A method and apparatus for measuring the attitude of a rotary steerable drilling tool

Abstract

This invention discloses a method and apparatus for measuring the attitude of a rotary steered drilling tool, belonging to the field of oil drilling parameter measurement technology. It includes establishing a state equation using gravity tool face angle, well inclination angle, azimuth angle, and magnetometer magnetic interference as state variables; establishing an observation equation using accelerometer and magnetometer measurements as observation variables; adaptively adjusting the adjustment coefficient of the observation noise covariance matrix to adjust the observation noise covariance matrix; updating the Kalman gain using the observation noise covariance matrix, Jacobian matrix, and predicted covariance matrix; and obtaining the updated gravity tool face angle, well inclination angle, and azimuth angle through Kalman gain, prior estimation of state variables, observed values ​​of observation variables, and prior estimation of observation variables. This invention achieves data fusion from multiple sensors and adaptively adjusts the weight of the acceleration component in the observation noise covariance matrix, effectively suppressing the interference of high-frequency, strong downhole vibration noise on the estimation of well inclination angle and tool face angle.

Description

Technical Field

[0001] This invention belongs to the field of oil drilling parameter measurement technology, specifically relating to a method and device for measuring the attitude of a rotary steerable drilling tool. Background Technology

[0002] Precise control in rotary steerable drilling technology heavily relies on the real-time and accurate measurement of three core attitude parameters of the downhole tool: gravity tool facet angle, inclination angle, and azimuth angle. In rotary steerable drilling systems, accurate and real-time measurement of these attitude parameters is the prerequisite and core for achieving precise wellbore trajectory control.

[0003] However, the downhole environment is harsh, and tools are subjected to intense vibration, impact, high temperature, and high pressure during drilling, which places extremely high demands on the measurement accuracy of sensors and the robustness of data processing algorithms. Traditional attitude measurement schemes often rely on a single type of sensor, such as using only a triaxial accelerometer to calculate the static gravity tool face angle and well inclination angle, or only a triaxial magnetometer to calculate the azimuth angle. These methods have large errors under vibration, high temperature, and magnetic interference, and cannot meet the requirements for full-attitude, real-time measurement. In other schemes, the three attitude angles are calculated by estimating the triaxial gravity and magnetic components and then using trigonometric functions. However, since trigonometric functions are nonlinear, the process of calculating the attitude angles may introduce unpredictable nonlinear errors, thereby reducing the estimation accuracy.

[0004] Therefore, combined measurement technology integrating data from multiple sensors (such as gyroscopes, accelerometers, and magnetometers) has become an important development direction in this field. However, existing fusion schemes have bottlenecks in model construction. Some schemes have inappropriate selection of state variables, failing to uniformly describe the three core attitude parameters. In particular, when using magnetometers, the magnetometer changes slowly due to interference, making it difficult to filter out interference using conventional Kalman filtering methods. Other schemes have complex model structures (e.g., using high-dimensional nonlinear models or complex parallel filtering structures), resulting in huge computational loads and making it difficult to achieve real-time processing within the limited resources of downhole embedded systems. When dealing with high-frequency dynamic vibrations in downhole, these models generally face a prominent contradiction between computational complexity and real-time performance and estimation accuracy. Complex models have a high computational load, while overly simplified models lack sufficient accuracy, making it difficult to reliably achieve high-precision steering control in actual drilling.

[0005] Secondly, the downhole environment is complex, and sensor measurement noise has a significant impact on the measurement results. In existing technologies, sensor measurement noise is considered to be constant, which leads to inaccurate estimation results.

[0006] Furthermore, the installation position of sensors within the drill string has a decisive impact on attitude measurement accuracy. Existing devices generally use rectangular or hierarchical compartments to install circuits and sensors, making it difficult to position accelerometers and gyroscopes at the drill string's rotation axis. When sensors are eccentrically mounted, the measured acceleration and angular velocities will be superimposed with the centrifugal force and vibration amplification effects caused by the eccentricity. The centrifugal acceleration generated by the eccentricity is superimposed with the gravitational component, making it impossible for the accelerometer to accurately reflect the direction of gravity; the gyroscope output is superimposed with the angular velocity component induced by strong vibration, resulting in cumulative drift in the attitude angular integral. Compensating for eccentric interference requires higher-order nonlinear models or parallel filtering structures, which significantly increases computational load and power consumption, making it difficult to perform high refresh rate attitude calculations in real time on embedded platforms with limited downhole computing power.

[0007] Chinese patent application publication number CN112963093B proposes a dynamic attitude measurement method based on the fusion of a dual triaxial magnetometer and an optoelectronic positioning sensor. This scheme measures the spatial magnetic field components using two magnetometers and combines this with positioning information obtained by the optoelectronic sensor when triggered at a specific gravity tool face angle. An analytical relationship between the magnetic field measurements and attitude parameters is established, and the analytical solutions for the gravity tool face angle, well inclination angle, and azimuth angle are obtained by directly solving the equations. The advantage of this method is that it avoids the use of vibration-sensitive accelerometers, attempting to avoid iterative calculations through purely analytical computation. The disadvantages of this scheme are: although it avoids the use of accelerometers, its complete reliance on magnetic measurements causes a significant decrease in measurement accuracy or even failure when downhole magnetic interference (such as drill collar magnetization or interference from adjacent wells) is present. Furthermore, this method depends on the precise installation of a sophisticated optoelectronic positioning mechanism and magnetometers, resulting in a complex and costly system; its analytical solution process cannot further suppress random noise through filtering techniques, making it unsuitable for complex operating conditions where vibration and magnetic noise coexist. Furthermore, this method fails to incorporate the high dynamic characteristics of the gyroscope, and the dynamic response performance may degrade due to photoelectric sampling delay when the tool rotates at high speed.

