coefficient r2 according to the fluctuation amplitude of different types of data, wherein the judgment interval of the bit pressure, the torque, the rotating speed, the displacement and the drilling time data is [ M ₃- M₃*r2,M₃+ M₃ x r2], and deleting the data outside the judgment interval.

In a preferred embodiment of the present invention, the data enhancement method is further configured to complete and smooth the data by performing blank value completion and data encryption interpolation on the data cleaned along the well depth.

In the preferred embodiment of the invention, the method for dividing the constructed sample set into the training set, the verification set and the test set to train, verify and test on the machine learning model is further configured to obtain the test set as samples of one or more wells, obtain the training set and the verification set as samples obtained from other wells different from the test set, and extract 80% of data from the samples of the different wells in an out-of-order manner to serve as the training set, and obtain the remaining 20% of data as the verification set.

A interwell geosteering closed loop control system, the control system comprising:

The measuring data acquisition module is used for acquiring underground continuous attitude measuring data and LWD measuring parameter data;

The real drilling track calculation module is used for fitting according to the continuous attitude measurement data to obtain a continuous real drilling track;

When the geological deviation delta ₀ is larger than a threshold epsilon ₀, automatically designing a first track to be drilled of m meters in front of the drill bit by using an artificial intelligent prediction model according to a geosteering track adjustment principle, and updating the first design track according to the first track to be drilled to obtain a second design track;

The geometrical deviation calculation module is used for comparing the continuous real drilling track with the first/second design track when the geological deviation delta ₀ is smaller than or equal to a threshold epsilon ₀, calculating the geometrical deviation delta ₁, calculating the second track to be drilled of m meters in front of the drill bit according to the first/second design track when the geometrical deviation delta ₁ is larger than the threshold epsilon ₁, completely jointing the joint of the second track to be drilled and the first/second design track when the geometrical deviation delta ₁ is smaller than or equal to the threshold epsilon ₁, keeping drilling parameters and control instructions unchanged, enabling the control instructions to be sent to the rotary guiding tool, and continuously acquiring continuous attitude measurement data and LWD measurement parameter data;

the rotary guiding tool control instruction prediction module is used for inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into the rotary guiding tool control instruction prediction model to obtain new control instructions, issuing the control instructions to the rotary guiding tool, and continuously obtaining continuous attitude measurement data and LWD measurement parameter data.

The embodiment of the invention has the beneficial effects that:

1. When the geological deviation is overlarge, the first track to be drilled of the drill bit of the real drilling track is redesigned according to the geological guiding track adjustment principle, the first track to be drilled of the drill bit of the real drilling track of the front m meters is generated by an artificial intelligent prediction model, the first track to be drilled is spliced on the first design track to form a second design track, and when the geological deviation accords with the rule, the real drilling track and the design track are compared, so that a later control instruction can be subjected to guide control instruction adjustment in real time according to geological factors, the control precision of underground drilling is improved, and the quality and efficiency of the drilling track are ensured;

2. the second design track and the first m meters of the track to be drilled are determined by an artificial intelligent prediction model, the intelligent design shortens the track design time and the decision output time of the control instruction, so that the working efficiency is improved;

3. in the control instruction giving link, an artificial intelligent prediction model is adopted and is used for outputting the control instruction of the rotary guiding tool, so that the intelligent level of the rotary guiding tool can be improved, the track control precision, quality and efficiency are improved, and the error condition of the control instruction is reduced.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a method of inter-well geosteering closed loop control in accordance with an embodiment of the present invention;

FIG. 2 is a schematic diagram of a downhole drilling assembly connection according to an embodiment of the present invention;

FIG. 3 is a flowchart of a method for constructing an artificial intelligence prediction module for control instructions according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of the main functional units of the inter-well geosteering closed-loop control system according to an embodiment of the present invention.

The icons comprise a continuous attitude measurement sensor 001, an LWD short section 002, a cabled drill rod 003, a measurement data acquisition module M1, a real drilling track calculation module M2, a geological deviation calculation module M3, a geometric deviation calculation module M4 and a rotary guiding tool control instruction prediction module M5.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

It should be noted that like reference numerals and letters refer to like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.

