CN115095318B - A Particle Filter-Based Wellbore Trajectory Prediction Method and Device - Google Patents

Abstract

This specification provides a particle filter-based wellbore trajectory prediction method and device. The method includes: obtaining the adjacent well data around the target well to be predicted, and obtaining preprocessing data based on the adjacent well data; constructing the well inclination angle observation equation and the state equation, and the azimuth angle observation equation and the state equation based on the preprocessing data; Particle filtering, solving the inclination angle observation equation and the state equation, and the azimuth angle observation equation and the state equation to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well. Based on the above method, the technical problem of poor adaptive ability of the wellbore trajectory prediction algorithm existing in the existing methods can be solved, and the accurate prediction of the wellbore trajectory and the uncertain description of the prediction results of the wellbore trajectory can be realized.

Description

A method and device for predicting wellbore trajectory based on particle filtering

Technical Field

The present invention relates to the technical field of oil and gas exploration and development, and in particular to a method and device for predicting wellbore trajectory based on particle filtering.

Background Art

In the field of oil and gas development, the accuracy of the wellbore trajectory during drilling will affect the oil well production. Under normal circumstances, the wellbore trajectory is difficult to measure in real time with high frequency while drilling. On-site construction personnel mainly rely on personal experience and interval measurement information to adjust the directional construction parameters and determine the wellbore trajectory based on the adjusted directional construction parameters. Unreasonable decisions on directional construction parameter adjustment will lead to inaccurate prediction of the wellbore trajectory, which will in turn reduce the wellbore quality and construction efficiency, and increase safety risks.

Existing wellbore trajectory prediction methods include: geometric extrapolation, mechanical analysis and statistical methods. Among them, the geometric extrapolation method does not consider the impact of drilling parameters and formation characteristics on the wellbore trajectory, and the calculation accuracy is insufficient; the mechanical analysis method uses a model that considers many factors, and the input parameters are not easy to obtain, making it difficult to be widely used in the drilling site; the statistical method uses data from neighboring wells or drilled wells to establish a regression model, and the regression coefficient reflects the formation characteristics of the wellbore to be drilled, but its adaptive ability is poor.

Therefore, there is an urgent need for a wellbore trajectory prediction method that can solve the problem of poor adaptability.

Summary of the invention

This specification provides a wellbore trajectory prediction method and device based on particle filtering, which can solve the technical problem of poor adaptability of wellbore trajectory prediction algorithms in existing methods and obtain wellbore trajectory parameters and wellbore trajectory parameter confidence intervals of the well section to be predicted.

The purpose of the embodiment of this specification is to provide a wellbore trajectory prediction method based on particle filtering, comprising:

Acquire the data of neighboring wells around the target well to be predicted, and obtain preprocessed data based on the data of neighboring wells;

According to the preprocessed data, the observation equation and state equation of the well inclination angle, as well as the observation equation and state equation of the azimuth angle are constructed;

Based on particle filtering, the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation are solved to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well.

Furthermore, in another embodiment of the method, the step of acquiring the data of adjacent wells around the target well to be predicted and obtaining the preprocessed data based on the adjacent well data includes:

Performing a first segmentation of the wellbore based on the trajectory measurement data to obtain a first segmentation result of the wellbore;

Obtaining drilling data according to the first segmentation result of the wellbore and the logging meter data;

According to the first segmentation result of the wellbore and the directional well construction record data, an average value of the tool face angle is obtained;

The wellbore is segmented for the second time based on the parameter data of the adjacent wells to obtain the second segmentation result of the wellbore.

Furthermore, in another embodiment of the method, constructing the well inclination angle observation equation and state equation, and the azimuth angle observation equation and state equation according to the preprocessed data includes:

The well inclination angle observation equation is constructed according to the following formula:

Inc _{b, i} = (θ _{0, i} × cos (tf _{b, i} ) + θ _{1, i} × WOB _{b, i} + θ _{2, i} ) × L _{b, i} + θ _{3, i} × L _{h, i} + θ _4,i +Inc _b,i-1

Wherein, i and i-1 represent the number of the well section to be predicted, b represents the bottom, Inc _b,i is the well inclination angle at the bottom of the well section to be predicted, tf _b,i is the average value of the tool face angle, WOB _b,i is the average drilling pressure of sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Inc _b,i-1 is the well inclination angle at the top of the well section to be predicted, θ _0,i , θ _1,i , θ _2,i , θ _3,i , θ _4,i are the well inclination state parameters of the i well section, wherein θ _0,i is the well inclination angle build-up rate correction coefficient of the i well section, θ _1,i is the well inclination angle drilling pressure influence coefficient of the i well section, θ _2,i is the well inclination angle build-up rate error of the i well section, θ _3,i is the slope change law coefficient of the well inclination angle composite drilling distance of the i well section, and θ _4,i is the well inclination angle system error of the i well section.

Furthermore, in another embodiment of the method, constructing the well inclination angle observation equation and state equation, and the azimuth angle observation equation and state equation according to the preprocessed data includes:

The state equation of well inclination is constructed according to the following formula:

θ _0,i =θ _0,i-1 +w _0,i-1

θ _{1, i} = θ _{1, i-1} + w _{1, i-1}

θ _2,i =θ _2,i-1 +w _2,i-1

θ _{3, i} = θ _{3, i-1} + w _{3, i-1}

θ _{4, i} = θ _{4, i-1} + w _{4, i-1}

Wherein, w0 _,i-1 , _w1,i-1 , _w2,i-1 , _w3,i-1 , _w4,i-1 are the Gaussian white noise of the well inclination angle of the i-1 well section, _θ0,i is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i is the well inclination angle build-up rate error of the i-1 well section, _θ3,i is the slope variation law coefficient of the well inclination angle composite drilling distance of the i-1 well section, _θ4,i is the well inclination angle system error of the i-1 well section, _θ0,i-1 is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i-1 is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i-1 is the well inclination angle build-up rate error of the i-1 well section, θ _{3, i-1} is the slope variation coefficient of the well inclination angle composite drilling distance of the i-1 well section, θ _{4, i-1} is the systematic error of the well inclination angle of the i-1 well section.

Furthermore, in another embodiment of the method, constructing the well inclination angle observation equation and state equation, and the azimuth angle observation equation and state equation according to the preprocessed data includes:

Construct the azimuth observation equation according to the following formula:

Azi _b,i =(φ _0,i ×cos(tf _b,i )+φ _1,i ×WOB _b,i +φ _2,i )×L _b,i +φ _3,i ×L _h,i + φ _4,i +Azi _b,i-1

where i and i-1 represent the numbers of the well sections to be predicted, b represents the bottom, Azi _b,i is the azimuth of the bottom of the well section to be predicted, tf _b,i is the average tool face angle, WOB _b,i is the average drilling pressure during sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Azi _b,i-1 is the azimuth of the top of the well section to be predicted, φ _0,i , φ _1,i , φ _2,i , φ _3,i , φ _1,i are the azimuth state parameters of the i well section, where φ _0,i is the azimuth buildup rate correction coefficient of the i well section, φ _1,i is the azimuth buildup rate influence coefficient of the i well section, φ _2,i is the azimuth buildup rate error of the i well section, φ _3,i is the azimuth composite drilling distance slope variation coefficient of the i well section, and φ _4,i is the azimuth system error of the i well section.

