A smart drill bit attitude segmented adaptive filtering method, system and storage medium

Abstract

This invention discloses an intelligent drill bit attitude segmented adaptive filtering method, system, and storage medium. The intelligent drill bit attitude segmented adaptive filtering method includes: acquiring multi-source measurement signals from the drill bit and performing preprocessing and dynamic-static separation; extracting features to construct discrimination indices, dividing the working conditions into drilling start-up impact, continuous cutting, and idling separation; calculating acceleration reliability coefficients; establishing a full-state filtering model including attitude and zero bias for attitude prediction and temperature drift compensation; matching measurement update strategies according to identified working conditions; adaptively adjusting the noise covariance matrix based on the reliability coefficients; performing smoothing processing on the preliminary results; and outputting the final attitude parameters. This invention can dynamically switch filtering strategies and adaptively adjust noise weights according to changing working conditions, isolate strong impact interference, overcome zero bias drift caused by high temperatures, and improve the accuracy and robustness of attitude calculation in complex drilling environments.

Description

Technical Field

[0001] This invention relates to the field of downhole measurement data processing, specifically to an intelligent drill bit attitude segmented adaptive filtering method, system, and storage medium. Background Technology

[0002] In the drilling and development of deep and ultra-deep oil and gas resources, the real-time attitude parameters of the drill bit are the core basis for precise wellbore trajectory control and dynamic optimization of drilling parameters. With the continuous development of oil and gas drilling engineering, downhole measurement-while-drilling (MWD) and inertial navigation signal processing technologies are also constantly advancing. Accurately acquiring the drill bit attitude is crucial for improving drilling efficiency, reducing costs, and ensuring drilling safety. Inertial measurement units (IMUs) installed near the drill bit have become key equipment for acquiring the real-time attitude of the drill bit, driving the continuous evolution of related technologies.

[0003] In existing technologies, traditional extended Kalman filtering and unscented Kalman filtering methods are commonly used to solve the drill bit attitude calculation problem. These methods can calculate the drill bit attitude to a certain extent. Traditional extended Kalman filtering approximates the nonlinear system through linearization, thereby achieving state estimation; unscented Kalman filtering uses a deterministic sampling strategy to more accurately describe the nonlinear characteristics of the system. However, these methods all have certain limitations in practical applications.

[0004] However, these traditional methods have significant drawbacks. They employ fixed filtering parameters and uniform measurement strategies, making them unsuitable for adapting to the signal characteristics of different drilling conditions such as start-up impact, stable cutting, and idle separation. In high-impact environments, they are prone to attitude miscorrection and even filter divergence. They assume that acceleration measurements can always be used as a gravity reference without quantitatively evaluating the reliability of acceleration. When non-gravitational acceleration accounts for a large proportion, it leads to severe distortion of attitude calculation results. Furthermore, they fail to deeply integrate temperature drift compensation with the filtering model. Zero-bias temperature drift from gyroscopes and accelerometers causes a continuous increase in accumulated errors during long-term drilling, failing to meet the measurement accuracy requirements of deep and ultra-deep wells. In addition, they do not specifically optimize the tool face angle, a core parameter for directional drilling. During continuous cutting, the tool face angle is prone to non-physical jumps, affecting the precise control of the wellbore trajectory. Simultaneously, the decoupling of the working condition identification and attitude filtering processes prevents dynamic adjustment of the filtering strategy based on the real-time working status of the drill bit, resulting in a tradeoff between anti-interference capability and calculation accuracy. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned technical deficiencies and to propose an intelligent drill bit attitude segmented adaptive filtering method, system and storage medium.

[0006] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: This invention provides a smart drill bit attitude segmented adaptive filtering method, comprising the following steps: S1. Synchronously acquire multi-source measurement signals output by the drill bit or near-drill bit measurement unit, wherein the multi-source measurement signals include at least: triaxial angular velocity signal, triaxial acceleration signal, ambient temperature signal, and drill pressure / load signal; S2. Perform unified time base synchronization, anomaly identification and repair on the acquired multi-source measurement signals to obtain preprocessed signals, and separate the preprocessed triaxial acceleration signal and triaxial angular velocity signal into a trend term characterizing attitude gradual change and a dynamic term characterizing vibration disturbance, respectively. S3. Based on the dynamic terms, drilling pressure / load signals, and ambient temperature signals, extract multi-dimensional working condition features and construct quantitative working condition discrimination indicators. And determine the indicators based on the said working conditions. The working conditions of the drill bit are divided into three categories: drilling start-up impact condition, continuous cutting condition, and idle separation condition. S4. Calculate the acceleration reliability coefficient based on the deviation between the magnitude of the triaxial acceleration signal and the gravitational constant, and the high-frequency vibration energy. ; S5. Establish a full-state filtering model that includes attitude quaternions, gyroscope zero bias parameters and accelerometer zero bias parameters. Complete attitude prediction and update based on three-axis angular velocity signals. At the same time, combine the ambient temperature signal to perform temperature drift compensation for gyroscope zero bias and accelerometer zero bias. S6. Based on the drill bit working status identified in step S3, switch the matching measurement update strategy: turn off or reduce the weight of accelerated measurement updates under drilling impact conditions; introduce low-frequency gravity reference constraints and tool face continuity pseudo-observation constraints under continuous cutting conditions; introduce static stability constraints and zero bias online correction under idling separation conditions. S7. Based on the acceleration confidence coefficient The measurement noise covariance matrix of the full-state filtering model is adaptively adjusted based on the filter residual energy and load fluctuation intensity. With process noise covariance matrix Output the preliminary attitude calculation results; S8. Perform fixed-delay smoothing and quaternion normalization on the preliminary attitude calculation results, and output the final drill bit attitude parameters.