The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the application. One skilled in the relevant art will recognize, however, that the application may be practiced without one or more of the specific details, or with other methods, components, devices, steps, etc. In other instances, well-known methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the application.

The block diagrams depicted in the figures are merely functional entities and do not necessarily correspond to physically separate entities. That is, the functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.

The flow diagrams depicted in the figures are exemplary only, and do not necessarily include all of the elements and operations/steps, nor must they be performed in the order described. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the order of actual execution may be changed according to actual situations.

First embodiment

Referring to fig. 1, the present embodiment provides a method for inter-well geosteering closed-loop control, the method comprising:

s1, acquiring underground continuous attitude measurement data and LWD measurement parameter data;

S2, fitting according to continuous attitude measurement data to obtain a continuous real drilling track, wherein the continuous real drilling track is obtained by calculating the continuous attitude measurement data by utilizing a track calculation method commonly used in the industry, such as a minimum curvature method and the like;

S3, continuously comparing the LWD measurement parameter data with the geological deviation delta 0 between geological requirements along the real drilling track, when the geological deviation delta 0 is larger than a threshold epsilon 0, automatically designing a first track to be drilled of m meters in front of the drill bit by using an artificial intelligent prediction model according to a geosteering track adjustment principle, and updating the first design track according to the first track to be drilled to obtain a second design track;

S4, when the geometric deviation delta 1 is larger than a threshold epsilon 1, recalculating a second track to be drilled of m meters in front of the drill bit according to the first/second design tracks, wherein the joint of the second track to be drilled and the first/second design tracks is completely jointed, when the geometric deviation delta 1 is smaller than or equal to the threshold epsilon 1, keeping drilling parameters and control instructions unchanged, and enabling the control instructions to be sent to a rotary guiding tool, and continuously acquiring continuous attitude measurement data and LWD measurement parameter data to form closed-loop control;

S5, inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into a rotary guiding tool control instruction prediction model to obtain a new control instruction, and issuing the control instruction to the rotary guiding tool to continuously obtain continuous attitude measurement data and LWD measurement parameter data.

In this embodiment, please refer to fig. 2, the downhole continuous attitude measurement data and LWD measurement parameter data are respectively obtained by 1 continuous attitude measurement sensor 001 arranged on the rotary guiding tool and a plurality of multi-type LWD pup joints 002 arranged on the cabled drill pipe, the various types of LWD pup joints 002 are arranged at intervals, and the more the types of LWD pup joints 002 are, the better. LWD nipple 002 measures at least azimuthal gamma or azimuthal resistivity. The real-time uploading and the instruction downloading of the underground measurement information are realized between the measurement sensor and the ground system/inter-well geosteering closed-loop control system through the cabled drill rod 003.

Because in the drilling operation, geological factors have great influence on the safety of drilling engineering, geological prediction misalignment can lead to obvious promotion of the occurrence probability of drilling engineering accidents, and great loss is likely to be caused to staff and oilfield enterprises.

When the drilling engineering design is carried out, geological analysis work of the exploration engineering is carried out deeply according to actual conditions, then a preset design track is compared with an actual drilling track before the drilling engineering is carried out, whether track adjustment is needed or not is determined according to the comparison conditions, and when errors exist between geological conditions in the actual drilling process and geological data acquired in advance, a large error exists in the later drilling track, so that the whole drilling operation is affected.

S3 in the embodiment, geological deviation amount calculation is newly added before track comparison, so that when the design track can be matched with the later track according to the geological deviation result, the first design track which is designed in advance or the second design track which is redesigned again according to geological data is used, and the well drilling operation is ensured to be smoother, safer and more efficient.