[0008] Chinese invention patent CN106907142B proposes a dynamic measurement device and method for near-drill bit azimuth. The core of this scheme lies in using three sensors—one axially mounted accelerometer and two radially mounted magnetometers—to measure the dynamic azimuth angle. The technical approach involves using the axial accelerometer to measure the axial component of gravitational acceleration to calculate the wellbore inclination angle, and using the two magnetometers to measure the two radial components of the geomagnetic field. Combined with the known local total geomagnetic field strength and magnetic inclination, the axial magnetic field component of the drill string is indirectly calculated, thus avoiding axial magnetic interference from drill string magnetization. Simultaneously, because it does not rely on radial accelerometers, it also avoids interference from rotational centrifugal acceleration and radial vibration. The disadvantages of this scheme are: while this technology improves integration and reduces power consumption by simplifying the sensors, its method has inherent limitations. This method fails to integrate gyroscope data and cannot provide independent angular velocity information for auxiliary correction when the drill string rotates at high speed or when magnetic interference is present. Furthermore, this scheme can only achieve dynamic measurement of the azimuth angle and does not provide real-time calculation capabilities for the gravity tool face angle and well inclination angle. Therefore, it cannot meet the complete requirements of the rotary steering system for closed-loop control of the entire attitude (well inclination angle, azimuth angle, gravity tool face angle). Although its algorithm avoids specific disturbances, it does not construct a universal state estimation model that can optimally fuse multi-sensor data.

[0009] In summary, existing technologies generally suffer from problems such as the inability to uniformly estimate the three attitude angles in the same model, the complexity of algorithm calculations making it difficult to achieve real-time downhole processing, and the difficulty in guaranteeing the accuracy of attitude calculation in complex magnetic interference environments. Summary of the Invention

[0010] To address the technical problems in the prior art, this invention provides a method and apparatus for measuring the attitude of rotary steered drilling tools. It integrates multiple downhole sensors to establish a model and dynamically adjusts the influence of noise on the estimation results, thereby obtaining accurate measurement data.

[0011] This invention addresses the problems of strong vibration and impact, complex magnetic interference, and decreased measurement accuracy due to sensor misalignment during downhole drilling. It constructs a unified technical system from three levels: model building, algorithm mechanism, and device structure. It proposes a rotary steered drilling tool attitude measurement algorithm and device based on multi-sensor data fusion, achieving real-time estimation of three key attitude parameters: gravity tool face angle, well inclination angle, and azimuth angle. This invention offers the following advantages:

[0012] A method for measuring the attitude of a rotary steerable drilling tool, the rotary steerable drilling tool including an accelerometer, a gyroscope, and a magnetometer, the method comprising:

[0013] State equations are established using the gravity tool face angle, well inclination angle, azimuth angle, and magnetometer magnetic interference as state variables, and observation equations are established using accelerometer measurements and magnetometer measurements as observation variables.

[0014] The observation noise covariance matrix is ​​adaptively adjusted by multiplying the acceleration component in the initial value of the observation noise covariance matrix with the adjustment coefficient.

[0015] The Kalman gain is updated using the observation noise covariance matrix, Jacobian matrix, and prediction covariance matrix. The updated gravity tool face angle, well inclination angle, and azimuth angle are obtained by using the Kalman gain, prior estimation of state variables, observed values ​​of observed variables, and prior estimation of observed variables.

[0016] Furthermore, establishing the state equations includes:

[0017] Gravity tool face angle and stable platform speed The relationship is:

[0018] ;

[0019] in, For time indexing, Indicates the sampling time. Noise from gravity tool face angle system; well inclination angle Azimuth , Shaft magnetic interference , Shaft magnetic interference and Shaft magnetic interference Modeled as a random walk process:

[0020] ;

[0021] in For well inclination angle system noise, For azimuth system noise, for Shaft magnetic interference system noise, for Shaft magnetic interference system noise, for The noise of the shaft magnetic interference system is as follows:

[0022] ;

[0023] in, , For the system matrix, For the input matrix, Let be the system process noise vector, and its covariance matrix be... , The system process noise variance represents the angle of the gravity tool face. The system process noise variance representing the well inclination angle, The variance of system process noise represents the azimuth angle. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference; Indicates the first Rotational speed at any given moment;

[0024] Using the measurements from the accelerometer and magnetometer as observed variables, the observation equation is:

[0025] ;

[0026] in, The local magnetic inclination, It is the acceleration due to gravity. The local magnetic field strength, The observation noise vector has the following covariance matrix: , Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft.

[0027] Furthermore, the updated observation noise covariance matrix is:

[0028] ;

[0029] in This is for adjusting the coefficient.

[0030] Furthermore, the adjustment coefficients of the observation noise covariance matrix are obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration:

[0031] Relative deviation is The adjustment coefficient is ;in, The current acceleration magnitude, If the magnitude of the local gravitational acceleration is given, then the observation noise covariance matrix is:

[0032] .