Furthermore, in another embodiment of the method, constructing the well inclination angle observation equation and state equation, and the azimuth angle observation equation and state equation according to the preprocessed data includes:

The azimuth state equation is constructed according to the following formula:

φ _0,i =φ _0,i-1 +n _0,i-1

φ _1,i =φ _1,i-1 +n _1,i-1

φ _2,i =φ _2,i-1 +n _2,i-1

φ _3,i =φ _3,i-1 +n _3,i-1

φ _4,i =φ _4,i-1 +n _4,i-1

Wherein, n0 _,i-1 , _n1,i-1 , _n2,i-1 , _n3,i-1 , _n4,i-1 are the azimuth Gaussian white noise of well section i-1, _φ0,i is the azimuth buildup rate correction coefficient of well section i, _φ1,i is the azimuth drilling pressure influence coefficient of well section i, φ2 _,i is the azimuth buildup rate error of well section i, _φ3,i is the azimuth composite drilling distance slope variation law coefficient of well section i, _φ4,i is the azimuth system error of well section i, _φ0,i-1 is the azimuth buildup rate correction coefficient of well section i-1, φ1 _,i-1 is the azimuth drilling pressure influence coefficient of well section i-1, _φ2,i-1 is the azimuth buildup rate error of well section i-1, φ _{3, i-1} is the coefficient of slope variation of azimuth composite drilling distance of well section i-1, φ _{4, i-1} is the azimuth system error of well section i-1.

Further, in another embodiment of the method, the particle filtering-based method of solving the well inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation to determine the wellbore trajectory parameters and the wellbore trajectory parameter confidence interval of the target well includes:

The wellbore trajectory parameters and wellbore trajectory parameter confidence intervals of the current well section to be predicted in the target well are determined by solving the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation in the following manner:

According to the well inclination state equation and the azimuth state equation, the state parameters of the current well section to be predicted are obtained;

According to the state parameters of the current well section to be predicted, the well inclination observation equation and the azimuth observation equation are solved to obtain the weighted normalized particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted;

According to the particle swarm normalized by the weight of the current well section to be predicted, a resampled particle swarm of the current well section to be predicted is obtained;

Based on the resampled particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted, the final bottom well inclination angle of the current well section to be predicted, the final bottom azimuth angle of the current well section to be predicted, the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted, and the preset prediction confidence interval of the bottom azimuth angle of the current well section to be predicted are obtained.

Furthermore, in another embodiment of the method, obtaining the state parameters of the current well section to be predicted according to the well inclination state equation and the azimuth state equation comprises:

Determine whether the current well section to be predicted is the first well section;

When the current well section to be predicted is the first well section, the prior Gaussian distribution of the target well to be predicted is obtained according to the preprocessed data, and an initial particle group is generated based on the prior Gaussian distribution; the well inclination state equation and the azimuth state equation are solved according to the initial particle group to obtain the state parameters of the current well section to be predicted;

When the current well section to be predicted is not the first well section, the particle swarm after resampling of the previous well section is obtained, and the well inclination state equation and the azimuth state equation are solved according to the particle swarm after resampling of the previous well section to obtain the state parameters of the current well section to be predicted.

Furthermore, in another embodiment of the method, solving the well inclination angle observation equation and the azimuth angle observation equation according to the state parameters of the current well section to be predicted to obtain the weighted normalized particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted, comprises:

According to the state parameters of the current well section to be predicted, the well inclination angle observation equation and the azimuth angle observation equation are solved to obtain the bottom well inclination angle and the bottom azimuth angle of the current well section to be predicted;

According to the bottom well inclination angle of the current well section to be predicted and the bottom azimuth angle of the current well section to be predicted, the absolute error of the well inclination angle of the current well section to be predicted and the absolute error of the azimuth angle of the current well section to be predicted are obtained;

Recalculate the weight of each particle in the particle swarm of the current well section to be predicted according to the absolute error of the well inclination angle and the absolute error of the azimuth angle of the current well section to be predicted, and obtain a re-weighted particle swarm of the current well section to be predicted;

According to the re-weighted particle swarm of the current well section to be predicted, a particle swarm with normalized weights of the current well section to be predicted is obtained.

On the other hand, the present application provides a wellbore trajectory prediction device based on particle filtering, comprising:

A preprocessing module is used to obtain the data of neighboring wells around the target well to be predicted, and obtain preprocessing data based on the data of the neighboring wells;

A construction module, used for constructing the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation according to the preprocessed data;

The particle filter module is used to solve the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation based on particle filtering, so as to determine the wellbore trajectory parameters and the wellbore trajectory parameter confidence interval of the target well.

On the other hand, the present application also provides a computer-readable storage medium having computer instructions stored thereon, and the computer-readable storage medium implements the above-mentioned particle filtering-based wellbore trajectory prediction method when executing the instructions.

The present specification provides a method and device for predicting borehole trajectory based on particle filtering, which obtains data of neighboring wells around the target well to be predicted and obtains preprocessed data based on the data of neighboring wells; constructs the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation according to the preprocessed data; solves the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation based on particle filtering to determine the wellbore trajectory parameters and wellbore trajectory parameter confidence intervals of the target well. According to the method provided in the present specification, the technical problem of poor adaptability of the wellbore trajectory prediction algorithm existing in the existing method can be solved, and the technical effect of wellbore trajectory prediction and uncertainty description of the wellbore trajectory can be achieved.

Moreover, when determining the borehole trajectory parameters and the confidence interval of the borehole trajectory parameters of the target well based on particle filtering, the state parameters of the well section to be predicted are calculated by the particle group of the previous well section to be predicted, and the bottom well inclination angle and the bottom azimuth angle of the well section to be predicted are calculated according to the state parameters of the well section to be predicted. After normalizing and resampling the particle group, the final well inclination angle and the final azimuth angle of the bottom of the well section to be predicted are obtained according to the resampled particle group, so as to predict the borehole trajectory and achieve the purpose of online adaptive calibration of the borehole trajectory prediction parameters, thereby adjusting the directional construction parameters and drilling data.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the embodiments of this specification, the drawings required for use in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

FIG1 is a flow chart of an embodiment of a wellbore trajectory prediction method based on particle filtering provided in this specification;

FIG2 is a diagram of preprocessed data in one embodiment of the present specification;

FIG3 is a diagram showing the result of obtaining a priori Gaussian distribution of well inclination in one embodiment of this specification;

FIG4 is a diagram showing the result of obtaining the azimuth prior Gaussian distribution in one embodiment of the present specification;

FIG5 is a flowchart of an adaptive calibration in one embodiment of the present specification;

FIG6 is a diagram showing the prediction result of the well inclination angle of a target well in one embodiment of the present specification;

FIG. 7 is a schematic diagram of the module structure of an embodiment of a wellbore trajectory prediction device based on particle filtering provided in this specification.

DETAILED DESCRIPTION

In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be described clearly and completely below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

Considering that the existing wellbore trajectory prediction methods based on statistical methods usually use the data of neighboring wells or drilled wells to establish regression models, and reflect the formation characteristics of the wellbore to be drilled through regression coefficients, it is difficult to adjust its own model parameters in time according to the data of neighboring wells, resulting in poor adaptability.

Furthermore, it is also considered that the existing wellbore trajectory prediction method based on geometric extrapolation does not consider the influence of formation characteristics on the wellbore trajectory, resulting in poor accuracy of its calculation results.

In view of the above problems existing in the existing methods and the specific reasons for the above problems, this application considers introducing particle filtering for wellbore trajectory prediction to improve the adaptability of the method and the accuracy of the prediction results.

Based on the above ideas, this specification proposes a wellbore trajectory prediction method based on particle filtering. First, the neighboring well data around the target well to be predicted is obtained, and pre-processed data is obtained based on the neighboring well data; then, according to the pre-processed data, the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation are constructed; finally, based on particle filtering, the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation are solved to determine the wellbore trajectory parameters of the target well and the wellbore trajectory parameter confidence interval. Referring to FIG1, an embodiment of this specification provides a wellbore trajectory prediction method based on particle filtering. In specific implementation, the method may include the following contents.

S101: Acquire the neighboring well data around the target well to be predicted, and obtain pre-processed data based on the neighboring well data.

In some embodiments, the adjacent well data may specifically include: trajectory measurement data, logging meter data, directional well construction record data, and adjacent well parameter data; the preprocessed data may specifically include: drilling data, tool face angle average value, and wellbore second segmentation result.