The operation is described in fig. 1:

s31, continuously differencing LWD measurement parameter data and geological requirement data along the real drilling track to obtain geological deviation delta 0;

S32, when the geological deviation delta 0 is larger than a threshold epsilon 0, designing a first track to be drilled of m meters in front of the drill bit by using an artificial intelligent prediction model according to a geosteering track adjustment principle, and updating the first design track according to the first track to obtain a second design track, wherein the second design track is a starting point of the first track to be drilled which is jointed with an end point of the real track, the end point of the first track to be drilled is jointed with a joint point or a joint section of the first design track, and the spatial position and the gesture of each joint point or joint section are completely the same;

And S33, when the geological deviation delta 0 is smaller than or equal to the threshold epsilon 0, further track judgment is carried out, namely, the continuous real drilling track is compared with the first/second design track, and the geometric deviation delta 1 is calculated.

And then, track judgment is carried out, in the specific step S4, when the geometric deviation delta 1 exceeds a given threshold epsilon 1, the real drilling track is required to be adjusted, firstly, the track to be drilled of m meters in front of the drill bit is redesigned by the artificial intelligent prediction model, and the track to be drilled and the designed track are required to be completely jointed, namely, the spatial position and the gesture of the joint are completely consistent. Specifically, before a time t for track determination, all geological deviation amounts delta 0 in a time t period are smaller than or equal to a threshold epsilon 0, comparing and determining the first designed track with the real drilling track, and when the geological deviation amount delta 0 at any moment in the time t period is larger than the threshold epsilon 0, comparing and determining the second designed track with the real drilling track;

When the calculation of the track to be drilled exists, the generated track to be drilled and drilling operation data are input into a rotary guiding tool control instruction prediction model to generate a new guiding control instruction. The trajectory to be drilled in this embodiment may be obtained in a manner conventional in the art, or may be obtained using a rotary steerable tool control command prediction model. It should be noted that, in selecting the length of the trajectory to be drilled, a rotation guiding tool control command prediction model must be used to determine the output speed of the decision command.

Specifically, referring to fig. 3, the method for constructing the prediction model of the control instruction of the rotary steerable tool in this embodiment includes the following steps:

The method comprises the steps of firstly, acquiring drilling operation data and rotation guiding operation instruction data of a drilled well, and cleaning and enhancing abnormal data, wherein the drilling operation data comprise drilling deviation a at a drill bit, azimuth b at the drill bit, drilling tool form data c in n meters behind the drill bit, average drilling pressure d in m meters in front of the drill bit, average rotating speed e, average torque f, g during drilling, average displacement h, well deviation change rate i and change direction j in m meters in front of the drill bit, and the rotation guiding operation instruction data comprise resultant force magnitude average k and direction average l of guiding tools in m meters in front of the drill bit;

Constructing a sample set, namely constructing the sample set by using the data subjected to cleaning enhancement, taking drilling operation data as input and taking rotation guiding operation instruction data as output;

And thirdly, selecting a machine learning model with a regression algorithm, dividing the constructed sample set into a training set, a verification set and a test set, and training, verifying and testing on the machine learning model to obtain a mature control instruction artificial intelligent prediction model.

In the construction method of the rotary guiding tool control instruction prediction model, the selection and preprocessing modes of data and the establishment of a sample set have the most key and core innovation elements for the output of control instructions, and the selection of the model has diversity.

In terms of data selection, drilling operation data only need to select well deviation a at a drill bit, azimuth b at the drill bit, drilling tool form data c in n meters behind the drill bit, average value d of drilling pressure in m meters in front of the drill bit, average value e of rotating speed, average value f of torque, g during drilling, average value h of displacement, well deviation change rate i in m meters in front of the drill bit and change direction j, so that the output speed of a control instruction can be improved.

In order to improve the accuracy of the control command, the embodiment also adopts different modes to clean and enhance different drilling operation data.