[0033] Furthermore, the adjustment coefficients of the observation noise covariance matrix are obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration:

[0034] Relative deviation is Based on the magnitude of the relative deviation, the interference is divided into several levels, and the corresponding adjustment coefficient is obtained according to the level. Furthermore, the Kalman gain is updated using the observed noise covariance matrix, Jacobian matrix, and predicted covariance matrix. The Kalman gain is calculated as follows:

[0035] ;

[0036] in, For Jacobian matrices, To predict the covariance matrix, To observe the noise covariance matrix; the state variables are updated as follows:

[0037] ;

[0038] in For prior estimates of state variables, , For the observed values ​​of the observed variable, For the reason The prior estimates of the observed variables are calculated. For the system matrix, The input matrix is ​​denoted as .

[0039] Furthermore, the predicted covariance matrix is ​​obtained using the system matrix, the state estimation covariance matrix from the previous time step, and the system process noise covariance matrix. The predicted covariance matrix is:

[0040] ;

[0041] in For the system matrix, Let be the covariance matrix of the system process noise;

[0042] The state estimation covariance matrix at the current time is obtained using the Kalman gain, Jacobian matrix, and prediction covariance matrix:

[0043] ;

[0044] in To predict the covariance matrix.

[0045] A rotary steerable drilling tool attitude measurement device includes a pressure-resistant outer cylinder and a main control module. It also includes a triaxial accelerometer, a gyroscope, and a magnetometer coaxially mounted with the pressure-resistant outer cylinder. The angular velocity meter data, accelerometer data, and magnetometer data are transmitted to the main control module, which implements the measurement method described above.

[0046] Furthermore, the measuring device also includes an installation frame, an angular velocity measurement module, an acceleration measurement module, and a magnetometer installed inside the pressure-resistant outer cylinder. The angular velocity measurement module and the acceleration measurement module are inserted into the installation frame perpendicular to its central axis and are fixedly connected to the installation frame. The accelerometer is installed at the center of the acceleration measurement module, and the x-axis of the triaxial accelerometer is coaxial with the pressure-resistant outer cylinder. The gyroscope is installed at the center of the angular velocity measurement module and is coaxial with the pressure-resistant outer cylinder. The magnetometer is installed at the end of the installation frame away from the main control module, the angular velocity measurement module, and the acceleration measurement module. The angular velocity measurement module, the acceleration measurement module, and the magnetometer are connected to the main control module to transmit angular velocity data, acceleration data, and magnetometer data to the main control module.

[0047] Furthermore, the mounting frame is coaxial with the pressure-resistant outer cylinder, and two mounting slots are formed on the circumferential surface of the mounting frame perpendicular to the central axis. The acceleration measurement module and the angular velocity measurement module are installed in the mounting slots by fixing blocks. The fixing blocks protrude outward in the circumferential direction to form a fixing part. The angular velocity measurement module and the acceleration measurement module are fixed to the fixing block by fixing parts. The fixing block is fixedly connected to the side wall of the mounting slot by the fixing part. A third mounting slot is provided at the end of the mounting frame away from the main control module, the angular velocity measurement module and the acceleration measurement module. The magnetometer is installed in the third mounting slot.

[0048] First, a full attitude estimation model incorporating magnetic interference was constructed, effectively solving the measurement challenges in magnetic environments. This invention establishes a nonlinear state-space model with gravity tool facet angle, well inclination angle, azimuth angle, and triaxial magnetic interference as unified state variables. This model uses slowly changing drill string magnetization interference or environmental magnetic interference as state variables for real-time estimation and compensation, enabling the system to calculate accurate azimuth angles even under conditions of continuous magnetic interference, thus improving the system's adaptability and measurement accuracy in complex magnetic formation environments. This model not only fully describes the dynamic attitude changes of the drill string but also models complex downhole environmental magnetic interference as a system state, utilizing the data redundancy characteristics of magnetometers and gyroscopes for real-time estimation. This unified modeling approach enables data fusion from multiple sensors, providing an accurate mathematical description for subsequent optimal filtering estimation in multi-disturbance environments.

[0049] Secondly, based on the aforementioned mathematical model, this invention proposes an adaptive extended Kalman filter algorithm based on acceleration interference detection to improve the estimation accuracy of the system under strong vibration environments. To address the sudden, strong vibration noise that is difficult to quantify in the model, this invention introduces an acceleration interference detection and adaptive compensation mechanism into the filtering algorithm. The algorithm monitors the deviation between the accelerometer measurement magnitude and the standard gravity in real time. Once strong vibration is detected (abnormal acceleration magnitude), it automatically determines that the reliability of the current accelerometer data has decreased, and dynamically reduces the impact of accelerometer data on attitude updates by adaptively increasing the weight of the acceleration component in the observation noise covariance matrix, instead relying more on the angular velocity integral information from the gyroscope. This mechanism effectively suppresses the interference of high-frequency strong vibration noise from downhole on the estimation of wellbore inclination angle and tool face angle, ensuring the smoothness and stability of attitude calculation.