In some embodiments, the trajectory measurement data may specifically include: trajectory measurement depth, well inclination, and azimuth; the logging meter data may specifically include: logging meter depth, drilling pressure, and rotation speed; the directional well construction record data may specifically include tool face angle; the adjacent well parameter data may specifically include: wellbore size, formation, and drill bit combination; the drilling data may specifically include: sliding drilling distance, sliding drilling average drilling pressure, and composite drilling distance.

In some embodiments, the above-mentioned obtaining of pre-processed data based on adjacent well data may include:

S1: Perform the first segmentation of the wellbore based on the trajectory measurement data to obtain the first segmentation result of the wellbore;

S2: Obtain drilling data according to the first segmentation result of the wellbore and the logging meter data;

S3: obtaining an average tool face angle according to the first segmentation result of the wellbore and the directional well construction record data;

S4: Perform a second segmentation on the wellbore based on the parameter data of the adjacent wells to obtain a second segmentation result of the wellbore.

In some embodiments, the above-mentioned first segmentation of the wellbore based on the trajectory measurement data to obtain the first segmentation result of the wellbore can be specifically implemented as follows: segmenting the wellbore trajectory according to the trajectory measurement data, using the trajectory measurement depth, well inclination, and azimuth of a certain measuring point as the bottom depth, bottom well inclination, and bottom azimuth of the well section, respectively, and using the trajectory measurement depth, well inclination, and azimuth of the previous measuring point as the top depth, top well inclination, and top azimuth of the well section, respectively, to obtain the first segmentation result of the wellbore.

In some embodiments, the drilling data is obtained according to the first segmentation result of the wellbore and the logging meter data. When implemented specifically, it may include: extracting the logging meter depth from the logging meter data according to the first segmentation result of the wellbore, placing each logging meter depth into the corresponding first segmentation result, and extracting the number of record rows with a rotation speed equal to 0 in each first segmentation result, recording it as the sliding drilling distance, and calculating the average drilling pressure in the rows with a rotation speed equal to 0, recording it as the sliding drilling average drilling pressure; extracting the number of record rows with a rotation speed greater than 0 in each first segmentation result, recording it as the composite drilling distance.

In some embodiments, the above-mentioned obtaining the average tool face angle based on the first segmentation result of the wellbore and the directional well construction record data may include: obtaining the tool face angle of each segment in the first segmentation result of the wellbore based on the directional well construction record data, and calculating the average value thereof, which is recorded as the average tool face angle.

In some embodiments, the second segmentation of the wellbore based on the adjacent well parameter data to obtain the second segmentation result of the wellbore may include: dividing the first segmentation result of the wellbore according to the wellbore size to obtain the first segmentation data, wherein each segment in the first segmentation data has the same wellbore size; dividing the first segmentation data according to the formation to obtain the second segmentation data, wherein each segment in the second segmentation data has the same formation; dividing the second segmentation data according to the drilling tool assembly to obtain the second segmentation result of the wellbore, wherein each segment in the second segmentation result of the wellbore has the same drilling tool assembly; wherein the minimum number of measuring points in the wellbore segment in the above second segmentation result of the wellbore cannot be less than 10; and the inclination angle of the last measuring point in each segment of the second segmentation result of the wellbore is used as the bottom inclination angle of the segment, and also the top inclination angle of the next segment; the azimuth angle of the last measuring point in each segment of the second segmentation result of the wellbore is used as the bottom azimuth angle of the segment, and also the top azimuth angle of the next segment.

In some embodiments, the pre-processed data is visualized on a software interactive page of a computer.

Through the above embodiment, the wellbore is segmented by taking into account the three factors of wellbore size, formation and drilling tool assembly, which can make the segmentation result more accurate and improve the accuracy of subsequent particle filtering method calculation.

S102: constructing the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation according to the preprocessed data.

In some embodiments, the state equation is used to combine the state prediction result of the system at the current moment with the state measurement to obtain the optimal estimate of the system; the observation equation is used to perform linear or nonlinear observation of the state to obtain the observation value of the system at the current moment. In this specification, the observation value of the borehole trajectory parameter of the previous well section can be used in combination with the preprocessed data to correct the value of the borehole trajectory parameter of the well section to be predicted to obtain a more accurate observation value; the state parameter value of the previous well section is used in combination with Gaussian white noise to infer the state parameter value of the well section to be predicted.

In some embodiments, the above-mentioned construction of the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation based on the preprocessed data may include:

The well inclination angle observation equation is constructed according to the following formula:

Inc _{b, i} = (θ _{0, i} × cos (tf _{b, i} ) + θ _{1, i} × WOB _{b, i} + θ _{2, i} ) × L _{b, i} + θ _{3, i} × L _{h, i} + θ _4,i +Inc _b,i-1 (1)

Wherein, i and i-1 represent the number of the well section to be predicted, Inc _b,i is the well inclination angle at the bottom of the well section to be predicted, tf _b,i is the average tool face angle, WOB _b,i is the average drilling pressure during sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Inc _b,i-1 is the well inclination angle at the top of the well section to be predicted, θ _0,i , θ _1,i , θ _2,i , θ _3,i , θ _4,i are the well inclination state parameters of the i well section, wherein θ _0,i is the well inclination angle build-up rate correction coefficient of the i well section, θ _1,i is the well inclination angle drilling pressure influence coefficient of the i well section, θ _{2,i is} the well inclination angle build-up rate error of the i well section, θ _3,i is the slope variation law coefficient of the well inclination angle composite drilling distance of the i well section, and θ _4,i is the well inclination angle system error of the i well section.

The state equation of well inclination is constructed according to the following formula:

Wherein, w0 _,i-1 , _w1,i-1 , _w2,i-1 , _w3,i-1 , _w4,i-1 are the Gaussian white noise of the well inclination angle of the i-1 well section, _θ0,i is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i is the well inclination angle build-up rate error of the i-1 well section, _θ3,i is the slope variation law coefficient of the well inclination angle composite drilling distance of the i-1 well section, _θ4,i is the well inclination angle system error of the i-1 well section, _θ0,i-1 is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i-1 is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i-1 is the well inclination angle build-up rate error of the i-1 well section, θ _{3, i-1} is the slope variation coefficient of the well inclination angle composite drilling distance of the i-1 well section, θ _{4, i-1} is the systematic error of the well inclination angle of the i-1 well section.

Construct the azimuth observation equation according to the following formula:

Azi _b,i =(φ _0,i ×cos(tf _b,i )+φ _1,i ×WOB _b,i +φ _2,i )×L _b,i +φ _3,i ×L _h,i + φ _4,i +Azi _b,i-1 (3)

where i and i-1 represent the number of the well section to be predicted, Azi _b,i is the azimuth of the bottom of the well section to be predicted, tf _b,i is the average tool face angle, WOB _b,i is the average drilling pressure during sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Azi _b,i-1 is the azimuth of the top of the well section to be predicted, φ _0,i , φ _1,i , φ _2,i , φ _3,i , φ _1,i are the azimuth state parameters of the i well section, where φ _0,i is the azimuth buildup rate correction coefficient of the i well section, φ _1,i is the azimuth buildup rate influence coefficient of the i well section, φ _2,i is the azimuth buildup rate error of the i well section, φ _3,i is the azimuth composite drilling distance slope variation coefficient of the i well section, and φ _4,i is the azimuth system error of the i well section.