The method for cleaning abnormal data of drilling operation data of the first type, namely drilling deviation a at the drill bit, azimuth b at the drill bit and drilling tool form data c in n meters behind the drill bit, is further configured to:

sequencing dynamic well deviation and dynamic azimuth data in the single well drilling operation process according to time sequence;

forming a first sliding window with the width W by the length W of the drill rod;

setting the deflecting capability of a guiding tool as X degrees/Ym, and calculating the median M1 of all measured values in a first sliding window to obtain a well deviation a at the drill bit, an azimuth b at the drill bit and a judging section [ M1-X W/Y, M1+X W/Y ] of drilling tool form data c in n meters behind the drill bit;

And (3) using a first sliding window to slide, judge and clean drilling tool form data c in drill bit position deviation a, drill bit position b and drill bit rear n meters of the whole well section, and deleting data outside the judging section.

Sequencing dynamic well deviation and dynamic azimuth data in the single well drilling operation process according to time sequence;

forming a first sliding window with a width W by using a length W of the drill rod, for example, selecting a conventional drill rod with a length of 10 meters in the embodiment to form a square sliding window with a length of 10 x 10;

Setting the deflecting capability of a guiding tool to be X degrees/Ym, for example, the maximum deflecting rate of the current downhole tool to be 18 degrees/30M, calculating the median M ₁ of all measured values in a first sliding window, and obtaining a judging section [ M ₁-X*W/Y,M₁ +X W/Y ] of drilling tool form data c in a drill bit position a, a drill bit position b and n meters behind the drill bit;

And (3) using a first sliding window to slide, judge and clean drilling tool form data c in drill bit position deviation a, drill bit position b and drill bit rear n meters of the whole well section, and deleting data outside the judging section.

The method comprises the steps of judging and cleaning all dynamic well deviation and dynamic azimuth data of a whole well section by using a first sliding window, deleting the dynamic well deviation and dynamic azimuth data of the well section corresponding to the sliding window outside a judging section, and reserving normal dynamic well deviation and dynamic azimuth data so as to improve prediction precision, wherein the dynamic well deviation and dynamic azimuth data are continuous in the drilling process, so that the two types of data are continuously changed.

And then, enhancing the cleaned dynamic well deviation and dynamic azimuth data, and smoothing the enhanced dynamic well deviation and dynamic azimuth data. And data enhancement, namely performing null value completion and data encryption interpolation on the data cleaned along the well depth, wherein an interpolation algorithm is required to meet the characteristics of curve smoothness and interpolation while providing data points, such as an Akima interpolation method and the like. In this embodiment, a cubic spline interpolation method is used for all data enhancement.

The selection of the interpolation interval is critical, if the interval is too small, the data volume is increased, redundant data is easy to generate, and if the interval is too large, the data continuity is poor, so that the stability of a subsequent prediction model is affected. In this embodiment, the track change is a slow process, taking the current maximum slope of the downhole tool of 18 °/30m as an example, the average change rate per meter is 0.6 °, and the pitch is about 0.17m calculated according to the well deviation change rate of 0.1 °, so that the interpolation pitch of the available data is 0.1m for fine description of the track change, which accords with engineering practice.

Wherein the second type of drilling operation data, namely the abnormal data cleaning method of the well deviation change rate i and the change direction j in m meters in front of the drill bit, is further configured to:

Screening out the data of the deflecting force and the direction corresponding to the instructions according to the different instructions;

The size and direction data of the deflecting force are arranged according to the starting depth and the ending depth of the instruction;

Calculating the median value M ₂ of all the oblique force magnitude and direction data under the instruction condition;

And determining a threshold adjustment proportionality coefficient r1 according to the magnitude of the deflecting force and the fluctuation amplitude of the direction data under the instruction condition, wherein a judgment section of the well deflection change rate h and the change direction j is [ M ₂- M₂*r1,M₂+ M₂ x r1], and deleting the data outside the judgment section.

For the data of the magnitude and the direction of the guiding force, the difference between different control instructions is large, so that the two data have discontinuous characteristics in the instruction switching process, and a plurality of instruction data are difficult to clean at the same time, but the two types of data basically fluctuate in a small section in one instruction section.