[0050] Finally, to facilitate the implementation of the aforementioned measurement algorithm, this invention designs a measurement device with a centralized sensor mounting system, effectively reducing the impact of eccentricity at the physical level. This device physically integrates a triaxial accelerometer and a single-axis gyroscope at the geometric center of the drill string's cross-section, ensuring that their sensitive axes are highly aligned with the drill string's rotation axis. This design reduces the centrifugal acceleration and tangential vibration interference caused by eccentric mounting at the physical source, not only lowering the noise level of the raw data but also avoiding the introduction of complex high-order eccentricity compensation models into the algorithm. This reduces the computational load on the embedded system, ensuring the efficient real-time operation of the high-precision measurement algorithm on the downhole hardware platform.

[0051] In summary, this invention, through the dual algorithm guarantee of "magnetic interference state estimation" and "vibration interference adaptive suppression" and the hardware support of "centralized physical layout", realizes dynamic attitude measurement of rotary steerable drilling tools under harsh downhole conditions. Attached Figure Description

[0052] Figure 1 This is a flowchart of the rotary steerable drilling tool attitude measurement method of the present invention;

[0053] Figure 2 This is a flowchart illustrating a specific implementation of the present invention;

[0054] Figure 3 The accelerometer X-axis signal curve;

[0055] Figure 4 The accelerometer Y-axis signal curve;

[0056] Figure 5 The accelerometer Z-axis signal curve;

[0057] Figure 6 The X-axis signal curve of the magnetometer;

[0058] Figure 7This is the Y-axis signal curve of the magnetometer;

[0059] Figure 8 This is the Z-axis signal curve of the magnetometer;

[0060] Figure 9 The curve of gyroscope angular velocity measurement;

[0061] Figure 10 Comparison curves of the estimation results for the face angle of gravity tools;

[0062] Figure 11 A comparison curve showing the absolute error of the gravity tool face angle estimation results;

[0063] Figure 12 Comparison curves of well inclination angle estimation results;

[0064] Figure 13 The curve shows the comparison of the absolute error of the well inclination angle estimation results;

[0065] Figure 14 Comparison curves of azimuth estimation results;

[0066] Figure 15 This is a curve showing the absolute error of the azimuth estimation results.

[0067] Figure 16 Comparison curves of X-axis magnetic interference estimation results;

[0068] Figure 17 The curve showing the absolute error of the X-axis magnetic interference estimation results is shown.

[0069] Figure 18 Comparison curves of Y-axis magnetic interference estimation results;

[0070] Figure 19 The curve showing the absolute error comparison of the Y-axis magnetic interference estimation results;

[0071] Figure 20 Comparison curves of Z-axis magnetic interference estimation results;

[0072] Figure 21 The curve showing the absolute error comparison of the Z-axis magnetic disturbance estimation results;

[0073] Figure 22 This is a schematic diagram of the device structure of the present invention;

[0074] Figure 23 This is an exploded view of the device structure of the present invention;

[0075] Figure 24 This is a schematic diagram of the mounting frame structure of the device of the present invention;

[0076] Figure 25 This is a diagram showing the module connections of the device of the present invention;

[0077] Wherein: 1-Pressure-resistant outer cylinder; 2-Mounting frame; 201-Third mounting slot; 202-Magnetometer; 203-External thread; 204-Angular velocity measurement module; 205-Second fixing block; 206-Acceleration measurement module; 207-First fixing block; 208-Main control module; 209-Fourth mounting slot; 210-First mounting slot; 211-Second mounting slot; 212-Acceleration meter; 213-Gyroscope. Detailed Implementation

[0078] To address the problems in model building during data processing in existing technologies, this invention addresses the issues in rotary steered drilling tools, which include accelerometers, gyroscopes, and magnetometers. This invention proposes a method for measuring the attitude of rotary steered drilling tools, which involves modeling the accelerometers, gyroscopes, and magnetometers, fusing multi-sensor data, and processing the sensor measurement data. The method is as follows: Figure 1 As shown, it includes:

[0079] S101: Establish state equations using gravity tool face angle, well inclination angle, azimuth angle, and magnetometer magnetic interference as state variables, and establish observation equations using accelerometer and magnetometer measurements as observation variables.

[0080] Based on the principle of rotary steerable drilling, the gravity tool face angle can be derived by integrating the rotational speed of the stable platform. and stable platform speed The relationship is:

[0081] ;

[0082] in, For time indexing, Indicates the sampling time. Noise from gravity tool face angle system; well inclination angle and azimuth Modeled as a random walk process:

[0083] ;

[0084] in For well inclination angle system noise, This refers to the azimuth system noise.

[0085] In rotary steerable drilling operations, the rate of change of well inclination angle and azimuth angle is usually relatively small and difficult to model. Therefore, this invention models them as a random walk process.

[0086] Drilling tools are equipped with magnetometers to measure the surrounding magnetic field. The drilling environment may be subject to unknown magnetic interference, including magnetic interference from the drill string itself and that from adjacent casing. Because drilling speeds are slow, the magnetic interference from adjacent casing also changes very slowly. Therefore, most magnetic interference is a slowly changing constant value, rather than Gaussian white noise like accelerometer noise, making it impossible to filter out with Kalman filters. To estimate the magnetic field, this invention utilizes the slow-changing nature of magnetic interference. Drill string magnetic interference and adjacent casing magnetic interference can be vector-decomposed onto the three axes of the magnetometer. Shaft magnetic interference , Shaft magnetic interference and Shaft magnetic interference Modeled as a random walk model:

[0087] ;

[0088] in, for Shaft magnetic interference system noise, for Shaft magnetic interference system noise, for Shaft magnetic interference system noise;

[0089] State equations are established by taking the gravity tool face angle, well inclination angle, azimuth angle, and triaxial magnetic disturbance as state variables:

[0090] ;

[0091] in, , For the system matrix, For the input matrix, Let be the system process noise vector, and its covariance matrix be... , The system process noise variance represents the angle of the gravity tool face. The system process noise variance representing the well inclination angle, The variance of system process noise represents the azimuth angle. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference; Indicates the first Rotation speed at any given moment.