The azimuth state equation is constructed according to the following formula:

Wherein, n0 _,i-1 , _n1,i-1 , _n2,i-1 , _n3,i-1 , _n4,i-1 are the azimuth Gaussian white noise of well section i-1, _φ0,i is the azimuth buildup rate correction coefficient of well section i, _φ1,i is the azimuth drilling pressure influence coefficient of well section i, φ2 _,i is the azimuth buildup rate error of well section i, _φ3,i is the azimuth composite drilling distance slope variation law coefficient of well section i, _φ4,i is the azimuth system error of well section i, _φ0,i-1 is the azimuth buildup rate correction coefficient of well section i-1, φ1 _,i-1 is the azimuth drilling pressure influence coefficient of well section i-1, _φ2,i-1 is the azimuth buildup rate error of well section i-1, φ _{3, i-1} is the coefficient of slope variation of azimuth composite drilling distance of well section i-1, φ _{4, i-1} is the azimuth system error of well section i-1.

S103: Based on particle filtering, solving the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation, to determine the wellbore trajectory parameters of the target well and the wellbore trajectory parameter confidence interval.

In some embodiments, the wellbore trajectory parameters of the target well may include: the final well inclination angle at the bottom of the well section to be predicted, and the final azimuth angle at the bottom of the well section to be predicted; the confidence interval of the wellbore trajectory parameters of the target well may include: the preset prediction confidence interval of the well inclination angle at the bottom of the well section to be predicted, and the preset prediction confidence interval of the azimuth angle at the bottom of the well section to be predicted.

In some embodiments, particle filtering is a nonlinear, non-Gaussian system filtering method based on the Monte Carlo concept, which improves the defect that Kalman filtering is only applicable to linear systems and has no restrictions on the process noise and measurement noise of the system. Particle filtering uses a series of weighted particles to represent the posterior probability density function to obtain the value to be predicted. Since particle filtering is applicable to both nonlinear and non-Gaussian systems, it has a wider range of applications. In this specification, firstly, a particle filter is used to update the state parameters, and the posterior distribution of the previous well section is used as the prior distribution of the current well section. The well inclination observation equation and the azimuth observation equation are solved using the updated state parameters to obtain the well inclination at the bottom of the well section to be predicted and the azimuth at the bottom of the well section to be predicted; then, the particle group is weighted normalized and resampled to obtain the resampled particle group; finally, according to the well inclination at the bottom of the well section to be predicted, the azimuth at the bottom of the well section to be predicted, and the resampled particle group, the wellbore trajectory parameters and the wellbore trajectory parameter confidence interval of the well section to be predicted are calculated to realize the online adaptive calibration of the wellbore trajectory parameters, and then the directional construction parameters can be adjusted; the above-mentioned posterior distribution specifically refers to the particle group after the resampling of the previous well section.

In some embodiments, the above-mentioned particle filtering is used to solve the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation to determine the wellbore trajectory parameters and the wellbore trajectory parameter confidence interval of the target well. When implemented specifically, it may include:

The wellbore trajectory parameters and wellbore trajectory parameter confidence intervals of the current well section to be predicted in the target well are determined by solving the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation in the following manner:

S1: According to the well inclination state equation and the azimuth state equation, the state parameters of the current well section to be predicted are obtained;

S2: according to the state parameters of the current well section to be predicted, solve the well inclination angle observation equation and the azimuth angle observation equation to obtain the weighted normalized particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted;

S3: obtaining a resampled particle swarm of the current well section to be predicted according to the particle swarm normalized by the weight of the current well section to be predicted;

S4: Based on the resampled particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted, obtain the bottom final well inclination angle of the current well section to be predicted, the bottom final azimuth angle of the current well section to be predicted, the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted, and the preset prediction confidence interval of the bottom azimuth angle of the current well section to be predicted.

In some embodiments, the state parameters of the current well section to be predicted are obtained according to the well inclination state equation and the azimuth state equation. When implemented specifically, the following may be included:

S1: Determine whether the current well section to be predicted is the first well section;

S2: When the current well section to be predicted is the first well section, the prior Gaussian distribution of the target well to be predicted is obtained based on the preprocessed data, and an initial particle swarm is generated based on the prior Gaussian distribution; the well inclination state equation and the azimuth state equation are solved based on the initial particle swarm to obtain the state parameters of the current well section to be predicted;

S3: When the current well section to be predicted is not the first well section, the particle swarm after resampling of the previous well section is obtained, and the well inclination state equation and the azimuth state equation are solved according to the particle swarm after resampling of the previous well section to obtain the state parameters of the current well section to be predicted.

In some embodiments, the above-mentioned state parameters may specifically include: well inclination state parameters and azimuth state parameters.

In some embodiments, the above-mentioned initial particle group may specifically include: an initial particle group of well inclination angle and an initial particle group of azimuth angle; the initial particle group of well inclination angle and the initial particle group of azimuth angle are both a group of randomly generated data sets with the above-mentioned prior Gaussian distribution, and the weight of each particle in the initial particle group of well inclination angle is equal, and the weight of each particle in the initial particle group of azimuth angle is equal.

其为N×5的矩阵，N表示井斜角初始粒子群的粒子数，每个粒子的权重可以记为1/N；上述方位角初始粒子群可以记为

其为M×5的矩阵，M表示方位角初始粒子群的粒子数，每个粒子的权重可以记为1/M。Among them, the initial particle group of the above-mentioned well inclination angle can be recorded as

It is an N×5 matrix, where N represents the number of particles in the initial particle group of well inclination angle, and the weight of each particle can be recorded as 1/N; the above initial particle group of azimuth angle can be recorded as

It is an M×5 matrix, where M represents the number of particles in the azimuth initial particle group, and the weight of each particle can be recorded as 1/M.

In some embodiments, when the current well section to be predicted is the first well section, obtaining the prior Gaussian distribution of the target well to be predicted based on the preprocessed data may include:

S1: According to the second segmentation result of the wellbore, based on the top well inclination angle, sliding drilling distance, sliding drilling average drilling pressure, composite drilling average drilling pressure, and bottom well inclination angle, the well inclination angle observation equation is fitted to obtain the well inclination angle state parameters of each well section;

S2: According to the results of the second segmentation of the wellbore, the azimuth observation equation is fitted based on the top azimuth, sliding drilling distance, sliding drilling average drilling pressure, composite drilling average drilling pressure, and bottom azimuth to obtain the azimuth state parameters of each well section;

S3: Calculate the expectation and variance of the well inclination state parameter of each well section and the expectation and variance of the azimuth state parameter of each well section to obtain the prior Gaussian distribution of the target well to be predicted.

In some embodiments, the method of fitting the well inclination observation equation and the method of fitting the azimuth observation equation may specifically include: least square method, Newton iteration method, and constrained optimization by linear approximation (COBYLA) method.

In some embodiments, the above-mentioned solving the well inclination state equation and the azimuth state equation based on the initial particle swarm to obtain the state parameters of the current well section to be predicted may include:

S1: The first column of data of the initial particle group of well inclination angle is used as the well inclination angle build-up rate correction coefficient, the second column of data is used as the well inclination angle drilling pressure influence coefficient, the third column of data is used as the well inclination angle build-up rate error, the fourth column of data is used as the well inclination angle composite drilling distance slope variation law coefficient, and the fifth column of data is used as the well inclination angle system error, and the well inclination angle state equation is solved to obtain the well inclination angle state parameters of the current well section to be predicted;

S2: The data in the first column of the azimuth initial particle group is used as the azimuth slope correction coefficient, the data in the second column is used as the azimuth drilling pressure influence coefficient, the data in the third column is used as the azimuth slope error, the data in the fourth column is used as the azimuth composite drilling distance slope variation coefficient, and the data in the fifth column is used as the azimuth system error. The azimuth state equation is solved to obtain the azimuth state parameters of the current well section to be predicted.