Based on this, the scheme processes the data of each instruction interval separately, namely, takes the starting depth and the ending depth of each instruction as a processing unit. In the unit, a proper threshold value is selected according to the actual fluctuation amplitude of the guiding force magnitude and the direction by taking the median value M2 of all actually measured guiding force magnitudes or directions as a reference. For example, the threshold value may be set to a ratio of the reference value, and the ratio coefficient r1 may be adjusted according to the actual effect, and may be generally set to r=30%, and when the measured value q exceeds the interval [ M2-M2 r1, m2+m2 r1], it is determined that the measured value q is abnormal, the deletion process is performed, and the rest is determined as a normal value.

After the same processing is carried out on the units corresponding to each instruction interval, the data are cleaned, wherein the threshold value of the guiding force and the direction in each processing unit can be flexibly adjusted according to the actual effect, so that the optimal cleaning effect is achieved.

And after the cleaning is finished, performing blank value complementation and data encryption interpolation on the data by adopting a data enhancement mode which is designed to be the same as the first type of drilling operation.

The third type of drilling operation data, namely an abnormal data cleaning method of the average value d of the bit weight in m meters in front of the drill bit, the average value e of the rotating speed, the average value f of the torque, the average value g of the drilling time and the average value h of the displacement, is further configured to:

Arranging the data of the bit pressure, the rotating speed, the torque, the drilling time and the displacement according to the time sequence;

Forming a second sliding window with the width W by the length W of the drill rod;

Calculating a median value M ₃ of the corresponding data types in the second sliding window;

And determining a threshold adjustment proportionality coefficient r2 according to the fluctuation amplitude of different types of data, wherein the judgment interval of the bit pressure, the torque, the rotating speed, the displacement and the drilling time data is [ M ₃- M₃*r2,M₃+ M₃ x r2], and deleting the data outside the judgment interval.

The data of the weight on bit, the torque, the rotating speed, the displacement, the drilling time and the like are all obtained by ground measurement, the data are few in abnormality, and the actual change also has a certain characteristic of continuity. In this embodiment, a sliding window method is used to clean the abnormal value, the window length and the reference are the same as the cleaning mode of the dynamic well deviation, i.e. the window length w=10m, the median value M3 of the data in the window is taken as the reference value, the threshold value is selected to be the same as the cleaning mode of the guiding force, the ratio coefficient r can be adjusted according to the actual effect, and can be generally set to r2=30%, if the measured value s exceeds the interval [ M3-M3 r2, m3+m3 r2], it is determined as abnormal, the deletion processing is performed, and the rest is determined as normal value. According to the principle, the abnormal data cleaning of the whole well section is completed through window sliding.

After the data preprocessing is completed, a sample set is further established. Drilling operation data is used as input, rotation guide operation instruction data is used as output, and the rotation guide operation instruction data is arranged into a format shown in a table 1, so that a control instruction can be conveniently and rapidly output in a later period:

TABLE 1

The sample set construction of this step is highly correlated with the commanded output speed of the rotary steerable drilling operation process, which is one of the key technical points of the present invention.

The input of each sample comprises well inclination a at the drill bit, azimuth b at the drill bit, drilling tool form data c in n meters behind the drill bit, average value c, d, e, f, g of drilling pressure, rotating speed, torque, drilling time and displacement in m meters in front of the drill bit, well inclination change rate i and change direction j in m meters in front of the drill bit, and the output of each sample is average value of resultant force j and direction k of guiding tools in m meters in front of the drill bit.

In the aspect of the drill morphology in n meters behind the drill, the n meters are subdivided according to 1m intervals to form n drill morphology data so as to more finely describe the drill morphology. It should be noted that all parameters of each sample correspond to only one rotation-oriented control instruction and cannot span two instructions.

The method for training, verifying and testing the constructed sample set into the training set, the verification set and the test set on the machine learning model is further configured to enable the test set to be a sample of one or more wells, enable the training set and the verification set to be samples obtained from other wells different from the test set, and enable 80% of data to be extracted from the samples of the different wells in an out-of-order mode to serve as the training set, and enable the remaining 20% of data to serve as the verification set.