[0092] The accelerometer and magnetometer measurements are stable over long periods with minimal drift. Using the accelerometer and magnetometer measurements as observation variables, the observation equation is:

[0093] ;

[0094] in, The local magnetic inclination, It is the acceleration due to gravity. The local magnetic field strength, The observation noise vector has the following covariance matrix: , Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft.

[0095] S102: Adaptively adjust the adjustment coefficient of the observation noise covariance matrix. Update the observation noise covariance matrix by multiplying the acceleration component in the initial value of the observation noise covariance matrix with the adjustment coefficient.

[0096] The adjustment coefficient of the observation noise covariance matrix is ​​obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration. Specifically, the current acceleration magnitude is calculated as follows:

[0097] ;

[0098] Calculate the relative deviation:

[0099] .

[0100] Calculate the adjustment factor using relative deviation Then the observed noise covariance matrix is ​​updated as follows:

[0101] .

[0102] Specifically, set the initial values ​​for each parameter: Initial state estimation: , set as Or set to the attitude angle measured statically.

[0103] Initial error covariance matrix: It is usually set as an identity matrix.

[0104] Process noise covariance matrix: ;

[0105] Observation noise covariance matrix: .

[0106] Adjust the observation noise covariance matrix in real time based on the acceleration disturbance intensity:

[0107] .

[0108] The deviation of acceleration measurement values ​​is used to compensate for the observation noise covariance matrix, so that the observation noise covariance matrix can be adjusted in real time according to the acceleration measurement values. Moreover, the amount of calculation is small, which meets the needs of engineering implementation.

[0109] In another embodiment, the adjustment coefficient of the observation noise covariance matrix is ​​obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration:

[0110] Relative deviation is Based on the magnitude of the relative deviation, the interference is divided into several levels, and the corresponding adjustment coefficient is obtained according to the level. .

[0111] Specifically, based on relative deviation, they are divided into the following levels:

[0112] ;

[0113] in The threshold for moderate interference is... The threshold for severe interference, and the adjustment factor for moderate interference. Severe interference adjustment coefficient The observation noise covariance matrix is ​​then adjusted according to different levels as follows:

[0114] .

[0115] This embodiment adjusts the observation noise covariance matrix piecewise according to the magnitude of interference. When the acceleration magnitude changes only slightly, the observation noise covariance matrix does not need to be updated, while still achieving high measurement accuracy and reducing downhole data computation. This embodiment classifies interference into three levels, but other levels can also be used depending on the actual situation. These are constants set according to actual conditions.

[0116] S103: Update the Kalman gain using the observation noise covariance matrix, Jacobian matrix, and prediction covariance matrix. The updated gravity tool face angle, well inclination angle, and azimuth angle are obtained through the Kalman gain, prior estimation of state variables, observed values ​​of observed variables, and prior estimation of observed variables.

[0117] Specifically, prior estimates are calculated from the state equations. :

[0118] ;

[0119] Calculate the prediction covariance matrix:

[0120] ;

[0121] Kalman gain calculation:

[0122] Calculate the Jacobian matrix:

[0123] .

[0124] The Kalman gain is updated using the observation noise covariance matrix, Jacobian matrix, and prediction covariance matrix, including the Kalman gain calculation as follows:

[0125] ;

[0126] in, For Jacobian matrices, To predict the covariance matrix, To observe the noise covariance matrix.

[0127] Status Update:

[0128] ;

[0129] For the observed values ​​of the observed variable, For the reason The prior estimates of the observed variables are calculated. For the system matrix, The input matrix is ​​denoted as .

[0130] The covariance matrix at the current time step is updated using the Kalman gain, Jacobian matrix, and the covariance matrix at the current time step. The specific calculation is as follows:

[0131] .

[0132] In this invention, all three attitude angles are modeled as state variables, allowing for simultaneous estimation of all three angles while avoiding the introduction of additional nonlinear errors. This technical solution uses attitude angles and magnetic interference as state variables, employing a Kalman filter to estimate these variables. This solves the problem of the Kalman filter being unable to handle the slow changes in magnetic interference. By removing magnetic interference from the magnetometer, accurate magnetometer measurements are obtained, thus improving the magnetometer's measurement accuracy.

[0133] The following uses specific data to further illustrate the solution of the present invention.

[0134] The core algorithm of this invention is executed by the DSP processor. This algorithm, based on the ordinary extended Kalman filter (EKF), adds acceleration disturbance detection (i.e., vibration detection) and an adaptive compensation mechanism. The following is in conjunction with the appendix... Figure 2 (Data fusion algorithm flowchart) Detailed explanation of its algorithm implementation process:

[0135] S201: System Power-On and Initialization

[0136] (1) Hardware module initialization

[0137] After the DSP powers on, it first initializes the system clock, interrupt system, and AD acquisition module, and configures the sampling channels (corresponding to the three-axis accelerometer, three-axis magnetometer, and single-axis gyroscope) and sampling frequency (e.g., ...). Corresponding sampling time (and resolution).