In some embodiments, the above-mentioned particle swarm after resampling of the previous well section solves the well inclination state equation and the azimuth state equation to obtain the state parameters of the current well section to be predicted. When implemented specifically, it may include:

S1: The first column of the well inclination angle particle group after resampling of the previous well section is used as the well inclination angle build-up rate correction coefficient of the previous well section, the second column of data is used as the well inclination angle drilling pressure influence coefficient of the previous well section, the third column of data is used as the well inclination angle build-up rate error of the previous well section, the fourth column of data is used as the well inclination angle composite drilling distance slope variation law coefficient of the previous well section, and the fifth column of data is used as the well inclination angle system error of the previous well section. The well inclination angle particle group after resampling of the previous well section is substituted into the well inclination angle state equation to solve and obtain the well inclination angle state parameters of the current well section to be predicted;

S2: The first column of the azimuth particle group after resampling of the previous well section is used as the azimuth build-up rate correction coefficient of the previous well section, the second column of data is used as the azimuth drilling pressure influence coefficient of the previous well section, the third column of data is used as the azimuth build-up rate error of the previous well section, the fourth column of data is used as the azimuth composite drilling distance slope variation law coefficient of the previous well section, and the fifth column of data is used as the azimuth system error of the previous well section. The azimuth particle group after resampling of the previous well section is substituted into the azimuth state equation to solve and obtain the azimuth state parameters of the current well section to be predicted.

In some embodiments, the above method solves the well inclination observation equation and the azimuth observation equation according to the state parameters of the current well section to be predicted, and obtains the weighted normalized particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted. When implemented specifically, the following steps may be included:

S1: solving the well inclination angle observation equation and the azimuth angle observation equation according to the state parameters of the well section to be predicted, and obtaining the well inclination angle at the bottom of the well section to be predicted and the azimuth angle at the bottom of the well section to be predicted;

S2: according to the bottom well inclination angle of the current well section to be predicted and the bottom azimuth angle of the current well section to be predicted, obtaining the absolute error of the well inclination angle of the current well section to be predicted and the absolute error of the azimuth angle of the current well section to be predicted;

S3: recalculating the weight of each particle in the particle swarm of the current well section to be predicted according to the absolute error of the well inclination angle and the absolute error of the azimuth angle of the current well section to be predicted, and obtaining a re-weighted particle swarm of the current well section to be predicted;

S4: According to the re-weighted particle swarm of the current well section to be predicted, a particle swarm with normalized weights of the current well section to be predicted is obtained.

In some embodiments, the above-mentioned re-weighted particle group specifically includes: a particle group re-weighted by well inclination angle and a particle group re-weighted by azimuth angle.

In some embodiments, the weight-normalized particle group specifically includes: a particle group with normalized well inclination weight and a particle group with normalized azimuth weight.

In some embodiments, the above-mentioned method of obtaining the absolute error of the inclination angle of the current well section to be predicted and the absolute error of the azimuth angle of the current well section to be predicted based on the bottom well inclination angle of the current well section to be predicted and the bottom azimuth angle of the current well section to be predicted may include:

The absolute error of the well inclination angle of the current well section to be predicted is calculated according to the following formula:

为当前待预测井段的井斜角绝对误差，

为第k个当前待预测井段底部井斜角，

为第k个当前待预测井段底部井斜角的实测值。In the formula,

is the absolute error of the well inclination angle of the current well section to be predicted,

is the bottom inclination angle of the kth well section to be predicted,

is the measured value of the bottom well inclination angle of the kth well section to be predicted.

Calculate the absolute azimuth error of the current well section to be predicted according to the following formula:

为当前待预测井段的方位角绝对误差，

为第l个当前待预测井段底部方位角，

为第l个当前待预测井段底部方位角的实测值。In the formula,

is the absolute error of the azimuth of the well section to be predicted,

is the bottom azimuth of the lth well section to be predicted,

is the measured value of the bottom azimuth of the lth well section to be predicted.

In some embodiments, the weight of each particle in the particle swarm of the current well section to be predicted is recalculated according to the absolute error of the well inclination angle and the absolute error of the azimuth angle of the current well section to be predicted to obtain the re-weighted particle swarm of the current well section to be predicted. Specifically, the method includes:

The weight of the particle swarm re-weighted by the well inclination angle of the current well section to be predicted is calculated according to the following formula:

为当前待预测井段的井斜角重新赋权的粒子群中第k个粒子的权重，R_Inc为常数，根据经验确定。In the formula,

is the weight of the kth particle in the particle swarm that is re-weighted for the well inclination angle of the current well section to be predicted, and R _Inc is a constant determined based on experience.

The weight of the particle swarm whose azimuth is re-weighted for the current well section to be predicted is calculated according to the following formula:

为当前待预测井段的方位角重新赋权的粒子群中第l个粒子的权重，R_Azi为常数，根据经验确定。In the formula,

is the weight of the lth particle in the particle swarm that re-weights the azimuth of the current well section to be predicted, and R _Azi is a constant determined based on experience.

In some embodiments, the particle group with normalized weights of the current well section to be predicted is obtained based on the re-weighted particle group of the current well section to be predicted. When implemented specifically, the particle group may include:

The weight of the particle swarm after the normalization of the well inclination weight of the current well section to be predicted is calculated according to the following formula:

为井斜角权重归一化后的粒子群中第k个粒子的权重。In the formula,

is the weight of the kth particle in the particle swarm after the well inclination weight is normalized.

The weight of the particle swarm after the azimuth weight normalization of the current well section to be predicted is calculated according to the following formula:

为方位角权重归一化后的粒子群中第l个粒子的权重。In the formula,

is the weight of the lth particle in the particle swarm after the azimuth weight is normalized.

In some embodiments, the above-mentioned particle group normalized according to the weight of the current well section to be predicted is used to obtain the resampled particle group of the current well section to be predicted. When implemented specifically, it can include: resampling the particle group normalized according to the weight of the current well section to be predicted according to the weight size, the particles with higher weights have more resample times, after resampling, the total number of particles remains unchanged, particles with large weights are divided into multiple particles, particles with small weights are discarded, and after resampling, the weight of each particle is the same; wherein, the resampling method can specifically include: random resampling, systematic sampling, residual sampling, polynomial sampling; the resampling particle group can specifically include: well inclination resampling particle group, azimuth resampling particle group.

In some embodiments, the above-mentioned resampled particle group based on the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted are used to obtain the bottom final well inclination angle of the current well section to be predicted, the bottom final azimuth angle of the current well section to be predicted, the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted, and the preset prediction confidence interval of the bottom azimuth angle of the current well section to be predicted. When implemented specifically, it may include:

The final well inclination angle at the bottom of the current well section to be predicted is calculated according to the following formula:

为当前待预测井段的井斜角重采样粒子群中第k个粒子的值。Where, Inc _final,b,i is the final well inclination angle at the bottom of the well section to be predicted.

is the value of the kth particle in the well inclination resampling particle swarm for the current well section to be predicted.

The final azimuth of the bottom of the well section to be predicted is calculated according to the following formula:

为当前待预测井段的方位角重采样粒子群中第l个粒子的值。Where Azi _final,b,i is the final azimuth of the bottom of the well section to be predicted.

is the value of the lth particle in the azimuth resampling particle swarm of the current well section to be predicted.

In some embodiments, the final well inclination angle at the bottom of the well section to be predicted and the final azimuth angle at the bottom of the well section to be predicted are visualized on the software interactive page of the computer.

In some embodiments, after obtaining the wellbore trajectory parameters of the current well section to be predicted, the wellbore trajectory parameters can be used to predict the wellbore trajectory, that is, the wellbore trajectory parameters are used to calculate the position coordinates of the corresponding measuring points according to the trajectory shape, and finally the spatial data of the wellbore trajectory morphology is obtained. The on-site staff makes decisions based on the spatial data, providing a theoretical basis for directing the drilling work.

When the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted can be the 95% confidence interval of the bottom well inclination angle of the current well section to be predicted, the lower limit of the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted is calculated according to the following formula:

Inc _pred,min =μ _Inc -1.96σ _Inc (13)

Wherein, μ _Inc is the mean of the well inclination angle at the bottom of the well section to be predicted, σ _Inc is the variance of the well inclination angle at the bottom of the well section to be predicted, and Inc _pred,min is the lower limit of the preset prediction confidence interval of the well inclination angle at the bottom of the well section to be predicted.

