Real-time data management system based on near-bit geological parameter calibration

Abstract

The present application relates to the technical fields of measurement while drilling and data management, in particular to a real-time data management system based on near-bit geological parameter calibration, comprising: a data acquisition module capable of real-time acquisition of original data near the drill bit; an adaptive correction module capable of correcting original geological data, dynamically adjusting the sampling rate and compression strategy of the corrected geological data; an independent depth module capable of extracting axial motion acceleration using original acceleration data and original magnetic force data, obtaining downhole independent depth and error covariance thereof through inertial recursion; a depth alignment module capable of fusing the downhole independent depth and the ground measured depth to obtain optimal depth estimation and depth uncertainty thereof; aligning the corrected geological data with the optimal depth estimation to generate a logging-while-drilling curve with depth uncertainty; and an uncertainty management module capable of quality marking and visualization of the logging-while-drilling curve according to the depth uncertainty.

Description

Technical Field

[0001] This invention relates to the field of measurement while drilling and data management technology, and in particular to a real-time data management system based on near-bit geological parameter calibration. Background Technology

[0002] The core task of near-bit geological parameter calibration is to eliminate the influence of harsh downhole environments (such as instrument eccentricity and drilling fluid interference) on the raw measurement data. Common geological parameters that need to be calibrated include gamma count rate and resistivity.

[0003] In geological resource exploration and development, near-bit geological parameter data is fundamental for achieving precise geological guidance. However, existing technologies still face the following challenges in the real-time management of near-bit geological data: First, the mainstream communication method between downhole and surface is mud pulse transmission, which has extremely low bandwidth (typically only 0.5-12 bps), making it difficult to upload high-sampling-rate raw data in real time. This results in significant delays and loss of detail in the information obtained from the surface. Second, existing tools mostly use fixed algorithms for simple data compression and correction, failing to dynamically adjust processing strategies according to changes in formation characteristics, resulting in low levels of intelligence. Third, the drill pipe depth measured on the surface is affected by factors such as drill string tension, compression, and thermal expansion, leading to a large error compared to the actual downhole depth. Traditional depth tracking methods cannot correct these errors in real time, resulting in inaccurate matching between geological data and depth, affecting the reliability of directional decisions. Finally, existing real-time management systems typically only provide a single depth value, failing to quantify and convey the uncertainty of depth estimation. This makes it difficult for technicians to judge the reliability of the data, easily leading to misjudgments in thin reservoirs.

[0004] Therefore, how to overcome the bottleneck of data transmission, improve the level of intelligence in downhole data processing, and achieve high-precision real-time depth correction and uncertainty management is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] This invention, through the cooperation of an adaptive correction module, an independent depth module, a depth alignment module, and an uncertainty management module, achieves intelligent downhole processing of geological data, real-time depth correction, and quantitative management of depth uncertainty.

[0006] The technical solution proposed in this invention is: a real-time data management system based on near-bit geological parameter calibration, comprising: A data acquisition module is configured to acquire raw data near the drill bit in real time, including raw geological data, raw acceleration data, and raw magnetic data. An adaptive correction module is configured to correct the original geological data and dynamically adjust the sampling rate and compression strategy of the corrected geological data. An independent depth module is configured to extract axial motion acceleration using raw acceleration data and raw magnetic data, and obtain the downhole independent depth and its error covariance through inertial recursion. A depth alignment module is configured to utilize downhole independent depths and surface-measured depths to fuse an optimal depth estimate and its depth uncertainty; and to align the corrected geological data with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty. An uncertainty management module is configured to perform quality marking and visualization of the logging-while-drilling curves based on the depth uncertainty.

