Method and device for monitoring azimuth angle signal of drill rod in real time

Abstract

The invention relates to the technical field of signal real-time monitoring, and discloses a drill rod azimuth angle signal real-time monitoring method and device. The method comprises the following steps: acquiring drill rod relative azimuth data and drill rod rotation motion characteristic data; angular velocity integration and rotation state analysis are conducted through the drill rod rotation angle prediction model, and the real-time rotation deviation angle of the drill rod relative to the geographic coordinate system is obtained; based on the real-time rotation deviation angle, calculating a rotation transformation relation between an orientation device coordinate system and a geographic coordinate system to obtain an azimuth angle conversion compensation parameter; performing rotation compensation and coordinate system conversion on the relative azimuth angle data of the drill rod to obtain geographic absolute azimuth angle data; and performing sliding mode control on the geographic absolute azimuth data, and outputting a geographic absolute azimuth signal. According to the method, accurate real-time conversion from the relative azimuth angle measured by the orientation device to the absolute azimuth angle required by drilling guidance is realized by establishing the accurate mapping relation between the rotation state of the drill rod and the geographic coordinate system.

Description

Technical Field

[0001] This invention relates to the field of real-time signal monitoring technology, and in particular to a method and device for real-time monitoring of drill pipe azimuth signals. Background Technology

[0002] In modern oil drilling and geological exploration, directional instruments are crucial equipment for well guidance. They detect the azimuth information of the drill pipe using magnetic field sensors. However, existing drill pipe azimuth monitoring technologies generally suffer from significant technical deficiencies, primarily in their inability to effectively handle the impact of continuous drill pipe rotation on azimuth measurements. Traditional directional instrument measurement systems obtain the relative azimuth angle relative to the drill pipe's coordinate system, while well guidance actually requires the absolute azimuth angle relative to the geographic coordinate system. The continuous rotation of the drill pipe during drilling causes real-time rotational deviations between the two coordinate systems. Existing static angle compensation algorithms cannot adapt to this dynamic change, lacking an effective monitoring and compensation mechanism for the real-time rotation of the drill pipe. This results in accumulated azimuth measurement errors, severely affecting the accuracy of well guidance. Summary of the Invention

[0003] This invention provides a method and device for real-time monitoring of drill pipe azimuth signals. By establishing a precise mapping relationship between the drill pipe rotation state and the geographic coordinate system, this invention achieves accurate real-time conversion of the relative azimuth angle measured by the orientation instrument to the absolute azimuth angle required for drilling guidance.