[0007] In some embodiments, in step S1, the multi-source measurement signal further includes an auxiliary electrical signal, which includes a drill bit drive current signal or a drive voltage signal, used to assist in determining the rotation state of the drill bit. In step S2, the separation of the trend term and the dynamic term satisfies the following relationship: in, For the first The three-axis acceleration dynamics at the sampling time For the first Dynamic terms of triaxial angular velocity at the sampling time; For the first The triaxial acceleration signal after preprocessing at the sampling time. For the first The triaxial angular velocity signal after preprocessing at the sampling time; This is the three-axis acceleration sliding trend term. This is a three-axis angular velocity sliding trend term; the sliding trend term passes through a length of... The mean or mean value of the sliding window is calculated.

[0008] In some embodiments, in step S3, the operating condition discrimination index The following relationship must be satisfied: in, The root mean square of the dynamic acceleration. The root mean square of the dynamic angular velocity. As the center of the short-time spectrum, The standard deviation of drilling pressure / load. For the rate of change of load, The rate of temperature change; The preset weighting coefficients satisfy the following conditions: ,in The value range is 0.15 to 0.35. The value range is 0.10 to 0.30. The value range is 0.05 to 0.20. The value range is 0.10 to 0.25. The value range is 0.10 to 0.25. The value range is 0.02 to 0.10; The division of the drill bit's working state adopts a discrimination rule based on a dual-threshold joint duration constraint, satisfying the following relationship: in, Threshold for determining the impact condition at the start of drilling Threshold for continuous cutting conditions The load change rate threshold; The duration of the current operating condition. The minimum stable duration.

[0009] In some embodiments, in step S4, the acceleration confidence coefficient The following relationship must be satisfied: in, For the first The magnitude of the triaxial acceleration signal at the sampling time. The local gravitational acceleration constant is This refers to the high-frequency vibration energy of the triaxial acceleration signal. , The preset adjustment coefficient; The value range is [0,1].

[0010] In some embodiments, in step S5, the state vector of the full-state filtering model Defined as: in, For attitude quaternion components; These are the zero-bias parameters for a three-axis gyroscope; These are the zero-bias parameters for the triaxial accelerometer; The attitude prediction update satisfies the following relationship: in, For the first Predicting attitude quaternions at all times It is a quaternion normalization function; This is the antisymmetric attitude propagation matrix constructed from the angular velocity vector; For the first The three-axis angular velocity signal at any given moment; For the first The zero-bias vector of the gyroscope at any given time; The system sampling period; The zero-bias prediction update of the temperature drift compensation satisfies the following relationship: in, For the first Predict the zero bias vector of the gyroscope at all times. For the first Predict the zero bias vector of the accelerometer at any time; This is the temperature compensation coefficient for the gyroscope. This is the accelerometer temperature compensation coefficient; For the first Ambient temperature at all times; This refers to the noise during the gyroscope's zero-bias process. This refers to the noise during the zero-bias process of the accelerometer.

[0011] In some embodiments, the measurement update strategy for each operating condition in step S6 is specifically as follows: Under drilling impact conditions, the noise covariance of the acceleration measurement is amplified to a preset multiple or the acceleration measurement weight is reset to zero in order to turn off or reduce the weight of acceleration measurement updates. Under continuous cutting conditions, the observation model of the low-frequency gravity reference constraint satisfies the following relationship: in, This is the gravity reference observation vector; The low-frequency component of triaxial acceleration; For gravity reference observation equations; This is the direction cosine matrix obtained from the attitude quaternion transformation; Let be the unit gravity direction vector in the navigation coordinate system, taken in the navigation coordinate system with the gravity direction as the positive z-axis. ; The tool surface continuity pseudo-observation constraint satisfies the following relationship: in, This is a pseudo-observation reference value for the tool face angle; The first , Time-based tool face angle calculation results; , For continuous weighting coefficients, The value range is 0.5 to 0.8. The value range is 0.2 to 0.5, and satisfies the following condition: , making This forms a weighted average estimate of the tool face angle at the first two time points; For tool face angle continuity residual; Under the idling separation condition, the static stability constraint is a full update of the accelerometer measurement based on the pure gravity field; the zero bias online correction is a closed-loop update based on the zero bias convergence characteristic, which converges the gyroscope zero bias and the accelerometer zero bias to the preset nominal range.

[0012] In some embodiments, in step S7, the adaptive adjustment of the measurement noise covariance matrix and the process noise covariance matrix satisfies the following relationship: in, For the first The noise covariance matrix is ​​measured at any given time. The initial measurement noise covariance matrix; For the first The noise covariance matrix at any given time step. The initial process noise covariance matrix; The filter residual vector, This refers to the energy of the filtered residual. Standard deviation of drilling pressure / load; For judging working conditions; , , To measure the noise adjustment coefficient; , This is the process noise adjustment coefficient.