[0007] Preferably, the step of correcting the original geological data and dynamically adjusting the sampling rate and compression strategy of the corrected geological data includes: Based on one or more of the multi-physical quantity consistency model, stratigraphic statistical characteristic model and time series prediction model, obtain the predicted values ​​of geological parameters at the current time and their confidence levels; Based on the stratigraphic type identification probability, a basic correction coefficient is dynamically generated, and based on the confidence level and the residual between the predicted and measured correction values ​​of the geological parameters, the fine-tuning correction coefficient is adaptively updated to correct the original geological data and obtain the corrected geological parameters. Feature extraction is performed on the corrected geological parameters to obtain one or more of the multi-scale statistical features and interface detection features, and the sampling rate or compression strategy is dynamically adjusted according to the change rate of the corresponding features.

[0008] Preferably, the step of extracting axial motion acceleration using raw acceleration data and raw magnetic data, and obtaining the independent downhole depth and its error covariance through inertial recursion, includes: The axial motion acceleration is extracted from the raw acceleration data using the attitude calculation results, and the gravity component and the estimated accelerometer zero bias error are deducted. Integrate the axial motion acceleration to obtain the velocity and depth; Construct an error state vector that includes depth error, velocity error, and axial accelerometer bias error; The discrete-time dynamic equation of the error state is established, and the error state is estimated and corrected using a Kalman filter. At the same time, the estimated values ​​of the independent downhole depths and their error covariance are output.

[0009] Preferably, the step of fusing downhole independent depth and surface measured depth to obtain the optimal depth estimate and its depth uncertainty; aligning the corrected geological data with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty includes: Obtain the ground measurement depth and its error variance; An error-state Kalman filter is constructed, and the optimal depth estimate and its depth uncertainty are obtained by fusing the downhole independent depth and the surface measured depth as observations. Based on the time-depth conversion method, the corrected geological parameters are aligned with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty; The observables of the error-state Kalman filter include: At the moment of starting a single section or stopping drilling, the difference between the ground-measured depth and the inertial recursive depth is taken as the depth error observation. When the drill bit is stationary, the inertial recursive velocity is directly used as the velocity error observation.

[0010] Preferably, obtaining the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and physical confidence levels are obtained based on a multi-physical quantity consistency model, including: Using a pre-trained lightweight neural network model, the predicted mean of the current geological parameter is obtained by inverting the measured values ​​of other geological parameters at the same time, and the predicted variance is obtained through error propagation or network output. The confidence level is obtained by normalizing the inverse of the predicted variance. The measured values ​​of the other geological parameters include one or more of resistivity, density, and neutron porosity.

[0011] Preferably, obtaining the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and statistical confidence levels are obtained based on stratigraphic statistical characteristic models, including: Construct an adaptive sliding window that dynamically adjusts the window length based on the formation change rate; Robust statistics of the geological parameters calibrated within the adaptive sliding window are calculated as predicted values, and statistical confidence is calculated based on the coefficient of variation of the data within the window and the window length. The robust statistics include the corrected median of geological parameters.

[0012] Preferably, obtaining the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and time series confidence levels are obtained based on time series prediction models, including: The recursive least squares algorithm is used to update the parameters of the time series prediction model online. The calibrated geological parameters are used to make a one-step prediction of the current time, and the predicted values ​​of the geological parameters and their prediction variance are obtained. The inverse of the obtained prediction variance is normalized and used as the time series confidence level.

[0013] Preferably, the adaptive update of the fine-tuning correction coefficients includes: The correction coefficients are decomposed into basic coefficients and fine-tuning amounts based on stratigraphic identification. The overall confidence score is obtained by weighted fusion of physical confidence, statistical confidence, and time-series confidence. When the overall confidence level is higher than the preset confidence threshold and the residual does not exceed the anomaly threshold, the gradient descent method is used to update the fine-tuning amount, and the update step size is set based on the overall confidence level. A regression constraint term is introduced to prevent long-term drift of the fine-tuning amount.