When the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted can be the 95% confidence interval of the bottom well inclination angle of the current well section to be predicted, the upper limit of the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted is calculated according to the following formula:

Inc _pred,max =μ _Inc +1.96σ _Inc (14)

Inc _pred,max is the upper limit of the prediction confidence interval of the bottom well inclination angle of the well section to be predicted.

Therefore, the preset prediction confidence interval of the bottom well inclination angle of the well section to be predicted can be recorded as [Inc _pred,min ,Inc _pred,max ].

In the case where the preset prediction confidence interval of the bottom azimuth of the current well section to be predicted can be the 95% confidence interval of the bottom azimuth of the current well section to be predicted, the lower limit of the preset prediction confidence interval of the bottom azimuth of the current well section to be predicted is calculated according to the following formula:

Azi _pred,min =μ _Azi -1.96σ _Azi (15)

Wherein, μ _Azi is the mean azimuth of the bottom of the well section to be predicted, σ _Azi is the variance of the azimuth of the bottom of the well section to be predicted, and Azi _pred,min is the lower limit of the prediction confidence interval of the azimuth of the bottom of the well section to be predicted.

In the case where the preset prediction confidence interval of the bottom azimuth of the current well section to be predicted can be the 95% confidence interval of the bottom azimuth of the current well section to be predicted, the upper limit of the preset prediction confidence interval of the bottom azimuth of the current well section to be predicted is calculated according to the following formula:

Azi _pred,max =μ _Azi +1.96σ _Azi (16)

Where Azi _pred,min is the upper limit of the prediction confidence interval of the bottom azimuth of the well section to be predicted.

Therefore, the preset prediction confidence interval of the bottom azimuth of the well section to be predicted can be recorded as [Azi _pred,min ,Azi _pred,max ].

In some embodiments, the wellbore trajectory of the target well can be described uncertainly based on the preset prediction confidence interval of the bottom well inclination angle of the well section to be predicted and the preset prediction confidence interval of the bottom azimuth angle of the well section to be predicted.

In a specific scenario example, the wellbore trajectory prediction method based on particle filtering provided in this specification can be applied to obtain the wellbore trajectory parameters of the target wellbore. Among them, the preprocessed data obtained based on the adjacent well data is shown in Figure 2; the target well is divided into 4 well sections, and the prior Gaussian distribution of the well inclination state parameter is shown in Figure 3, θ _i,prior (i＝0,1,2,3,4) represents the prior Gaussian distribution of the well inclination state parameter; the prior Gaussian distribution of the azimuth state parameter is shown in Figure 4, φ _i,prior (i＝0,1,2,3,4) represents the prior Gaussian distribution of the azimuth state parameter; the adaptive calibration process is shown in Figure 5, the prior Gaussian distribution of the target well is obtained based on the adjacent well data, and the resampled particle group of the previous well section is used as the posterior distribution of the previous well section, and also as the prior distribution of the current well section to be measured. The wellbore trajectory prediction result is adaptively calibrated based on the calculation result of the particle filter. According to the wellbore trajectory prediction result, the drilling operation based on the drilling data is guided. In the adaptive calibration process, the relevant parameters of the particle filter can be adaptively adjusted; the wellbore inclination prediction result is shown in Figure 6. It can be seen from Figure 6 that the wellbore trajectory prediction method based on the particle filter can achieve better prediction effect.

Although this specification provides method operation steps or device structures as shown in the following embodiments or drawings, more or fewer operation steps or module units may be included in the method or device based on routine or no creative labor. In the steps or structures where there is no necessary causal relationship logically, the execution order of these steps or the module structure of the device is not limited to the execution order or module structure shown in the embodiments or drawings of this specification. When the method or module structure is applied in an actual device, server or terminal product, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiment or drawings (for example, a parallel processor or multi-threaded processing environment, or even a distributed processing, server cluster implementation environment).

Based on the above-mentioned wellbore trajectory prediction method based on particle filtering, this specification also proposes an embodiment of a wellbore trajectory prediction device based on particle filtering. As shown in FIG7 , the wellbore trajectory prediction device based on particle filtering specifically includes the following modules:

The preprocessing module 701 is used to obtain the data of the neighboring wells around the target well to be predicted, and obtain preprocessing data based on the data of the neighboring wells;

A construction module 702 is used to construct the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation according to the preprocessed data;

The particle filter module 703 is used to solve the well inclination observation equation and state equation, as well as the azimuth observation equation and state equation based on particle filtering to determine the wellbore trajectory parameters and wellbore trajectory parameter confidence intervals of the target well.

In some embodiments, the above-mentioned preprocessing module 701 can be specifically used to perform a first segmentation of the wellbore based on trajectory measurement data to obtain a first segmentation result of the wellbore; obtain drilling data based on the first segmentation result of the wellbore and logging data; obtain an average tool face angle based on the first segmentation result of the wellbore and directional well construction record data; perform a second segmentation of the wellbore based on adjacent well parameter data to obtain a second segmentation result of the wellbore.

In some embodiments, the building block 702 may be specifically used for:

The well inclination angle observation equation is constructed according to the following formula:

Inc _{b, i} = (θ _{0, i} × cos (tf _{b, i} ) + θ _{1, i} × WOB _{b, i} + θ _{2, i} ) × L _{b, i} + θ _{3, i} × L _{h, i} + θ _4,i +Inc _b,i-1 (17)

Wherein, i and i-1 represent the number of the well section to be predicted, b represents the bottom, Inc _b,i is the well inclination angle at the bottom of the well section to be predicted, tf _b,i is the average value of the tool face angle, WOB _b,i is the average drilling pressure of sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Inc _b,i-1 is the well inclination angle at the top of the well section to be predicted, θ _0,i , θ _1,i , θ _2,i , θ _3,i , θ _4,i are the well inclination state parameters of the i well section, wherein θ _0,i is the well inclination angle build-up rate correction coefficient of the i well section, θ _1,i is the well inclination angle drilling pressure influence coefficient of the i well section, θ _2,i is the well inclination angle build-up rate error of the i well section, θ _3,i is the slope change law coefficient of the well inclination angle composite drilling distance of the i well section, and θ _4,i is the well inclination angle system error of the i well section.

The state equation of well inclination is constructed according to the following formula:

Wherein, w0 _,i-1 , _w1,i-1 , _w2,i-1 , _w3,i-1 , _w4,i-1 are the Gaussian white noise of the well inclination angle of the i-1 well section, _θ0,i is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i is the well inclination angle build-up rate error of the i-1 well section, _θ3,i is the slope variation law coefficient of the well inclination angle composite drilling distance of the i-1 well section, _θ4,i is the well inclination angle system error of the i-1 well section, _θ0,i-1 is the well inclination angle build-up rate correction coefficient of the i-1 well section, _θ1,i-1 is the well inclination angle drilling pressure influence coefficient of the i-1 well section, _θ2,i-1 is the well inclination angle build-up rate error of the i-1 well section, θ _{3, i-1} is the slope variation coefficient of the well inclination angle composite drilling distance of the i-1 well section, θ _{4, i-1} is the systematic error of the well inclination angle of the i-1 well section.