[0013] In some embodiments, in step S8, the fixed-delay smoothing process satisfies the following relationship: in, The smoothed pose quaternion; For a fixed time delay window length; It is a spherical linear interpolation function; It is a unit quaternion; This refers to quaternion multiplication operations; The weighting coefficients are smoothed and satisfy the following conditions: ; When the time delay window length is fixed When the attitude changes are small and the changes in attitude between adjacent time points are small, the smoothing process simplifies to a weighted average approximation: in, To smooth the tool face corners; The final output drill bit attitude parameters include at least the well inclination angle, azimuth angle, and tool face angle, as well as the current operating condition label and attitude confidence score; wherein the attitude confidence score satisfies the following relationship: in, For attitude confidence; , Let be the weight coefficient, and satisfy... .

[0014] This invention also provides a downhole intelligent drill bit attitude segmentation adaptive filtering system, comprising: The sensor unit, installed in the drill bit or near the drill bit section, includes a triaxial gyroscope, a triaxial accelerometer, a temperature sensor, and a drill pressure / load sensor. An embedded processor, electrically connected to the sensor unit, has the following functional modules deployed on it: The data acquisition module is used to synchronously acquire the triaxial angular velocity signal, triaxial acceleration signal, ambient temperature signal, and drilling pressure / load signal output by the sensor unit; The data preprocessing module is communicatively connected to the data acquisition module and is used to perform unified time base synchronization, anomaly identification and repair on the acquired multi-source signals, and separate the preprocessed inertial signals into trend terms and dynamic terms. The working condition identification module is connected in communication with the data preprocessing module. It is used to extract multi-dimensional working condition features, construct working condition discrimination indicators, and divide the working state of the drill bit into drilling start-up impact working condition, continuous cutting working condition and idle separation working condition. The attitude filtering and calculation module is connected to the data preprocessing module and the working condition identification module respectively. It has a built-in full-state filtering model containing attitude quaternions, gyroscope zero bias, and accelerometer zero bias. It is used to perform attitude prediction update, temperature drift compensation, working condition matching measurement update, adaptive noise covariance adjustment, and output preliminary attitude calculation results. The result output module is communicatively connected to the attitude filtering and calculation module. It is used to perform fixed-delay smoothing and quaternion normalization on the preliminary attitude calculation results and output the final drill bit attitude parameters.

[0015] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the downhole intelligent drill bit attitude segmented adaptive filtering method is implemented.

[0016] Compared with existing technologies, the beneficial effects of the intelligent drill bit attitude segmented adaptive filtering method, system, and storage medium provided by this invention include: (1) By extracting multi-dimensional working condition features to construct quantitative working condition discrimination index, the working state of the drill bit is accurately divided into three types of working conditions: drilling start-up impact, continuous cutting and idling separation. Based on this, the matching measurement update strategy is dynamically switched. It can actively cut off the interference source to avoid attitude miscorrection during strong impact and use the gravity field to eliminate the cumulative error during stable idling, thereby improving the robustness and full-process adaptability of the filtering model in complex drilling environment.

[0017] (2) Based on the triaxial acceleration modulus deviation and high-frequency vibration energy, the acceleration reliability coefficient is calculated in real time and combined with the filter residual energy and load fluctuation intensity. The measurement noise covariance matrix and process noise covariance matrix of the full-state filtering model are adaptively and dynamically adjusted, realizing the real-time quantitative evaluation of the effectiveness of the acceleration signal as a gravity reference. When the drill bit vibrates violently and causes a surge in non-gravity acceleration, it can automatically reduce the weight of low signal-to-noise ratio measurements and simultaneously increase the process noise, thus isolating the negative impact of strong vibration interference on attitude calculation, enhancing the system's ability to isolate non-gravity acceleration interference, and avoiding attitude calculation distortion under harsh working conditions.

[0018] (3) A full-state filtering model including attitude quaternions and zero bias of inertial devices is established, and the environmental temperature compensation is deeply integrated into the state prediction link. At the same time, the tool face continuity pseudo-observation constraint is specifically introduced under continuous cutting conditions, and the preliminary solution results are smoothed by fixed time delay. This overcomes the long-term drift of sensors caused by the high temperature environment of deep or ultra-deep wells and reduces the cumulative error of long-term drilling. On the other hand, through pseudo-observation constraint and smoothing, the non-physical jump of tool face angle during continuous cutting is eliminated, ensuring the smooth and stable core control parameters of directional drilling and providing reliable data support for the precise control of wellbore trajectory. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the downhole intelligent drill bit attitude segmented adaptive filtering method provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the working condition identification and classification logic in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the segmented adaptive filtering strategy switching and noise covariance adjustment in Embodiment 1 of the present invention; Figure 4 This is a block diagram of the architecture of the downhole intelligent drill bit attitude segmented adaptive filtering system provided in Embodiment 2 of the present invention; Figure 5 This is a schematic diagram of system deployment and data flow provided in Embodiment 2 of the present invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. The described embodiments are only possible technical implementations of the present invention, but are not limited thereto. Other embodiments obtained by those skilled in the art in conjunction with the embodiments of the present invention without creative effort are also within the protection scope of the present invention.