[0014] Preferably, the step of extracting features from the corrected geological parameters to obtain multi-scale statistical features includes: The multi-scale statistical characteristics of the corrected geological parameters are calculated within an adaptive sliding window. The multi-scale statistical characteristics include one or more of the following: mean, variance, kurtosis, skewness, extreme value difference, and zero-crossing rate. The time-depth conversion method includes: The optimal depth time series is constructed using the optimal depth estimate, and the optimal depth time series is used as the interpolation benchmark. For each geological parameter with a timestamp, locate its immediate and next neighboring points in the optimal depth time series; The depth value at that moment is calculated using linear interpolation, cubic spline interpolation, or piecewise Hermite interpolation based on drilling rate. By considering the temporal correlation of depth error, the depth uncertainty of the geological parameter point is obtained by interpolation using the depth variance of neighboring points and their cross-covariance.

[0015] A computer-readable storage medium storing a computer program that is executed by a processor to implement the functions of the real-time data management system based on near-bit geological parameter calibration described above.

[0016] The beneficial effects of this invention are: 1. This invention, through adaptive correction and intelligent feature extraction downhole, dynamically adjusts the sampling rate and compression strategy of the corrected geological data, and uploads only high-value information, effectively overcoming the limitation of mud pulse transmission bandwidth, enabling the ground to acquire more geologically significant data in real time.

[0017] 2. This invention achieves high-precision continuous depth estimation by utilizing independent downhole depths to obtain optimal depth estimates and correcting systematic errors such as drill string extension and retraction in real time. It provides the uncertainty at each depth point, offering a scientific basis for geological steering decisions and avoiding misjudgments caused by depth errors in thin reservoirs.

[0018] 3. This invention dynamically generates basic correction coefficients based on the probability of stratigraphic type identification, and adaptively updates and fine-tunes the correction coefficients based on the confidence level and the residual between the predicted geological parameter values ​​and the measured correction values. This can effectively identify and suppress abnormal data, avoid the inability of traditional fixed parameter models to adapt to stratigraphic changes, and prevent errors in calibration parameters, thereby improving the robustness of data output. Attached Figure Description

[0019] Figure 1 This is a block diagram of a real-time data management system based on near-bit geological parameter calibration according to the present invention. Detailed Implementation

[0020] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious modifications will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.

[0021] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0022] refer to Figure 1 The technical solution provided by this invention is: a real-time data management system based on near-bit geological parameter calibration, comprising: a data acquisition module, an adaptive correction module, an independent depth module, a depth alignment module, and an uncertainty management module.

[0023] The data acquisition module is configured to acquire raw data near the drill bit in real time, including raw geological data, raw acceleration data, and raw magnetic data. This can be achieved through the following steps: The near-bit measurement sub is drilled along with the drill bit. Multiple integrated sensors on it collect data at a preset acquisition frequency to obtain raw near-bit data, specifically: The raw gamma count rate is acquired using a geological parameter sensor, such as a scintillation crystal gamma detector, at a sampling rate of 50 Hz. Formation resistivity was collected using a multi-frequency resistivity sensor at a sampling rate of 10 Hz. .

[0024] The original specific force is acquired by using motion sensors, such as a triaxial MEMS accelerometer and a triaxial magnetometer, at a sampling rate of 100Hz, to obtain the original scale vector. and geomagnetic field vector . for , , The original specific force in the axial direction, for , , Magnetic force in the axial direction.

[0025] Data collected by geological parameter sensors and motion sensors all include a unified timestamp and are stored in downhole memory in real time.

[0026] The adaptive correction module is configured to correct the raw geological data and dynamically adjust the sampling rate and compression strategy of the corrected geological data. This can be achieved through the following steps: Step 2.1: Based on one or more of the following models—multi-physical quantity consistency model, stratigraphic statistical characteristic model, and time series prediction model—obtain the predicted geological parameters and their confidence levels for the current moment. Specifically: Step 2.11: Obtain predicted geological parameters and physical confidence levels based on a multi-physical quantity consistency model, including: Using a pre-trained lightweight neural network model, the predicted mean of the current geological parameter is obtained by inverting the measured values ​​of other geological parameters at the same time, and the predicted variance is obtained through error propagation or network output.