Construct the azimuth observation equation according to the following formula:

Azi _b,i =(φ _0,i ×cos(tf _b,i )+φ _1,i ×WOB _b,i +φ _2,i )×L _b,i +φ _3,i ×L _h,i + φ _4,i +Azi _b,i-1 (19)

where i and i-1 represent the numbers of the well sections to be predicted, b represents the bottom, Azi _b,i is the azimuth of the bottom of the well section to be predicted, tf _b,i is the average tool face angle, WOB _b,i is the average drilling pressure during sliding drilling, L _b,i is the sliding drilling distance, L _h,i is the composite drilling distance, Azi _b,i-1 is the azimuth of the top of the well section to be predicted, φ _0,i , φ _1,i , φ _2,i , φ _3,i , φ _1,i are the azimuth state parameters of the i well section, where φ _0,i is the azimuth buildup rate correction coefficient of the i well section, φ _1,i is the azimuth buildup rate influence coefficient of the i well section, φ _2,i is the azimuth buildup rate error of the i well section, φ _3,i is the azimuth composite drilling distance slope variation coefficient of the i well section, and φ _4,i is the azimuth system error of the i well section.

The azimuth state equation is constructed according to the following formula:

Wherein, n0 _,i-1 , _n1,i-1 , _n2,i-1 , _n3,i-1 , _n4,i-1 are the azimuth Gaussian white noise of well section i-1, _φ0,i is the azimuth buildup rate correction coefficient of well section i, _φ1,i is the azimuth drilling pressure influence coefficient of well section i, φ2 _,i is the azimuth buildup rate error of well section i, _φ3,i is the azimuth composite drilling distance slope variation law coefficient of well section i, _φ4,i is the azimuth system error of well section i, _φ0,i-1 is the azimuth buildup rate correction coefficient of well section i-1, φ1 _,i-1 is the azimuth drilling pressure influence coefficient of well section i-1, _φ2,i-1 is the azimuth buildup rate error of well section i-1, φ _{3, i-1} is the coefficient of slope variation of azimuth composite drilling distance of well section i-1, φ _{4, i-1} is the azimuth system error of well section i-1.

In some embodiments, the particle filter module 703 can be specifically used to solve the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation in the following manner, through particle filtering, to determine the borehole trajectory parameters and the confidence interval of the borehole trajectory parameters of the current well section to be predicted in the target well: according to the well inclination angle state equation and the azimuth angle state equation, the state parameters of the current well section to be predicted are obtained; according to the state parameters of the current well section to be predicted, the well inclination angle observation equation and the azimuth angle observation equation are solved to obtain the particle group after the weight normalization of the current well section to be predicted , the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted; according to the particle group after weight normalization of the current well section to be predicted, the resampled particle group of the current well section to be predicted is obtained; based on the resampled particle group of the current well section to be predicted, the bottom well inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted, the final well inclination angle of the bottom of the current well section to be predicted, the preset prediction confidence interval of the bottom well inclination angle of the current well section to be predicted, and the preset prediction confidence interval of the bottom azimuth angle of the current well section to be predicted are obtained.

The embodiments of the present specification also provide a computer storage medium for a wellbore trajectory prediction method based on particle filtering, wherein the computer storage medium stores computer program instructions, which, when executed, implement the following: obtaining the data of neighboring wells around the target well to be predicted, and obtaining preprocessed data based on the neighboring well data; constructing the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation according to the preprocessed data; solving the well inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation based on particle filtering to determine the wellbore trajectory parameters of the target well and the wellbore trajectory parameter confidence interval.

In this embodiment, the storage medium includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a cache, a hard disk (HDD), or a memory card. The memory may be used to store computer program instructions. The network communication unit may be an interface for network connection communication set in accordance with the standard specified by the communication protocol.

Although the present specification provides method operation steps as described in the embodiments or flow charts, more or less operation steps may be included based on conventional or non-creative means. The order of steps listed in the embodiments is only one way of executing the order of many steps, and does not represent the only execution order. When the device or client product in practice is executed, it can be executed in sequence or in parallel according to the method shown in the embodiments or the drawings (for example, a parallel processor or a multi-threaded processing environment, or even a distributed data processing environment). The term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, product or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such process, method, product or device. In the absence of more restrictions, it is not excluded that there are other identical or equivalent elements in the process, method, product or device including the elements. The first, second, etc. words are used to represent the name, and do not represent any particular order.

This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

Through the description of the above embodiments, it can be known that those skilled in the art can clearly understand that the present specification can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present specification can essentially be embodied in the form of a software product, which can be stored in a storage medium such as ROM/RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a mobile terminal, a server, or a network device, etc.) to execute the methods described in each embodiment of the present specification or some parts of the embodiments.

Although the present specification is described through embodiments, those skilled in the art will appreciate that there are many modifications and changes to the present specification without departing from the spirit of the present specification, and it is intended that the appended claims include these modifications and changes without departing from the spirit of the present specification.

Claims (9)