The present embodiment divides the sample set into a training set, a validation set, and a test set. To test the generalization of the predictive model, the test set may be selected as a sample of one or more wells, while the training set and validation set are obtained from samples of other wells.

In terms of division of the training set and the verification set, due to the specificity of drilling engineering, input and output parameters in samples of different operation well sections have larger difference, for example, the well deviation change rate in samples of a deflecting section is usually larger, the well deviation change rate in samples of a horizontal section is usually smaller, in order that the training set can cover samples with wider parameter range, samples can be randomly extracted from samples of multiple wells in an out-of-order manner according to a certain proportion (for example, 80%) to serve as the training set, and the rest 20% of samples serve as the verification set.

The construction principle of the training set, the verification set and the test set is closely related to the drilling operation process of the rotary guiding tool, and is one key technical point of the invention.

The machine learning regression model comprises SVM, BP, LSTM, any one or more regression model combinations in deep learning. Aiming at the training set and the verification set constructed by the method, a machine learning regression prediction method, such as SVM, BP, LSTM, deep learning and the like, is selected to train the training set consisting of a plurality of samples, the parameter adjusting process in the specific training process is required to be carried out according to the characteristics of a specific algorithm, and a control instruction prediction model of the rotary guiding tool, namely a resultant force magnitude and direction instruction prediction model of the guiding tool is obtained through training. The more the number of drilled samples used to construct the training set, the more accurate the resulting predictive model, and the predictive model is dynamically updatable, i.e., drilled depth data for the current well may be added to the training set for predicting control commands in the subsequent drilling process for the current well.

Specifically, the control instruction prediction model is in an implicit form, and may be represented by f, i.e., (k, l) =f (a, b, c, d, e, f, g, h, I, j). Taking an LSTM machine learning prediction method as an example, training a training set constructed by the drilled data, and optimizing a training model by using a verification set to obtain an optimal model f ₁ under a current sample, wherein the prediction example is shown in the following table 2:

TABLE 2

In the actual drilling process, the data of the actual drilling are arranged into input parameters shown in the table 1, the input parameters are input into a prediction model, the optimal control instruction can be obtained in real time, the problem of inaccurate control instruction caused by insufficient artificial experience is avoided, and a new control instruction can be adjusted according to the actually measured geological data, so that the drilling time efficiency and the well track quality are improved.

Second embodiment

Referring to fig. 4, a system for inter-well geosteering closed loop control, the control system comprising:

The measuring data acquisition module M1 is used for acquiring underground continuous attitude measuring data and LWD measuring parameter data;

the real drilling track calculation module M2 is used for fitting according to the continuous attitude measurement data to obtain a continuous real drilling track;

When the geological deviation delta 0 is larger than a threshold epsilon 0, automatically designing a first track to be drilled of M meters in front of the drill bit by using an artificial intelligent prediction model according to a geosteering track adjustment principle, and updating the first design track according to the first track to be drilled to obtain a second design track;

The geometrical deviation calculation module M4 is used for comparing the continuous real drilling track with the first/second design track when the geological deviation delta 0 is smaller than or equal to a threshold epsilon 0, calculating the geometrical deviation delta 1, recalculating the second track to be drilled of M meters in front of the drill bit according to the first/second design track when the geometrical deviation delta 1 is larger than the threshold epsilon 1, completely jointing the joint of the second track to be drilled and the first/second design track, keeping drilling parameters and control instructions unchanged when the geometrical deviation delta 1 is smaller than or equal to the threshold epsilon 1, and sending the control instructions to the rotary guiding tool, and continuously acquiring continuous attitude measurement data and LWD measurement parameter data;

The rotary guiding tool control instruction prediction module M5 is used for inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into the rotary guiding tool control instruction prediction model to obtain new control instructions, issuing the control instructions to the rotary guiding tool, and continuously obtaining continuous attitude measurement data and LWD measurement parameter data.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the elements and steps of the examples have been generally described in terms of function in the foregoing description to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. In addition, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices, or elements, or may be an electrical, mechanical, or other form of connection.