[0027] In this embodiment, taking gamma counting as an example, a lightweight neural network method is used. This lightweight neural network includes three input layers, a hidden layer (containing eight neurons using the ReLU activation function), and two output layers. For the measured values ​​of other input geological parameters (resistivity, density, neutron porosity), the average of the predicted gamma values ​​is output through the output layers. and logarithmic variance .

[0028] The confidence level is obtained by normalizing the inverse of the prediction variance. ,Right now ; This is the normalized logarithmic variance.

[0029] Step 2.12: Obtain predicted geological parameters and statistical confidence levels based on the stratigraphic statistical characteristic model, including: Construct an adaptive sliding window and dynamically adjust its length based on the formation change rate. Specifically: Construct an adaptive sliding window. Initial adaptive sliding window width That is, the adaptive sliding window includes 5 sample points.

[0030] If the first difference of the calibrated geological parameters exceeds the rate of change threshold Then the adaptive sliding window width will be adjusted to the minimum width. Otherwise, in order to satisfy At the same time, the width of the adaptive sliding window is increased, for example, by gradually adding one sample point, which is the adjusted length of the adaptive sliding window. .

[0031] In this embodiment, the calibrated gamma count rate (geological parameter) is taken as an example. The calibrated gamma count rate is the value obtained after calibrating the original gamma count rate using a preset calibration model, that is: , Indicates the current time as specified. gamma count rate, This represents the correction coefficient. If the first-order difference of the calibrated gamma count rate is greater than the preset value at three consecutive time points... ,Right now Then adjust the adaptive window length to Otherwise, extend the window to an adaptive length.

[0032] Robust statistics of the geological parameters calibrated within the adaptive sliding window are calculated as predicted values, and statistical confidence is calculated based on the coefficient of variation of the data within the window and the window length. Specifically: The median of the geological parameters calibrated within the adaptive sliding window is used as the predicted value.

[0033] In this embodiment, taking the calibrated gamma count rate as an example, the predicted value is... This indicates that the median is being retrieved. This is for all calibrated gamma count rates within the adaptive sliding window.

[0034] Calculate the coefficient of variation within the adaptive window ,in, This represents the standard deviation of the calibrated gamma count rate within an adaptive sliding window. Then statistical confidence level ;in For reference window width, in this embodiment, Taking 10 means including 10 sample points.

[0035] Step 2.13: Obtain the predicted values ​​and time series confidence scores of geological parameters based on the time series prediction model, including: The recursive least squares algorithm is used to update the parameters of the time series prediction model online. A one-step prediction is performed on the current time using calibrated geological parameters to obtain the predicted values ​​of the geological parameters and their prediction variance. Specifically: A first-order autoregressive model was used to predict geological parameters.

[0036] Taking the calibrated gamma count rate as an example, the calibrated gamma count rate predicted by the first-order autoregressive model. ;in , These are the model parameters for a first-order autoregressive model. It is a positive number term.

[0037] Online estimation using the recursive least squares (RLS) algorithm , The forgetting factor is set to 0.98, and the initial covariance matrix is... ,in It is an identity matrix.

[0038] The inverse of the obtained prediction variance is normalized and used as the time series confidence score, i.e. ;in, This represents the predicted variance of geological parameters.

[0039] Step 2.2: Based on the stratigraphic type identification probability, dynamically generate basic correction coefficients, and adaptively update and fine-tune the correction coefficients based on the confidence level and the residuals between the predicted and measured correction values ​​of the geological parameters to correct the original geological data and obtain the corrected geological parameters. This includes the following steps: Step 2.21: Decompose the correction coefficients into basic coefficients and fine-tuning amounts based on stratigraphic identification, specifically as follows: Using a pre-defined real-time lithology classifier (built on a support vector machine, with the input being the mean and variance of the gamma count rate within an adaptive sliding window), the probability of the current stratum belonging to a particular rock type is output, for example, the probability of belonging to sandstone, mudstone, or carbonate rock. , , .