[0138] (2) Initialization of the extended Kalman filter

[0139] a. Initial value of state vector: Based on static measurement or default attitude setting after the system is powered on.

[0140] ;

[0141] in The initial gravity tool face angle, The initial well inclination angle, The initial azimuth angle, These are the initial triaxial magnetic interferences.

[0142] b. Initial error covariance matrix: Represents the uncertainty of the initial state estimate, and is usually set as a diagonal matrix.

[0143] ;

[0144] in, These are process noises from gravity tool face angle, well inclination angle, azimuth angle, and triaxial magnetic interference, respectively.

[0145] c. Process noise covariance matrix: set according to the characteristics of drill bit attitude change.

[0146] ;

[0147] d. Measurement noise covariance matrix: set according to sensor characteristics.

[0148] ;

[0149] The first three diagonal elements represent the accelerometer noise variance, and the last three represent the magnetometer noise variance.

[0150] S202: Real-time Data Acquisition and Filtering Loop

[0151] After the DSP completes initialization, the system enters a real-time data acquisition and filtering loop, performing the following steps in each sampling cycle:

[0152] (1) Real-time sensor data acquisition

[0153] The analog voltage signals output from the triaxial accelerometer, triaxial magnetometer, and single-axis gyroscope are synchronously acquired by the AD acquisition module and converted into digital signals to obtain the measurement value vector at the current moment. and and the angular velocity measurement value of a single-axis gyroscope. ,in:

[0154] ;

[0155] (2) Acceleration disturbance detection and adaptive adjustment

[0156] The current acceleration interference level is determined based on accelerometer measurements, and the observation noise covariance matrix is ​​adaptively adjusted. :

[0157] a. Calculate the magnitude of acceleration:

[0158] ;

[0159] b. Calculate the relative acceleration deviation:

[0160] ;

[0161] c. Adaptive adjustment This embodiment uses an adjustment coefficient. The calculation method is as follows:

[0162] ;

[0163] (3) State prediction

[0164] Calculate the prior state estimate based on the state equation. and predict covariance matrix :

[0165] ;

[0166] ;

[0167] in,

[0168] .

[0169] (4) Observation update

[0170] Calculate the observed predicted values:

[0171] ;

[0172] Calculate the Jacobian matrix :

[0173] ;

[0174] Where the observation function The expression is:

[0175] ;

[0176] Calculate the Kalman gain:

[0177] ;

[0178] (5) State update and covariance update

[0179] Status Update:

[0180] ;

[0181] in, This is the current observation vector.

[0182] Covariance update:

[0183] ;

[0184] (6) Attitude angle output

[0185] From the updated state vector Extract three attitude angles:

[0186] .

[0187] To simulate the strong vibrations caused by the drill bit impacting rocks under real drilling conditions, a random variance of 1 / 2 was introduced. The strong interference acceleration is as follows:

[0188] Sampling time The actual magnetic field strength Magnetic tilt angle The variance of the triaxial accelerometer measurements is The variance of the gyroscope measurements is The variance of the magnetometer measurements is The three-axis fixed magnetic interferences are respectively .

[0189] Kalman filter settings:

[0190] Initial error covariance matrix ,

[0191] Process noise covariance matrix ,

[0192] Measurement noise covariance matrix .

[0193] The results are as follows Figure 3-21 As shown in Tables 1 and 2, it can be seen that the method proposed in this invention has higher accuracy in parameter estimation compared with the prior art.

[0194] Table 1. Performance Comparison of Attitude Angle Estimation (RMSE)

[0195] method Gravity tool facet angle (°) Well inclination angle (°) Azimuth (°) Ordinary EKF 0.9938 0.3545 1.4541 This invention 0.7626 0.2867 0.8454

[0196] Table 2 Comparison of Magnetic Interference Estimation Performance (RMSE)

[0197] method X-axis (Gauss) Y-axis (Gauss) Z-axis (Gauss) Ordinary EKF 0.08222 0.07797 0.08588 This invention 0.08135 0.07541 0.08492

[0198] The aforementioned measurement method is applied to rotary steerable drilling tools to measure their attitude angles. However, in the prior art, due to the eccentric mounting of the accelerometer and gyroscope, the measured acceleration and angular velocity will be superimposed with the centrifugal force generated by the eccentricity, resulting in a vibration amplification effect. Therefore, this invention also provides a rotary steerable drilling tool attitude measurement device.

[0199] like Figure 22 and 25 As shown, the device includes a pressure-resistant outer cylinder 1, a main control module 208, and a triaxial accelerometer 212, a gyroscope 213, and a magnetometer 202 coaxially mounted with the pressure-resistant outer cylinder 1. The gyroscope data, accelerometer data, and magnetometer data are transmitted to the main control module 208, which implements the measurement method described above.

[0200] Specifically, the main control module 208 includes a power supply and a processor. The power supply provides power to the processor, accelerometer 212, gyroscope 213, and magnetometer 202. The processor can be a DSP, MCU, or other processor that meets performance requirements. The processor receives signals from the accelerometer 212, gyroscope 213, and magnetometer 202, processes them using the measurement method described above, and calculates and outputs the three key attitude parameters of the drill string: gravity tool face angle, well inclination angle, and azimuth angle. The accelerometer 212 and gyroscope 213 are coaxially connected to the pressure-resistant outer cylinder 1, avoiding the generation of centrifugal force and the amplification of vibration.