[0021] Example 1 This invention provides a segmented adaptive filtering method for the attitude of a downhole intelligent drill bit, applicable to deep, ultra-deep, and directional drilling scenarios in oil and gas drilling engineering. This method can be deployed in a measurement system installed on the drill bit body or near a drill bit sub. The measurement system includes a three-axis gyroscope, a three-axis accelerometer, a temperature sensor, a drill pressure / load sensor, and an embedded processor. The system sampling frequency can be configured within the range of 100Hz to 1000Hz depending on the field conditions; in this embodiment, a sampling frequency of 200Hz is preferred to balance attitude calculation accuracy with the real-time computing power of the downhole processor.

[0022] Please refer to Figure 1 The downhole intelligent drill bit attitude segmentation adaptive filtering method provided in this embodiment of the invention includes the following steps: S1. Synchronous acquisition of multi-source measurement signals First, multi-source measurement signals output from the drill bit or near-drill bit measurement unit are acquired synchronously. The multi-source measurement signals include at least: triaxial angular velocity signal, triaxial acceleration signal, ambient temperature signal, and drill pressure / load signal.

[0023] Among them, the three-axis angular velocity signal is output by the three-axis gyroscope and is used to characterize the real-time rotation state and attitude change rate of the drill bit; the three-axis acceleration signal is output by the three-axis accelerometer and is used to provide gravity direction reference and vibration information; the ambient temperature signal is output by the temperature sensor and is used to reflect downhole temperature changes and participate in the temperature drift compensation of inertial devices; the drill pressure / load signal is output by the drill pressure / load sensor and is used to characterize the contact state and force fluctuation between the drill bit and the formation.

[0024] In this embodiment, a unified hardware clock triggering method is preferably used to synchronously sample various sensors to ensure the consistency of multi-source measurement signals on the time axis. Preferably, the time synchronization error of each channel is controlled within 1ms, thereby reducing the fusion error introduced by time base deviation.

[0025] Furthermore, the multi-source measurement signal may also include an auxiliary electrical signal, which includes a drill bit drive current signal or a drive voltage signal, to help determine the rotation state and load changes of the drill bit and improve the accuracy of working condition identification.

[0026] S2, Signal preprocessing and separation of trend / dynamic terms The acquired multi-source measurement signals are synchronized with a unified time base, and anomaly identification and repair are performed to obtain preprocessed signals. The preprocessed triaxial acceleration signals and triaxial angular velocity signals are then separated into a trend term representing attitude change and a dynamic term representing vibration disturbance, respectively.

[0027] Specifically, firstly, the signals of each channel are resampled or interpolated based on a unified time reference, so that the outputs of different sensors are mapped to the same standard sampling time. Then, anomaly identification and repair are performed on the preprocessed signal. Anomalies can be caused by strong vibrations, electromagnetic interference, or transient communication anomalies. A sliding window statistical method is preferably used to identify anomaly samples, and interpolation is performed using adjacent valid samples to suppress the impact of outliers on subsequent filtering.

[0028] After preprocessing, the trend and dynamic terms of the triaxial acceleration and angular velocity signals are separated. The separation satisfies the following relationship: in, For the first The three-axis acceleration dynamics at the sampling time For the first Dynamic terms of triaxial angular velocity at the sampling time; For the first The triaxial acceleration signal after preprocessing at the sampling time. For the first The triaxial angular velocity signal after preprocessing at the sampling time; This is the three-axis acceleration sliding trend term. This represents the sliding trend term of the three-axis angular velocity.

[0029] In this embodiment, the sliding trend term is through a length of The mean value is calculated using a sliding window; in other embodiments, it can also be calculated using the mean value within a sliding window to enhance the suppression capability against pulse-type impulse interference. Preferably, when the system sampling frequency is 200Hz, the mean value is taken as... This corresponds to a window length of approximately 100ms, which effectively balances the preservation of attitude change information and the separation of vibration disturbances.

[0030] S3. Operating condition feature extraction and operating condition classification Please refer to Figure 2 Based on the dynamic terms, drilling pressure / load signals, and ambient temperature signals, multi-dimensional working condition features are extracted to construct quantitative working condition discrimination indicators. And determine the indicators based on the said working conditions. The working conditions of the drill bit are divided into three categories: drilling start-up impact condition, continuous cutting condition, and idle separation condition.

[0031] In this embodiment, the extracted operating condition features include at least: root mean square dynamic acceleration. Root mean square of dynamic angular velocity Short-time spectrum center Standard deviation of drilling pressure / load Load change rate and the rate of temperature change Based on the above characteristics, the following operating condition discrimination index is constructed: in, The preset weighting coefficients satisfy the following conditions: In this embodiment, the weighting coefficient can be set based on historical drilling data statistics and field calibration results. Preferably, , , , , , .