[0040] Read the basic correction coefficients corresponding to different rock formations from the pre-stored correction coefficient database downhole. , , , constitute the basic coefficient ; Correction coefficient , ,in , This is for fine-tuning.

[0041] Step 2.22: Combine the physical confidence score, statistical confidence score, and time series confidence score, and then weight and fuse them to obtain the overall confidence score. .

[0042] Step 2.23: When the overall confidence level is higher than the preset confidence threshold and the residual does not exceed the anomaly threshold, the fine-tuning amount is updated using gradient descent, with the update step size set based on the overall confidence level. Specifically: Calculate residuals ,in, , for Correction coefficients at time points.

[0043] like If a geological upheaval or sensor malfunction occurs, updates are paused, and the fine-tuning amount is maintained. . For sensor noise standard deviation, such as 5 API, like If so, updates will be paused.

[0044] like and Then update the fine-tuning amount: ; ;in , This is the updated fine-tuning amount. Learning rate. . This is a normalization constant, and in this embodiment, it is set to 200.

[0045] Step 2.24: Introduce a regression constraint term to prevent long-term drift of the fine-tuning amount. The regression constraint term is a small, zero-oriented regression term: ; Among them, the regression coefficients .

[0046] The final correction coefficient is obtained as follows The corrected geological parameters are obtained by correcting the calibrated geological parameters using correction coefficients.

[0047] Taking gamma count rate as an example, the corrected Other geological parameters, such as resistivity, density, and neutron porosity, can be referenced in the gamma count rate correction process.

[0048] Step 2.3: Extract features from the corrected geological parameters to obtain multi-scale statistical features, and dynamically adjust the sampling rate or compression strategy according to the rate of change of the corresponding features. This includes the following steps: Step 2.31: Calculate the multi-scale statistical characteristics of the corrected geological parameters within an adaptive sliding window. The multi-scale statistical characteristics include one or more of the following: mean, variance, kurtosis, skewness, extreme value difference, and zero-crossing rate. Kurtosis is used to represent the sharpness of a geological parameter curve. A geological parameter curve is a curve plotted using corrected geological parameters, with the horizontal axis representing time and the vertical axis representing the parameter values. The formula for calculating kurtosis is: The formula for calculating skewness is: .

[0049] The extreme value difference is the difference between the maximum and minimum values ​​of the corrected geological parameter. The zero-crossing rate is the number of times the geological parameter curve crosses its mean line.

[0050] in, To determine the variance of the corrected geological parameters within the adaptive sliding window, This represents the mean of the corrected geological parameters within the adaptive sliding window.

[0051] Step 2.32: Dynamically adjust the sampling rate or compression strategy based on the rate of change of the corresponding features, including: If the rate of change of any two of the following characteristics—mean, variance, kurtosis, skewness, extreme value difference, and zero-crossing rate—is lower than a preset rate of change threshold... Then adjust the sampling rate or compression rate of the corrected geological parameters. Specifically: if Then reduce the sampling rate and compress the uploaded data: reduce the acquisition frequency and only transmit the maximum and minimum values ​​and extreme value differences of the corrected geological parameters to the well. if If the sampling rate is maintained, the corrected geological parameters and their mean, variance, kurtosis, and extreme value difference within the adaptive sliding window are uploaded. The formula for calculating the rate of change is... ; For the first The eigenvalues ​​of the feature (one of the following: mean, variance, kurtosis, skewness, extreme value difference, and zero-crossing rate).