[0201] The measuring device also includes a mounting frame 2, an angular velocity measuring module 204, and an acceleration measuring module 206 installed inside the pressure-resistant outer cylinder 1. The angular velocity measuring module 204 and the acceleration measuring module 206 are inserted into the mounting frame 2 perpendicular to the central axis of the mounting frame 2 and are fixedly connected to the mounting frame 2. The accelerometer 212 is installed at the center of the acceleration measuring module 206, and the x-axis of the triaxial accelerometer 212 is coaxial with the pressure-resistant outer cylinder 1. The gyroscope 213 is installed at the center of the angular velocity measuring module 204 and is coaxial with the pressure-resistant outer cylinder 1. The angular velocity measuring module 204 and the acceleration measuring module 206 are connected to the main control module 208, which transmits the angular velocity data measured by the gyroscope 213 and the acceleration data measured by the accelerometer 212 to the main control module 208. The main control module 208 is installed in the fourth mounting slot 209 at the top of the mounting frame 2.

[0202] Specifically, the accelerometer 212 is installed at the center of the acceleration measurement module 206. The x-axis of the triaxial accelerometer is coaxial with the pressure-resistant outer cylinder 1, and the plane formed by the y-axis and z-axis is parallel to the acceleration measurement module 206. The acceleration measurement module 206 includes a first AD acquisition module, which acquires the acceleration signal of the triaxial accelerometer and sends the signal to the processor.

[0203] The gyroscope 213 is installed at the center of the angular velocity measurement module 204 and is coaxial with the pressure-resistant outer cylinder 1. The angular velocity measurement module 204 includes a second AD acquisition module, which acquires the gyroscope signal and sends it to the processor.

[0204] The triaxial accelerometer is integrated at the very center of the acceleration measurement module 206 and mounted to the bottom of the mounting frame 2 with screws. The gyroscope 213 is integrated at the very center of the angular velocity measurement module 204 and mounted to the middle of the mounting frame 2 with screws. This layout is designed to ensure that the accelerometer and gyroscope are mounted in the center of the device, guaranteeing the accuracy of sensor measurements and reducing the amplification of centrifugal acceleration and vibration caused by eccentric mounting.

[0205] The mounting frame 2, serving as the main load-bearing structure of the entire device, is tightly fixed to the interior of the pressure-resistant outer cylinder 1 by threads to withstand the high pressure and impact underground. Two mounting slots, the first mounting slot 210 and the second mounting slot 211, are formed perpendicular to the central axis on its circumference. The angular velocity measurement module 204 and the acceleration measurement module 206 are both circular and are respectively installed in the second mounting slot 211 and the first mounting slot 210.

[0206] Furthermore, both the angular velocity measurement module 204 and the acceleration measurement module 206 have a fixing block on one side, which is fixedly connected to the mounting frame 2. Specifically, the angular velocity measurement module 204 is mounted on the second fixing block 205 using screws or other fasteners. The second fixing block 205 has a fixing part, which is fixedly connected to the side wall of the mounting groove using screws or other fasteners. The acceleration measurement module 206 adopts the same installation method, being mounted on the first fixing block 207 using screws or other fasteners. The fixing part is fixedly connected to the side wall of the mounting groove using screws or other fasteners. Since the angular velocity measurement module 204 and the acceleration measurement module 206 are coaxial with the mounting frame 2 after installation, and the mounting frame 2 is coaxial with the pressure-resistant outer cylinder 1, the angular velocity measurement module 204 and the acceleration measurement module 206 are coaxial with the pressure-resistant outer cylinder 1.

[0207] Furthermore, the mounting frame 2 is provided with a connecting part, which is fixedly connected to the pressure-resistant outer cylinder 1. The connecting part can be an external thread 203. The end of the mounting frame is provided with a force-applying end, which serves as a force-applying point when the mounting frame 2 is assembled with the pressure-resistant outer cylinder 1 to insert the mounting frame 2 into the pressure-resistant outer cylinder 1.

[0208] For example, one end of the mounting frame 2 is provided with an external thread 203, and the corresponding position inside the pressure-resistant outer cylinder 1 is provided with an internal thread. The mounting frame 2 is fixed inside the pressure-resistant outer cylinder 1 by the thread.

[0209] A third mounting slot 201 is provided on the end of the mounting frame 2 away from the main control module 208, the angular velocity measurement module 204 and the acceleration measurement module 206. The triaxial magnetometer 202 is independently mounted in the third mounting slot 201 on the mounting frame 2 by screws. This layout is intended to keep the magnetometer away from potential magnetic interference sources on the circuit board and ensure the accuracy of magnetic field measurement.