The elements described as separate components may or may not be physically separate, and as such, those skilled in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, where the elements and steps of the examples are generally described functionally in the foregoing description of the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or a part contributing to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a grid device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. The storage medium includes a usb disk, a removable hard disk, a Read-only memory (ROM), a random access memory (RAM, random Access Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes.

The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.

Claims (10)

1. A method of inter-well geosteering closed loop control, the method comprising:

acquiring underground continuous attitude measurement data and LWD measurement parameter data;

fitting according to the continuous attitude measurement data to obtain a continuous real drilling track;

Continuously comparing the LWD measurement parameter data with the geological deviation delta ₀ between the geological requirements along the real drilling track, automatically designing a first track to be drilled of m meters in front of the drill bit according to a geosteering track adjustment principle when the geological deviation delta ₀ is larger than a threshold epsilon ₀, updating the first design track according to the first track to be drilled to obtain a second design track, and comparing the continuous real drilling track with the first/second design tracks when the geological deviation delta ₀ is smaller than or equal to the threshold epsilon ₀ to calculate the geometric deviation delta ₁;

When the geometric deviation delta ₁ is larger than the threshold epsilon ₁, recalculating a second track to be drilled of m meters in front of the drill bit according to the first/second design tracks, wherein the joint of the second track to be drilled and the first/second design tracks is completely jointed;

When the geometric deviation delta ₁ is smaller than or equal to a threshold epsilon ₁, keeping drilling parameters and control instructions unchanged, and sending the control instructions to a rotary guiding tool;

Inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into a rotary guiding tool control instruction prediction model to obtain a new control instruction, and issuing the control instruction to the rotary guiding tool to continuously obtain continuous attitude measurement data and LWD measurement parameter data.

2. The inter-well geosteering closed loop control method of claim 1, wherein the downhole continuous attitude measurement data and LWD measurement parameter data are acquired by continuous attitude measurement sensors provided on a rotary steerable tool and multi-type LWD sub provided on a cabled drill pipe, respectively, with each type of LWD sub being spaced apart.

3. The inter-well geosteering closed loop control method of claim 1, wherein the LWD sub measures at least azimuth gamma or azimuth resistivity.

4. The inter-well geosteering closed loop control method of claim 1, wherein the method of constructing a rotary steerable tool control command prediction model comprises:

The method comprises the steps of acquiring drilling operation data and rotary guiding operation instruction data of a drilled well, and carrying out abnormal data cleaning and data enhancement, wherein the drilling operation data comprise drilling deviation a at a drill bit, azimuth b at the drill bit, drilling tool form data c in n meters behind the drill bit, average drilling pressure d in m meters in front of the drill bit, average rotating speed e, average torque f, g during drilling, average displacement h, well deviation change rate i in m meters in front of the drill bit and change direction j;

Constructing a sample set, namely constructing the sample set by using the data subjected to cleaning enhancement, taking the drilling operation data as input and taking the rotation guiding operation instruction data as output;

selecting a machine learning model with a regression algorithm, dividing a constructed sample set into a training set, a verification set and a test set, and training, verifying and testing on the machine learning model to obtain a mature control instruction artificial intelligent prediction model.

5. The inter-well geosteering closed loop control method of claim 1, wherein the anomaly data cleansing method of drill bit at well deviation a, bit position b, and drill bit morphology data c n meters behind the bit is further configured to:

sequencing dynamic well deviation and dynamic azimuth data in the single well drilling operation process according to time sequence;

forming a first sliding window with the width W by the length W of the drill rod;

Setting the deflecting capability of a guiding tool as X degrees/Ym, and calculating the median M ₁ of all measured values in the first sliding window to obtain a well deviation a at the drill bit, a position b at the drill bit and a judging section [ M ₁-X*W/Y,M₁ +X W/Y ] of drilling tool form data c in n meters behind the drill bit;

and using the first sliding window to slide, judge and clean drilling tool form data c in drill bit position well deviation a, drill bit position azimuth b and drill bit rear n meters of the whole well section, and deleting data outside the judging section.