[0052] The independent depth module is configured to extract axial motion acceleration from raw acceleration and magnetic data, and obtain the downhole independent depth and its error covariance through inertial recursion. This can be achieved through the following steps: Step 3.1: Extract the axial motion acceleration from the raw acceleration data using the attitude calculation results, subtracting the gravity component and the estimated accelerometer bias error. Specifically: The original relative force is transformed to the navigation coordinate system using attitude quaternions. ,in The attitude quaternion is determined by the TRIAD (two-vector attitude determination) or Quest (vector observation) algorithm.

[0053] Subtract gravitational acceleration in the navigation coordinate system To obtain the acceleration of motion ; Axial acceleration is ,in ;in The well inclination angle, This is the azimuth angle.

[0054] The axial acceleration after deducting the gravitational component and the estimated accelerometer bias error is: ,in, For the gravitational component, This is the estimated zero bias error of the accelerometer.

[0055] Step 3.2: Integrate the axial motion acceleration to obtain the velocity and depth; velocity ,in Start time; depth .

[0056] Construct an error state vector including depth error, velocity error, and axial accelerometer bias error. ;in, For the actual measured depth, For actual speed measurement, This represents the measured well inclination angle.

[0057] Step 3.3: Establish the discrete-time dynamic equation of the error state, and use a Kalman filter to estimate and correct the error state. Simultaneously, output the estimated values ​​of the independent downhole depths and their error covariance, specifically: Constructing state equations Among them, the state transition matrix noise variance matrix ;in, , , These are the variances of depth noise, velocity noise, and acceleration noise, respectively. The filter period is 1.

[0058] Step 3.4: Perform time and measurement updates for the Kalman filter, and output the predicted error state vector. Correction is achieved through the error state vector. , and , obtained the corrected , and And upload it to the ground.

[0059] The corrected depth is used as an estimate of the independent depth. The error covariance of the depth error and velocity error is calculated using the depth error and velocity error in the error state vector.

[0060] The depth alignment module is configured to fuse downhole independent depths and surface-measured depths to obtain an optimal depth estimate and its depth uncertainty; aligning the corrected geological data with the optimal depth estimate generates a logging-while-drilling curve with depth uncertainty, specifically achieved through the following steps: The ground measurement depth and its error variance are obtained as follows: Ground depth measurements are obtained from the winch encoder; Obtain the depth error and calculate the variance of the ground-measured depth error. The depth error originates from the error state vector. The first item.

[0061] After normalizing the ground measurement depth, depth error, and corrected depth, the data are input into the pre-trained optimal depth estimation model, which outputs the optimal depth estimate. The optimal depth estimation model is constructed based on a convolutional neural network.

[0062] Based on the time-depth conversion method, the corrected geological parameters are aligned with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty, including: The optimal depth time series is constructed using the optimal depth estimates and used as the interpolation benchmark. For example, the optimal depth estimates arranged in chronological order within a borehole inspection cycle constitute an optimal depth time series.

[0063] For each geological parameter with a timestamp, locate its neighboring points before and after the optimal depth in the optimal depth time series; that is, locate the optimal depth estimate corresponding to the time points before and after the timestamp.

[0064] Because the sampling frequency of geological parameters and the frequency of data transmission to the ground affect the depth obtained by the ground system, it results in a series of discrete values. However, for each geological parameter with a timestamp, its depth and uncertainty need to be determined. Therefore, depth interpolation is required.

[0065] Linear interpolation, cubic spline interpolation, or piecewise Hermite interpolation based on drilling rate can be used to calculate the depth value at that moment (corresponding to the geological parameter timestamp); the piecewise Hermite interpolation method is used in this embodiment.

[0066] By considering the time correlation of depth error, and using the depth variances and cross-covariances of neighboring points, the depth uncertainty, i.e., the depth variance, of this geological parameter point can be interpolated. Among them, interpolation weights Time correlation coefficient ;in, Indicates the end time of the borehole inspection cycle. Indicates the interpolation time point. for The depth covariance over time.