[0210] The above embodiments are used to explain the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A method for measuring the attitude of a rotary steerable drilling tool, wherein the rotary steerable drilling tool includes an accelerometer, a gyroscope, and a magnetometer, characterized in that... The method includes: State equations are established using the gravity tool face angle, well inclination angle, azimuth angle, and magnetometer magnetic interference as state variables, and observation equations are established using accelerometer measurements and magnetometer measurements as observation variables. The observation noise covariance matrix is ​​adaptively adjusted by multiplying the acceleration component in the initial value of the observation noise covariance matrix with the adjustment coefficient. The Kalman gain is updated using the observation noise covariance matrix, Jacobian matrix, and prediction covariance matrix. The updated gravity tool face angle, well inclination angle, and azimuth angle are obtained through the Kalman gain, prior estimates of state variables, observed values ​​of observed variables, and prior estimates of observed variables. The establishment of the state equation includes: Gravity tool face angle and stable platform speed The relationship is: ； in, For time indexing, Indicates the sampling time. Noise from gravity tool face angle system; well inclination angle Azimuth , Shaft magnetic interference , Shaft magnetic interference and Shaft magnetic interference Modeled as a random walk process: ； in For well inclination angle system noise, For azimuth system noise, for Shaft magnetic interference system noise, for Shaft magnetic interference system noise, for The noise of the shaft magnetic interference system is as follows: ； in, , For the system matrix, For the input matrix, Let be the system process noise vector, and its covariance matrix be... , The system process noise variance represents the angle of the gravity tool face. The variance of system process noise, representing the well inclination angle. The variance of system process noise represents the azimuth angle. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference. express System process noise variance of shaft magnetic interference; Indicates the first Rotational speed at any given moment; Using the measurements from the accelerometer and magnetometer as observed variables, the observation equation is: ； in, The local magnetic inclination, It is the acceleration due to gravity. The local magnetic field strength, The observation noise vector has the following covariance matrix: , Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates accelerometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft, Indicates magnetometer Measurement noise variance of the shaft.

2. The method for measuring the attitude of a rotary steerable drilling tool according to claim 1, characterized in that... The updated observation noise covariance matrix is: ； in This is for adjusting the coefficient.

3. The method for measuring the attitude of a rotary steerable drilling tool according to claim 2, characterized in that... The adjustment coefficient of the observation noise covariance matrix is ​​obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration: The relative deviation is The adjustment coefficient is ;in, The current acceleration magnitude, If the magnitude of the local gravitational acceleration is given, then the observation noise covariance matrix is: 。 4. The method for measuring the attitude of a rotary steerable drilling tool according to claim 2, characterized in that... The adjustment coefficient of the observation noise covariance matrix is ​​obtained based on the relative deviation between the current acceleration magnitude and the gravitational acceleration: The relative deviation is Based on the magnitude of the relative deviation, the interference is divided into several levels, and the corresponding adjustment coefficient is obtained according to the level. .

5. The method for measuring the attitude of a rotary steerable drilling tool according to claim 1, characterized in that... The Kalman gain is updated using the observation noise covariance matrix, Jacobian matrix, and prediction covariance matrix. The Kalman gain is calculated as follows: ； in, For Jacobian matrices, To predict the covariance matrix, To observe the noise covariance matrix; the state variables are updated as follows: ； in For prior estimates of state variables, , For the observed values ​​of the observed variable, For the reason The prior estimates of the observed variables are calculated. For the system matrix, The input matrix is ​​denoted as .

6. The method for measuring the attitude of a rotary steerable drilling tool according to claim 5, characterized in that... The prediction covariance matrix is ​​obtained using the system matrix, the state estimation covariance matrix of the previous time step, and the system process noise covariance matrix. The prediction covariance matrix is: ； in For the system matrix, Let be the covariance matrix of the system process noise; The state estimation covariance matrix at the current time is obtained using the Kalman gain, Jacobian matrix, and prediction covariance matrix: ； in To predict the covariance matrix.

7. A rotary steerable drilling tool attitude measurement device, comprising a pressure-resistant outer cylinder and a main control module, characterized in that... It also includes a triaxial accelerometer, gyroscope, and magnetometer coaxially mounted with the pressure-resistant outer cylinder. The angular velocity meter data, accelerometer data, and magnetometer data are transmitted to the main control module, which implements the measurement method as described in any one of claims 1-6.

8. The rotary steerable drilling tool attitude measuring device according to claim 7, characterized in that... The measuring device also includes an installation frame, an angular velocity measurement module, an acceleration measurement module, and a magnetometer installed inside the pressure-resistant outer cylinder. The angular velocity measurement module and the acceleration measurement module are inserted into the installation frame perpendicular to its central axis and are fixedly connected to the installation frame. The accelerometer is installed at the center of the acceleration measurement module, and the x-axis of the triaxial accelerometer is coaxial with the pressure-resistant outer cylinder. The gyroscope is installed at the center of the angular velocity measurement module and is coaxial with the pressure-resistant outer cylinder. The magnetometer is installed at the end of the installation frame away from the main control module, the angular velocity measurement module, and the acceleration measurement module. The angular velocity measurement module, the acceleration measurement module, and the magnetometer are connected to the main control module and transmit the angular velocity data, acceleration data, and magnetometer data to the main control module.

9. The rotary steerable drilling tool attitude measuring device according to claim 8, characterized in that... The mounting frame is coaxial with the pressure-resistant outer cylinder. Two mounting slots are opened on the circumference of the mounting frame perpendicular to the central axis. The acceleration measurement module and the angular velocity measurement module are installed in the mounting slots by fixing blocks. The fixing blocks protrude outward in the circumferential direction to form a fixing part. The angular velocity measurement module and the acceleration measurement module are fixed to the fixing block by fixing parts. The fixing block is fixedly connected to the side wall of the mounting slot by the fixing part. A third mounting slot is provided at the end of the mounting frame away from the main control module, the angular velocity measurement module and the acceleration measurement module. The magnetometer is installed in the third mounting slot.

Images