[0032] Furthermore, the working state of the drill bit is determined using a dual-threshold combined duration constraint discrimination rule, satisfying the following relationship: in, Threshold for determining the impact condition at the start of drilling Threshold for continuous cutting conditions The load change rate threshold; The duration of the current operating condition. The minimum stable duration.

[0033] In this embodiment, it is preferred , , , Among them, the drilling start-up impact condition mainly corresponds to the strong impact stage when the drill bit initially contacts the formation; the continuous cutting condition corresponds to the stage of normal drilling with a stable cutting load; and the idle separation condition corresponds to the stage when the drill bit is lifted from the bottom of the well and basically separated from the formation.

[0034] S4. Calculation of Acceleration Reliability Coefficient Based on the deviation between the magnitude of the triaxial acceleration signal and the gravitational constant, and the high-frequency vibration energy, the acceleration reliability coefficient is calculated. .

[0035] During downhole drilling, the accelerometer output includes not only the gravity component but also potentially strong motion acceleration and impact vibration components. Therefore, it is necessary to evaluate in real time whether the acceleration signal is suitable as a gravity reference. The acceleration reliability coefficient satisfies the following relationship: in, For the first The magnitude of the triaxial acceleration signal at the sampling time. The local gravitational acceleration constant is This refers to the high-frequency vibration energy of the triaxial acceleration signal. , The preset adjustment coefficient; The value range is [0,1].

[0036] In this embodiment, it is preferred to take High-frequency vibration energy The mean square energy can be calculated by high-pass filtering the triaxial acceleration signal, for example, using a 20Hz high-pass filter to extract the high-frequency component, and then calculating the mean square energy based on this high-frequency component. Preferably, , .when When the value is close to 1, it indicates that the current acceleration signal is relatively reliable; when it is close to 0, it indicates that the current acceleration signal is greatly affected by non-gravitational disturbances and should not be used directly as a basis for attitude correction.

[0037] S5. Establish a full-state filtering model and perform attitude prediction and temperature drift compensation. A full-state filtering model including attitude quaternions, gyroscope bias parameters, and accelerometer bias parameters is established. Attitude prediction and update are completed based on the three-axis angular velocity signal. At the same time, temperature drift compensation is performed on the gyroscope bias and accelerometer bias by combining the ambient temperature signal.

[0038] In this embodiment, the state vector of the full-state filtering model is defined as: in, For attitude quaternion components; These are the zero-bias parameters for a three-axis gyroscope; These are the zero-bias parameters for the triaxial accelerometer.

[0039] The pose prediction update satisfies the following relationship: in, For the first Predicting attitude quaternions at all times The quaternion normalization function, The attitude propagation matrix is ​​constructed from the angular velocity vector. For the first The three-axis angular velocity signal at any given time. For the first The zero bias vector of the gyroscope at any given time. This is the system sampling period.

[0040] Meanwhile, the zero-bias temperature drift compensation prediction update satisfies the following relationship:

[0041] in, For the first Predict the zero bias vector of the gyroscope at all times. For the first Predict the zero bias vector of the accelerometer at any time; This is the temperature compensation coefficient for the gyroscope. This is the accelerometer temperature compensation coefficient; For the first Ambient temperature at all times; and These are the gyroscope zero-bias process noise and the accelerometer zero-bias process noise, respectively.

[0042] Preferably, the and This can be obtained in advance through factory temperature calibration experiments. By incorporating temperature drift compensation into the condition prediction process, long-term drift errors caused by high-temperature downhole environments can be effectively reduced.

[0043] S6. Measurement update strategy based on operating conditions Please refer to Figure 3 Based on the drill bit working status identified in step S3, the matching measurement update strategy is switched: under the drilling start-up impact condition, the measurement update is turned off or reduced in weight; under the continuous cutting condition, low-frequency gravity reference constraints and tool face continuity pseudo-observation constraints are introduced; under the idling separation condition, static stability constraints and zero bias online correction are introduced.

[0044] 1. Drilling start-up impact condition During drilling impact conditions, the acceleration signal contains a large amount of non-gravitational impact components due to the violent collision between the drill bit and the formation. Therefore, the acceleration measurement noise covariance is amplified to a preset factor, or the acceleration measurement weights are directly reset to zero to disable or deweight acceleration measurement updates. This avoids impact interference being directly introduced into the attitude correction process and improves filtering stability.

[0045] 2. Continuous cutting condition Under continuous cutting conditions, low-frequency gravity reference constraints and tool face continuity pseudo-observation constraints are introduced.

[0046] The observation model of the low-frequency gravity reference constraint satisfies the following relationship: in, This is the gravity reference observation vector. The low-frequency components of triaxial acceleration, For the gravity reference observation equation, The direction cosine matrix is ​​obtained from the attitude quaternion transformation. This is the unit gravity direction vector in the navigation coordinate system. (The text then abruptly shifts to a different topic, seemingly unrelated to the previous sentence.) In the navigation coordinate system along the positive axis, we can take: in, It can be obtained by low-pass filtering the acceleration signal, for example, using a 2Hz low-pass filter, to retain the gravity trend component and suppress high-frequency cutting vibration.