6. The inter-well geosteering closed loop control method of claim 1, wherein the anomaly data cleaning method for the rate of change i and direction of change j of well deviation within m meters ahead of the bit is further configured to:

screening out the data of the deflecting force and the direction corresponding to the instructions according to different instructions;

the data of the size and the direction of the deflecting force are arranged according to the starting depth and the ending depth of the instruction;

Calculating the median value M ₂ of all the oblique force magnitude and direction data under the instruction condition;

and determining a threshold adjustment proportionality coefficient r1 according to the magnitude of the deflecting force and the fluctuation amplitude of the direction data under the instruction condition, wherein a judgment section of the well deflection change rate h and the change direction j is [ M ₂- M₂*r1,M₂+ M₂ x r1], and deleting the data outside the judgment section.

7. The inter-well geosteering closed loop control method of claim 1, wherein the method of anomaly data cleaning for weight-on-bit average d, rotational speed average e, torque average f, g-while-drilling, displacement average h in m meters ahead of the bit is further configured to:

arranging the drilling pressure, the rotating speed, the torque, the drilling time and the displacement data according to time sequence;

Forming a second sliding window with the width W by the length W of the drill rod;

Calculating a median value M ₃ of the corresponding data types in the second sliding window;

and determining a threshold adjustment proportionality coefficient r2 according to fluctuation amplitude of different types of data, wherein a judgment section of the weight on bit, the torque, the rotating speed, the displacement and the drilling time data is [ M ₃- M₃*r2,M₃+ M₃ x r2], and deleting the data outside the judgment section.

8. The inter-well geosteering closed loop control method of claim 1, wherein the data enhancement method is further configured to complete and smooth data by performing null value filling and data encryption interpolation on the data cleaned down the well depth.

9. The inter-well geosteering closed loop control method of claim 1, wherein the method of dividing the constructed sample set into a training set, a validation set and a test set for training, validating and testing on a machine learning model is further configured such that the test set is a sample of one or more wells, the training set and the validation set are samples obtained from other wells different from the test set, and 80% of the data is extracted from the samples of the different wells in an out-of-order manner as the training set, and the remaining 20% of the data is used as the validation set.

10. An inter-well geosteering closed loop control system, the control system comprising:

The measuring data acquisition module is used for acquiring underground continuous attitude measuring data and LWD measuring parameter data;

The real drilling track calculation module is used for fitting according to the continuous attitude measurement data to obtain a continuous real drilling track;

When the geological deviation delta ₀ is larger than a threshold epsilon ₀, automatically designing a first track to be drilled of m meters in front of the drill bit by using an artificial intelligent prediction model according to a geological guiding track adjustment principle, and updating a first design track according to the first track to be drilled to obtain a second design track;

The geometrical deviation calculation module is used for comparing the continuous real drilling track with the first/second design track when the geological deviation delta ₀ is smaller than or equal to a threshold epsilon ₀, calculating geometrical deviation delta ₁, recalculating a second track to be drilled of m meters in front of the drill bit according to the first/second design track when the geometrical deviation delta ₁ is larger than the threshold epsilon ₁, completely jointing the joint of the second track to be drilled and the first/second design track, keeping drilling parameters and control instructions unchanged when the geometrical deviation delta ₁ is smaller than or equal to the threshold epsilon ₁, and enabling the control instructions to be issued to the rotary guiding tool, and continuously acquiring continuous attitude measurement data and LWD measurement parameter data;

The rotary guiding tool control instruction prediction module is used for inputting the first track to be drilled or the second track to be drilled and drilling operation parameters corresponding to the tracks to be drilled into the rotary guiding tool control instruction prediction model to obtain new control instructions, and sending the control instructions to the rotary guiding tool to continuously obtain continuous attitude measurement data and LWD measurement parameter data.