[0067] Based on the optimal depth time series after interpolation, a logging-while-drilling curve is generated. The horizontal axis represents depth information, and the vertical axis represents geological parameters that match the depth information. Each depth information includes a depth value and a depth variance.

[0068] The uncertainty management module is configured to perform quality marking and visualization of the logging-while-drilling curves based on the depth uncertainty. Specifically, different depth variances correspond to different colors of points on the logging-while-drilling curves.

[0069] Real-time geological guidance is achieved using logging-while-drilling (LWD) curves. For example, when the LWD curve shows an increase in a certain geological parameter and it is about to reach the depth uncertainty zone (i.e., the lower limit of the set depth variance range), a warning is issued to the technicians that the drill bit may have touched the reservoir edge and is about to enter a different rock formation, thus assisting the technicians in issuing drill bit adjustment instructions.

[0070] The present invention also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the functions of a real-time data management system based on near-bit geological parameter calibration as described above.

[0071] The processes described above with reference to the flowcharts in the embodiments disclosed in this invention can be implemented as computer software programs. The embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the functions defined in the methods of this application. It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wire segments, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless segments, wire segments, optical fibers, RF, etc., or any suitable combination thereof.

[0072] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0073] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved, and the functions and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the stated principles, the implementation of the present invention may have any changes or modifications.

Claims

1. A real-time data management system based on near-bit geological parameter calibration, characterized in that, include: A data acquisition module is configured to acquire raw data near the drill bit in real time, including raw geological data, raw acceleration data, and raw magnetic data. An adaptive correction module is configured to correct the original geological data and dynamically adjust the sampling rate and compression strategy of the corrected geological data. An independent depth module is configured to extract axial motion acceleration using raw acceleration data and raw magnetic data, and obtain the downhole independent depth and its error covariance through inertial recursion. A depth alignment module is configured to fuse downhole independent depths and surface-measured depths to obtain an optimal depth estimate and its depth uncertainty. Align the corrected geological data with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty; An uncertainty management module is configured to perform quality marking and visualization of the logging-while-drilling curves based on the depth uncertainty.

2. The real-time data management system based on near-bit geological parameter calibration according to claim 1, characterized in that, The process of correcting the original geological data and dynamically adjusting the sampling rate and compression strategy of the corrected geological data includes: Based on one or more of the multi-physical quantity consistency model, stratigraphic statistical characteristic model and time series prediction model, obtain the predicted values ​​of geological parameters at the current time and their confidence levels; Based on the stratigraphic type identification probability, a basic correction coefficient is dynamically generated, and based on the confidence level and the residual between the predicted and measured correction values ​​of the geological parameters, the fine-tuning correction coefficient is adaptively updated to correct the original geological data and obtain the corrected geological parameters. Feature extraction is performed on the corrected geological parameters to obtain one or more of the multi-scale statistical features and interface detection features, and the sampling rate or compression strategy is dynamically adjusted according to the change rate of the corresponding features.

3. A real-time data management system based on near-bit geological parameter calibration according to claim 2, characterized in that, The process involves extracting axial motion acceleration using raw acceleration and magnetic data, and obtaining the independent downhole depth and its error covariance through inertial recursion, including: The axial motion acceleration is extracted from the raw acceleration data using the attitude calculation results, and the gravity component and the estimated accelerometer zero bias error are deducted. Integrate the axial motion acceleration to obtain the velocity and depth; Construct an error state vector that includes depth error, velocity error, and axial accelerometer bias error; The discrete-time dynamic equation of the error state is established, and the error state is estimated and corrected using a Kalman filter. At the same time, the estimated values ​​of the independent downhole depths and their error covariance are output.