[0047] The tool surface continuity pseudo-observation constraint satisfies the following relationship: in, This is a pseudo-observation reference value for the tool face angle; The first , Time-based tool face angle calculation results; , The weighting coefficients are continuous and satisfy the following conditions: Preferably, , This constraint can reduce non-physical jumps in the tool face angle during continuous cutting, and improve the continuity of control parameter output.

[0048] 3. Idle Separation Condition Under idling separation conditions, the drill bit is essentially separated from the formation and is less affected by cutting loads and impact interference. Therefore, static stability constraints and online zero-bias correction are introduced. The static stability constraints are based on a full update of the accelerometer measurement using a pure gravity field; the online zero-bias correction is based on a closed-loop update using zero-bias convergence characteristics, which gradually converges the gyroscope zero bias and the accelerometer zero bias to a preset nominal range, thereby restoring the system state and reducing accumulated errors.

[0049] S7. Adaptively adjust the noise covariance matrix and output preliminary attitude results. Please refer to Figure 3 Based on the acceleration confidence coefficient The measurement noise covariance matrix of the full-state filtering model is adaptively adjusted based on the filter residual energy and load fluctuation intensity. With process noise covariance matrix Output the preliminary attitude calculation results.

[0050] The adaptive adjustment of the measurement noise covariance matrix and the process noise covariance matrix satisfies the following relationship: in, For the first The noise covariance matrix is ​​measured at any given time. The initial measurement noise covariance matrix; For the first The noise covariance matrix at any given time step. The initial process noise covariance matrix; The filter residual vector, This refers to the energy of the filtered residual. Standard deviation of drilling pressure / load; , , To measure the noise adjustment coefficient; , This is the process noise adjustment coefficient.

[0051] In this embodiment, preferably, , , , , Through the above adaptive adjustment, the filter gain can be changed in real time according to the current operating conditions and measurement reliability, making the system more robust in strong impact environments and more accurate in stable operating conditions.

[0052] S8. Fixed-delay smoothing and result output The preliminary attitude calculation results are subjected to fixed-delay smoothing and quaternion normalization to output the final drill bit attitude parameters.

[0053] The fixed-delay smoothing process satisfies the following relationship: in, The smoothed pose quaternion; For a fixed time delay window length; It is a spherical linear interpolation function; It is a unit quaternion; This refers to quaternion multiplication operations; The weighting coefficients are smoothed and satisfy: In this embodiment, preferably This corresponds to a fixed latency of approximately 50ms. To reduce the computational burden on the downhole processor, when When the values ​​are small and the attitude changes are small in adjacent time steps, the following weighted average approximation can be used: in, This is for smoothing the tool face angle.

[0054] The final output drill bit attitude parameters include at least the inclination angle, azimuth angle, and tool face angle, and can further output the current operating condition label and attitude confidence score. The attitude confidence score satisfies the following relationship: in, For attitude confidence, , Let be the weighting coefficients, and satisfy:

[0055] In this embodiment, preferably , The higher the attitude confidence level, the more reliable the current attitude calculation result.

[0056] Through the above steps, the present invention can dynamically switch the filtering strategy according to the actual working conditions of the drill bit, and combine acceleration reliability assessment, temperature drift compensation, noise covariance adaptive adjustment and fixed time delay smoothing to achieve high-precision and high-robust attitude calculation in complex downhole environments.

[0057] Example 2 Please refer to Figure 4 and Figure 5 This invention also provides a downhole intelligent drill bit attitude segmentation adaptive filtering system for implementing the above method, the system comprising: The sensor unit, installed in the drill bit or near the drill bit section, includes a triaxial gyroscope, a triaxial accelerometer, a temperature sensor, and a drill pressure / load sensor. An embedded processor, electrically connected to the sensor unit, has the following functional modules deployed on it: The data acquisition module is used to synchronously acquire triaxial angular velocity signals, triaxial acceleration signals, ambient temperature signals, and drill pressure / load signals; The data preprocessing module is used to perform unified time base synchronization, anomaly identification and repair on the acquired multi-source signals, and separate the preprocessed inertial signals into trend terms and dynamic terms; The working condition identification module is used to extract multi-dimensional working condition features, construct working condition discrimination indicators, and divide the working state of the drill bit into drilling start-up impact working condition, continuous cutting working condition and idle separation working condition. The attitude filtering and calculation module is used to perform attitude prediction updates, temperature drift compensation, measurement updates for working condition matching, and adaptive noise covariance adjustment, and outputs preliminary attitude calculation results. The results output module is used to perform fixed-delay smoothing and quaternion normalization on the preliminary attitude calculation results and output the final drill bit attitude parameters.

[0058] In this embodiment, the embedded processor is preferably a high-temperature resistant digital signal processor or a high-reliability microcontroller to adapt to the high-temperature, high-pressure, and strong vibration environment downhole. The result output module can be connected to the drilling telemetry transmission unit to send attitude parameters to the ground monitoring platform in real time.

[0059] Example 3 This invention also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the downhole intelligent drill bit attitude segmented adaptive filtering method described in Embodiment 1 above.

[0060] In this embodiment, the computer-readable storage medium can be a high-temperature resistant non-volatile memory, such as Flash memory, EEPROM, or other storage media that can operate stably in the downhole environment. The program can be deployed either in an embedded processor in the near-bit measurement unit or in a surface analysis platform to achieve real-time downhole calculations or offline surface analysis.