1. A method for predicting borehole trajectory based on particle filter, characterized in that, comprising: Obtain the data of adjacent wells around the target well to be predicted, and obtain preprocessing data based on the data of adjacent wells; According to the preprocessed data, the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation are constructed; Based on particle filtering, solve the inclination angle observation equation and state equation, as well as the azimuth angle observation equation and state equation to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well; Among them, based on the particle filter, the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation are solved to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well, including: The wellbore trajectory parameters and the confidence intervals of the wellbore trajectory parameters of the current well segment to be predicted in the target well are determined by solving the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation through particle filtering in the following manner: According to the inclination angle state equation and the azimuth angle state equation, obtain the state parameters of the current well section to be predicted; According to the state parameters of the current well section to be predicted, the inclination angle observation equation and the azimuth angle observation equation are solved to obtain the weighted normalized particle swarm of the current well section to be predicted, the bottom inclination angle of the current well section to be predicted, the current The bottom azimuth of the well section to be predicted; According to the particle swarm normalized by the weight of the current well section to be predicted, the resampling particle swarm of the current well section to be predicted is obtained; Based on the resampling particle swarm of the current to-be-predicted well section, the bottom inclination angle of the current to-be-predicted well section, and the bottom azimuth of the current to-be-predicted well section, obtain the bottom final inclination angle of the current to-be-predicted well section, the current to-be-predicted well section The final azimuth at the bottom of the section, the preset prediction confidence interval for the bottom inclination angle of the current well section to be predicted, and the preset prediction confidence interval for the bottom azimuth of the current well section to be predicted. 2. The method according to claim 1, wherein obtaining preprocessing data based on offset well data comprises: Based on the trajectory measurement data, the borehole of the target well is segmented for the first time, and the result of the first segment of the borehole is obtained; Obtain the drilling data according to the first segmentation result of the wellbore and the mud logging meter data; Obtaining the average value of the tool face angle according to the result of the first segmentation of the borehole and the construction record data of the directional well; Based on the parameter data of adjacent wells, the wellbore is segmented for the second time, and the result of the second segment of the wellbore is obtained. 3. method according to claim 1, is characterized in that, according to preprocessing data, constructs well inclination angle observation equation, comprises: Construct the well inclination angle observation equation according to the following formula: Inc _b,i = (θ _0,i ×cos(tf _b,i )+θ _1,i ×WOB _b,i +θ _2,i )×L _b,i +θ _3,i ×L _h,i + θ _4,i +Inc _b,i-1 In the formula, i and i-1 represent the number of the well section to be predicted, b represents the bottom, Inc _{b, i} is the inclination angle at the bottom of the well section to be predicted, tf _{b, i} is the average tool face angle, WOB _{b, i} is the average WOB of sliding drilling, L _{b, i} is the sliding drilling distance, L _{h, i} is the composite drilling distance, Inc _{b, i-1} is the inclination angle at the top of the well section to be predicted, θ _{0, i} , θ _{1, i} , θ _{2, i} , θ _{3, i} , θ _{4, i} are the well inclination angle state parameters of well section i, where θ _{0, i} is the inclination angle build-up rate correction coefficient of well section i, θ _{1 , i} is the well inclination angle WOB influence coefficient of i well section, θ _{2, i} is the well inclination angle build-up rate error of i well section, θ _{3, i} is the variation law of the well inclination compound drilling distance slope of i well section Coefficient, θ _{4, i} is the systematic error of inclination angle of well segment i. 4. method according to claim 1, is characterized in that, according to preprocessing data, constructs well inclination angle state equation, comprises: Construct the well inclination angle state equation according to the following formula: θ _0,i = θ _0,i-1 +w _0,i-1 θ _1,i = θ _1,i-1 +w _1,i-1 θ _2,i = θ _2,i-1 +w _2,i-1 θ _3,i = θ _3,i-1 +w _3,i-1 θ _4,i = θ _4,i-1 +w _4,i-1 In the formula, w _{0, i-1} , w _{1, i-1} , w _{2, i-1} , w _{3, i-1,} w _{4, i-1} are Gaussian white noise of inclination angle of well section i-1 , θ _{0, i} is the correction coefficient of inclination angle build-up rate of well section i, θ _{1, i} is the influence coefficient of well inclination angle on bit of well section i, θ _{2, i} is the error of inclination angle build-up rate of well section i , θ _{3, i} is the slope variation coefficient of the well inclination composite drilling distance of well section i, θ _{4, i} is the systematic error of well inclination angle of well section i, θ _{0, i-1} is the well inclination angle of well section i-1 Well inclination angle build-up rate correction coefficient, θ _{1, i-1} is the influence coefficient of well inclination angle on WOB of i-1 well section, θ _{2, i-1} is the well inclination angle build-up rate error of i-1 well section, θ _{3. i-1} is the slope variation coefficient of the well inclination composite drilling distance of the i-1 well section, θ _{4, i-1} is the well inclination angle systematic error of the i-1 well section. 5. method according to claim 1, is characterized in that, according to preprocessing data, constructs azimuth angle observation equation, comprises: Construct the azimuth observation equation according to the following formula: Azi _b,i ＝(φ _0,i ×cos(tf _b,i )+φ _1,i ×WOB _b,i +φ _2,i )×L _b,i +φ _3,i ×L _h,i + φ _4,i +Azi _b,i-1 In the formula, i and i-1 represent the number of the well segment to be predicted, b represents the bottom, Azi _{b, i} is the azimuth angle of the bottom of the well segment to be predicted, tf _{b, i} is the average tool face angle, WOB _{b, i} is Average WOB of sliding drilling, L _{b, i} is the sliding drilling distance, L _{h, i} is the compound drilling distance, azi _{b, i-1} is the top azimuth of the well section to be predicted, φ _{0, i} , φ _{1, i} , φ _2,i , φ _3,i , φ _1,i are the azimuth angle state parameters of well section i, where φ _0,i is the azimuth build-up rate correction coefficient of well section i, and φ _1,i is i The azimuth WOB influence coefficient of well section, φ _{2, i} is the azimuth build-up rate error of well section i, φ _{3, i} is the azimuth composite drilling distance slope variation coefficient of well section i, φ _{4, i} is Systematic error of the azimuth angle of well section i. 6. method according to claim 1, is characterized in that, according to preprocessing data, builds azimuth angle state equation, comprises: Construct the azimuth angle state equation according to the following formula: φ _0,i = φ _0,i-1 +n _0,i-1 φ _1,i = φ _1,i-1 +n _1,i-1 φ _2,i = φ _2,i-1 +n _2,i-1 φ _3,i = φ _3,i-1 +n _3,i-1 φ _4,i = φ _4,i-1 +n _4,i-1 In the formula, n _{0, i-1} , n _{1, i-1} , n _{2, i-1,} n _{3, i-1,} n _{4, i-1} is the azimuth Gaussian white noise of well section i-1, φ _{0, i} is the azimuth build-up rate correction coefficient of well section i, φ _{1, i} is the azimuth WOB influence coefficient of well section i, φ _{2, i} is the azimuth build-up rate error of well section i, φ _{3, i} is the azimuth compound drilling distance slope variation coefficient of well section i, φ _{4, i} is the azimuth systematic error of well section i, φ _{0, i-1} is the azimuth build-up rate correction coefficient of well section i-1 , φ _{1, i-1} is the azimuth WOB influence coefficient of well section i-1, φ _{2, i-1} is the azimuth build-up rate error of well section i-1, φ _{3, i-1} is i-1 The azimuth composite drilling distance slope variation coefficient of the well section, φ _{4, i-1} is the azimuth system error of the i-1 well section. 7. method according to claim 1, is characterized in that, according to inclination angle state equation and azimuth angle state equation, obtain the state parameter of current well section to be predicted, comprising: Judging whether the current well segment to be predicted is the first well segment; In the case that the current well segment to be predicted is the first well segment, the prior Gaussian distribution of the target well to be predicted is obtained according to the preprocessing data, and the initial particle swarm is generated based on the prior Gaussian distribution; the well inclination is calculated according to the initial particle swarm The state equation and the azimuth angle state equation are used to obtain the current state parameters of the well segment to be predicted; When the current well section to be predicted is not the first well section, obtain the resampled particle swarm of the previous well section, and solve the inclination angle state equation and azimuth state equation according to the resampled particle swarm of the previous well section , to get the state parameters of the current well section to be predicted. 8. The method according to claim 1, wherein, according to the state parameters of the current well section to be predicted, the well inclination observation equation and the azimuth observation equation are solved to obtain the weight normalization of the current well section to be predicted Particle swarm, the bottom inclination angle of the current well section to be predicted, and the bottom azimuth angle of the current well section to be predicted, including: According to the state parameters of the current well section to be predicted, the inclination angle observation equation and the azimuth angle observation equation are solved to obtain the bottom well inclination angle of the current well section to be predicted and the bottom azimuth angle of the current well section to be predicted; According to the bottom inclination angle of the current well section to be predicted and the bottom azimuth angle of the current well section to be predicted, the absolute error of the inclination angle of the current well section to be predicted and the absolute error of the azimuth angle of the current well section to be predicted are obtained; Recalculate the weight of each particle in the particle swarm of the current well section to be predicted according to the absolute error of the inclination angle of the current well section to be predicted and the absolute error of the azimuth angle of the current well section to be predicted, and obtain the reweighting of the current well section to be predicted the particle swarm; According to the re-weighted particle swarm of the current well section to be predicted, the weighted normalized particle swarm of the current well section to be predicted is obtained. 9. A particle filter-based wellbore trajectory prediction device, characterized in that it comprises: The preprocessing module is used to obtain the adjacent well data around the target well to be predicted, and obtain preprocessed data based on the adjacent well data; The construction module is used to construct the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation according to the preprocessed data; The particle filter module is used to solve the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation based on the particle filter, so as to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well; Among them, based on the particle filter, the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation are solved to determine the wellbore trajectory parameters and the confidence interval of the wellbore trajectory parameters of the target well, including: The wellbore trajectory parameters and the confidence intervals of the wellbore trajectory parameters of the current well segment to be predicted in the target well are determined by solving the inclination angle observation equation and the state equation, as well as the azimuth angle observation equation and the state equation through particle filtering in the following manner: According to the inclination angle state equation and the azimuth angle state equation, obtain the state parameters of the current well section to be predicted; According to the state parameters of the current well section to be predicted, the inclination angle observation equation and the azimuth angle observation equation are solved to obtain the weighted normalized particle swarm of the current well section to be predicted, the bottom inclination angle of the current well section to be predicted, the current The bottom azimuth of the well section to be predicted; According to the particle swarm normalized by the weight of the current well section to be predicted, the resampling particle swarm of the current well section to be predicted is obtained; Based on the resampling particle swarm of the current to-be-predicted well section, the bottom inclination angle of the current to-be-predicted well section, and the bottom azimuth of the current to-be-predicted well section, obtain the bottom final inclination angle of the current to-be-predicted well section, the current to-be-predicted well section The final azimuth at the bottom of the section, the preset prediction confidence interval for the bottom inclination angle of the current well section to be predicted, and the preset prediction confidence interval for the bottom azimuth of the current well section to be predicted.

Images