4. A real-time data management system based on near-bit geological parameter calibration according to claim 3, characterized in that, The optimal depth estimate and its depth uncertainty are obtained by fusing independent downhole depth and surface-measured depth. Align the corrected geological data with the optimal depth estimate to generate logging-while-drilling curves with depth uncertainty, including: Obtain the ground measurement depth and its error variance; An error-state Kalman filter is constructed, and the optimal depth estimate and its depth uncertainty are obtained by fusing the downhole independent depth and the surface measured depth as observations. Based on the time-depth conversion method, the corrected geological parameters are aligned with the optimal depth estimate to generate a logging-while-drilling curve with depth uncertainty; The observables of the error-state Kalman filter include: At the moment of starting a single section or stopping drilling, the difference between the ground-measured depth and the inertial recursive depth is taken as the depth error observation. When the drill bit is stationary, the inertial recursive velocity is directly used as the velocity error observation.

5. A real-time data management system based on near-bit geological parameter calibration according to claim 4, characterized in that, The acquisition of the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and physical confidence levels are obtained based on a multi-physical quantity consistency model, including: Using a pre-trained lightweight neural network model, the predicted mean of the current geological parameter is obtained by inverting the measured values ​​of other geological parameters at the same time, and the predicted variance is obtained through error propagation or network output. The confidence level is obtained by normalizing the inverse of the predicted variance. The measured values ​​of the other geological parameters include one or more of resistivity, density, and neutron porosity.

6. A real-time data management system based on near-bit geological parameter calibration according to claim 5, characterized in that, The acquisition of the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and statistical confidence levels are obtained based on stratigraphic statistical characteristic models, including: Construct an adaptive sliding window that dynamically adjusts the window length based on the formation change rate; Robust statistics of the geological parameters calibrated within the adaptive sliding window are calculated as predicted values, and statistical confidence is calculated based on the coefficient of variation of the data within the window and the window length. The robust statistics include the corrected median of geological parameters.

7. A real-time data management system based on near-bit geological parameter calibration according to claim 6, characterized in that, The acquisition of the predicted geological parameters and their confidence levels at the current moment includes: Geological parameter predictions and time series confidence levels are obtained based on time series prediction models, including: The recursive least squares algorithm is used to update the parameters of the time series prediction model online. The calibrated geological parameters are used to make a one-step prediction of the current time, and the predicted values ​​of the geological parameters and their prediction variance are obtained. The inverse of the obtained prediction variance is normalized and used as the time series confidence level.

8. A real-time data management system based on near-bit geological parameter calibration according to claim 7, characterized in that, The adaptive update of the fine-tuning correction coefficients includes: The correction coefficients are decomposed into basic coefficients and fine-tuning amounts based on stratigraphic identification. The overall confidence score is obtained by weighted fusion of physical confidence, statistical confidence, and time-series confidence. When the overall confidence level is higher than the preset confidence threshold and the residual does not exceed the anomaly threshold, the gradient descent method is used to update the fine-tuning amount, and the update step size is set based on the overall confidence level. A regression constraint term is introduced to prevent long-term drift of the fine-tuning amount.

9. A real-time data management system based on near-bit geological parameter calibration according to claim 8, characterized in that, The process of extracting features from the corrected geological parameters to obtain multi-scale statistical features includes: The multi-scale statistical characteristics of the corrected geological parameters are calculated within an adaptive sliding window. The multi-scale statistical characteristics include one or more of the following: mean, variance, kurtosis, skewness, extreme value difference, and zero-crossing rate. The time-depth conversion method includes: The optimal depth time series is constructed using the optimal depth estimate, and the optimal depth time series is used as the interpolation benchmark. For each geological parameter with a timestamp, locate its immediate and next neighboring points in the optimal depth time series; The depth value at that moment is calculated using linear interpolation, cubic spline interpolation, or piecewise Hermite interpolation based on drilling rate. By considering the temporal correlation of depth error, the depth uncertainty of the geological parameter point is obtained by interpolation using the depth variance of neighboring points and their cross-covariance.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to perform the functions of a real-time data management system based on near-bit geological parameter calibration as described in any one of claims 1-9.

Images