[0061] This invention introduces multi-dimensional working condition identification, acceleration reliability assessment, a measurement update strategy matching working conditions, temperature drift compensation, and an adaptive adjustment mechanism for noise covariance, enabling the attitude filtering process to be adjusted in real time according to dynamic downhole working conditions. Compared with traditional methods using fixed parameters and a single measurement update strategy, this invention can significantly reduce the risk of miscorrection under drilling start-up impact conditions, significantly improve tool face angle continuity under continuous cutting conditions, and effectively achieve zero-bias convergence and error recovery under idle separation conditions, thereby improving the overall accuracy, stability, and robustness of downhole attitude measurement.

[0062] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A piecewise adaptive filtering method for the attitude of a downhole intelligent drill bit, characterized in that, Includes the following steps: S1. Synchronously acquire multi-source measurement signals output by the drill bit or near-drill bit measurement unit, wherein the multi-source measurement signals include at least: triaxial angular velocity signal, triaxial acceleration signal, ambient temperature signal, and drill pressure / load signal; S2. Perform unified time base synchronization, anomaly identification and repair on the acquired multi-source measurement signals to obtain preprocessed signals, and separate the preprocessed triaxial acceleration signal and triaxial angular velocity signal into a trend term characterizing attitude gradual change and a dynamic term characterizing vibration disturbance, respectively. S3. Based on the dynamic terms, drilling pressure / load signals, and ambient temperature signals, extract multi-dimensional working condition features and construct quantitative working condition discrimination indicators. And determine the indicators based on the said working conditions. The working conditions of the drill bit are divided into three categories: drilling start-up impact condition, continuous cutting condition, and idle separation condition. S4. Calculate the acceleration reliability coefficient based on the deviation between the magnitude of the triaxial acceleration signal and the gravitational constant, and the high-frequency vibration energy. ; S5. Establish a full-state filtering model that includes attitude quaternions, gyroscope zero bias parameters and accelerometer zero bias parameters. Complete attitude prediction and update based on three-axis angular velocity signals. At the same time, combine the ambient temperature signal to perform temperature drift compensation for gyroscope zero bias and accelerometer zero bias. S6. Based on the drill bit working status identified in step S3, switch the matching measurement update strategy: turn off or reduce the weight of accelerated measurement updates under drilling impact conditions; introduce low-frequency gravity reference constraints and tool face continuity pseudo-observation constraints under continuous cutting conditions; introduce static stability constraints and zero bias online correction under idling separation conditions. S7. Based on the acceleration confidence coefficient The measurement noise covariance matrix of the full-state filtering model is adaptively adjusted based on the filter residual energy and load fluctuation intensity. With process noise covariance matrix Output the preliminary attitude calculation results; S8. Perform fixed-delay smoothing and quaternion normalization on the preliminary attitude calculation results, and output the final drill bit attitude parameters.

2. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 1, characterized in that, In step S1, the multi-source measurement signal also includes an auxiliary electrical signal, which includes a drill bit drive current signal or a drive voltage signal, used to help determine the rotation state of the drill bit. In step S2, the separation of the trend term and the dynamic term satisfies the following relationship: in, For the first The three-axis acceleration dynamics at the sampling time For the first Dynamic terms of triaxial angular velocity at the sampling time; For the first The triaxial acceleration signal after preprocessing at the sampling time. For the first The triaxial angular velocity signal after preprocessing at the sampling time; This is the three-axis acceleration sliding trend term. This is a three-axis angular velocity sliding trend term; the sliding trend term passes through a length of... The mean or mean value of the sliding window is calculated.

3. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 1 or 2, characterized in that, In step S3, the operating condition discrimination index The following relationship must be satisfied: in, The root mean square of the dynamic acceleration. The root mean square of the dynamic angular velocity. As the center of the short-time spectrum, The standard deviation of drilling pressure / load. For the rate of change of load, The rate of temperature change; The preset weighting coefficients satisfy the following conditions: ,in The value range is 0.15 to 0.

35. The value range is 0.10 to 0.

30. The value range is 0.05 to 0.

20. The value range is 0.10 to 0.

25. The value range is 0.10 to 0.

25. The value range is 0.02 to 0.10; The division of the drill bit's working state adopts a discrimination rule based on a dual-threshold joint duration constraint, satisfying the following relationship: in, Threshold for determining the impact condition at the start of drilling Threshold for continuous cutting conditions The load change rate threshold; The duration of the current operating condition. The minimum stable duration.

4. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 3, characterized in that, In step S4, the acceleration confidence coefficient The following relationship must be satisfied: in, For the first The magnitude of the triaxial acceleration signal at the sampling time. The local gravitational acceleration constant is This refers to the high-frequency vibration energy of the triaxial acceleration signal. , The preset adjustment coefficient; The value range is [0,1].

5. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 1, characterized in that, In step S5, the state vector of the full-state filtering model Defined as: in, For attitude quaternion components; These are the zero-bias parameters for a three-axis gyroscope; These are the zero-bias parameters for the triaxial accelerometer; The attitude prediction update satisfies the following relationship: in, For the first Predicting attitude quaternions at all times It is a quaternion normalization function; This is the antisymmetric attitude propagation matrix constructed from the angular velocity vector; For the first The three-axis angular velocity signal at any given moment; For the first The zero-bias vector of the gyroscope at any given time; The system sampling period; The zero-bias prediction update of the temperature drift compensation satisfies the following relationship: in, For the first Predict the zero bias vector of the gyroscope at all times. For the first Predict the zero bias vector of the accelerometer at any time; This is the temperature compensation coefficient for the gyroscope. This is the accelerometer temperature compensation coefficient; For the first Ambient temperature at all times; This refers to the noise during the gyroscope's zero-bias process. This refers to the noise during the zero-bias process of the accelerometer.

6. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 3, characterized in that, In step S6, the measurement update strategy for each working condition is as follows: Under drilling impact conditions, the noise covariance of the acceleration measurement is amplified to a preset multiple or the acceleration measurement weight is reset to zero in order to turn off or reduce the weight of acceleration measurement updates. Under continuous cutting conditions, the observation model of the low-frequency gravity reference constraint satisfies the following relationship: in, This is the gravity reference observation vector; The low-frequency component of triaxial acceleration; For gravity reference observation equations; This is the direction cosine matrix obtained from the attitude quaternion transformation; Let be the unit gravity direction vector in the navigation coordinate system, taken in the navigation coordinate system with the gravity direction as the positive z-axis. ; The tool surface continuity pseudo-observation constraint satisfies the following relationship: in, This is a pseudo-observation reference value for the tool face angle; The first , Time-based tool face angle calculation results; , For continuous weighting coefficients, The value range is 0.5 to 0.

8. The value range is 0.2 to 0.5, and satisfies the following condition: , making This forms a weighted average estimate of the tool face angle at the first two time points; For tool face angle continuity residual; Under the idling separation condition, the static stability constraint is a full update of the accelerometer measurement based on the pure gravity field; the zero bias online correction is a closed-loop update based on the zero bias convergence characteristic, which converges the gyroscope zero bias and the accelerometer zero bias to the preset nominal range.

7. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 4, characterized in that, In step S7, the adaptive adjustment of the measurement noise covariance matrix and the process noise covariance matrix satisfies the following relationship: in, For the first The noise covariance matrix is ​​measured at any given time. The initial measurement noise covariance matrix; For the first The noise covariance matrix at any given time step. The initial process noise covariance matrix; The filter residual vector, This refers to the energy of the filtered residual. Standard deviation of drilling pressure / load; For judging working conditions; , , To measure the noise adjustment coefficient; , This is the process noise adjustment coefficient.

8. The downhole intelligent drill bit attitude segmented adaptive filtering method according to claim 7, characterized in that, In step S8, the fixed-delay smoothing process satisfies the following relationship: in, The smoothed pose quaternion; For a fixed time delay window length; It is a spherical linear interpolation function; It is a unit quaternion; This refers to quaternion multiplication operations; The weighting coefficients are smoothed and satisfy the following conditions: ; When the time delay window length is fixed When the attitude changes are small and the changes in attitude between adjacent time points are small, the smoothing process simplifies to a weighted average approximation: in, To smooth the tool face corners; The final output drill bit attitude parameters include at least the well inclination angle, azimuth angle, and tool face angle, as well as the current operating condition label and attitude confidence score; wherein the attitude confidence score satisfies the following relationship: in, For attitude confidence; , Let be the weight coefficient, and satisfy... .

9. A downhole intelligent drill bit attitude segmented adaptive filtering system, characterized in that, include: The sensor unit, installed in the drill bit or near the drill bit section, includes a triaxial gyroscope, a triaxial accelerometer, a temperature sensor, and a drill pressure / load sensor. An embedded processor, electrically connected to the sensor unit, has the following functional modules deployed on it: The data acquisition module is used to synchronously acquire the triaxial angular velocity signal, triaxial acceleration signal, ambient temperature signal, and drilling pressure / load signal output by the sensor unit; The data preprocessing module is communicatively connected to the data acquisition module and is used to perform unified time base synchronization, anomaly identification and repair on the acquired multi-source signals, and separate the preprocessed inertial signals into trend terms and dynamic terms. The working condition identification module is connected in communication with the data preprocessing module. It is used to extract multi-dimensional working condition features, construct working condition discrimination indicators, and divide the working state of the drill bit into drilling start-up impact working condition, continuous cutting working condition and idle separation working condition. The attitude filtering and calculation module is connected to the data preprocessing module and the working condition identification module respectively. It has a built-in full-state filtering model containing attitude quaternions, gyroscope zero bias, and accelerometer zero bias. It is used to perform attitude prediction update, temperature drift compensation, working condition matching measurement update, adaptive noise covariance adjustment, and output preliminary attitude calculation results. The result output module is communicatively connected to the attitude filtering and calculation module. It is used to perform fixed-delay smoothing and quaternion normalization on the preliminary attitude calculation results and output the final drill bit attitude parameters.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the downhole intelligent drill bit attitude segmented adaptive filtering method as described in any one of claims 1 to 8.

Images