[0004] In a first aspect, the present invention provides a method for real-time monitoring of drill pipe azimuth angle signals, the method comprising: Collect relative azimuth angle data of the drill pipe and rotational motion characteristic data of the drill pipe; The drill pipe rotation motion characteristic data are input into the drill pipe rotation angle prediction model for angular velocity integration and rotation state analysis to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. Based on the real-time rotation deviation angle, the rotation transformation relationship between the orientation instrument coordinate system and the geographic coordinate system is calculated to obtain the azimuth angle transformation compensation parameters; Based on the azimuth transformation compensation parameters, the relative azimuth data of the drill pipe is rotated and the coordinate system is transformed to obtain the geographic absolute azimuth data. Sliding mode control is applied to the geographic absolute azimuth data to output the geographic absolute azimuth signal.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the acquisition of drill pipe relative azimuth angle data and drill pipe rotational motion characteristic data includes: The three-axis magnetic field strength components detected by the orientation instrument's magnetic field sensor and the six-axis inertial signals detected by the drill pipe's accelerometer gyroscope are simultaneously acquired to obtain the original magnetic field orientation signal and the original inertial motion signal of the drill pipe. The original magnetic field orientation signal and the original inertial motion signal of the drill rod are respectively input into a Kalman filter for noise cancellation and signal smoothing to obtain the denoised magnetic field orientation signal and the denoised inertial motion signal. Based on the sensor installation geometry, coordinate system alignment and magnetic declination correction are performed on the denoised magnetic field orientation signal to obtain a standardized magnetic field orientation signal in the drill pipe body coordinate system. The azimuth angle of the standardized magnetic field azimuth signal is calculated according to the magnetic field azimuth calculation algorithm to obtain the relative azimuth angle data of the drill pipe, and the coordinate system transformation is performed on the denoised inertial motion signal to obtain the rotational motion characteristic data of the drill pipe.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of inputting the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model for angular velocity integration and rotation state analysis to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system includes: The drill pipe rotation motion characteristic data are input into the data preprocessing layer of the drill pipe rotation angle prediction model for triaxial angular velocity separation and amplitude normalization to obtain a standardized triaxial angular velocity sequence. The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain the drill pipe rotation motion trend prediction parameters. The drill pipe rotation motion trend prediction parameters are input into the angular velocity integral calculation layer of the drill pipe rotation angle prediction model for numerical integration and cumulative rotation angle calculation to obtain the cumulative rotation angle value of the drill pipe's three axes. Nonlinear feature extraction and rotation pattern recognition are performed on the cumulative rotation angle values ​​of the three axes of the drill pipe to obtain the instantaneous rotation state parameters of the drill pipe. Then, coordinate system transformation and geographic reference correction are performed on the instantaneous rotation state parameters of the drill pipe to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the step of inputting the standardized triaxial angular velocity sequence into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain drill pipe rotation motion trend prediction parameters includes: The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model. In the time series analysis layer, the standardized triaxial angular velocity sequence is segmented into time windows and historical data is sampled to obtain the historical angular velocity data sequence. The historical angular velocity data sequence is input into the autoregressive moving average model parameter recognizer for model order optimization and optimal parameter search to obtain the autoregressive order p value and the moving average order q value. Based on the autoregressive order p and the moving average order q, the historical angular velocity data sequence is fitted with least squares and the coefficients are solved to obtain the autoregressive coefficient vector and the moving average coefficient vector. Based on the autoregressive coefficient vector and the moving average coefficient vector, angular velocity prediction and trend extrapolation are performed to obtain the drill pipe rotation motion trend prediction parameters.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the calculation of the rotational transformation relationship between the orientation coordinate system and the geographic coordinate system based on the real-time rotational deviation angle to obtain the azimuth angle transformation compensation parameters includes: The real-time rotational deviation angle is separated into three rotational components and mapped to coordinate axes to obtain independent rotational angle components of the drill pipe around the X-axis, Y-axis and Z-axis; Based on the independent rotation angle components, a three-dimensional rotation transformation matrix is ​​constructed and matrix elements are calculated to obtain the real-time rotation transformation matrix from the orientation coordinate system to the geographic coordinate system. Based on the real-time rotation transformation matrix, geographic north reference alignment and magnetic declination compensation are performed to obtain the corrected rotation transformation matrix. Based on the corrected rotation transformation matrix, the azimuth offset is extracted and the compensation coefficient is calculated to obtain the azimuth transformation compensation parameters.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of performing rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth transformation compensation parameters to obtain the geographic absolute azimuth data includes: The azimuth angle conversion compensation parameters are input into the adaptive compensation controller for compensation coefficient analysis and dynamic adjustment factor calculation to obtain the real-time rotation compensation coefficient and fuzzy PID control parameters. Based on the real-time rotation compensation coefficient, the relative azimuth angle data of the drill pipe is corrected for angular offset and rotation error is eliminated to obtain the azimuth angle data after rotation compensation. Based on the azimuth data after rotation compensation and the fuzzy PID control parameters, matrix multiplication and coordinate system mapping are performed to obtain the geographic absolute azimuth data.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the step of inputting the azimuth angle conversion compensation parameters into an adaptive compensation controller for compensation coefficient analysis and dynamic adjustment factor calculation to obtain real-time rotation compensation coefficients and fuzzy PID control parameters includes: Numerical analysis and error feature extraction are performed on the azimuth conversion compensation parameters to obtain the azimuth compensation error amount and error change rate; The azimuth angle compensation error and the error change rate are input into the adaptive compensation controller for membership function calculation and fuzzy rule matching to obtain the fuzzy control output. Based on the fuzzy control output, centroid method defuzzification and PID parameter adjustment are performed to obtain proportional coefficient adjustment, integral coefficient adjustment and derivative coefficient adjustment. The reference PID parameters are dynamically corrected and the compensation coefficients are updated based on the proportional coefficient adjustment, the integral coefficient adjustment, and the derivative coefficient adjustment to obtain the real-time rotational compensation coefficient and fuzzy PID control parameters.

[0011] Optionally, in a seventh implementation of the first aspect of the present invention, the step of performing sliding mode control on the geographic absolute azimuth data and outputting the geographic absolute azimuth signal includes: The expected trajectory tracking error is calculated and the error is differentiated by performing the expected trajectory tracking error calculation and error differentiation on the geographic absolute azimuth data to obtain the azimuth tracking error signal and the error change rate signal; Based on the azimuth tracking error signal and the error change rate signal, the sliding surface function is constructed and the switching control law is designed to obtain the sliding surface state quantity and the sliding switching control quantity. The sliding mode surface state quantity is input into an adaptive gain regulator to detect drilling vibration interference amplitude and dynamically adjust the control gain to obtain adaptive switching gain parameters. At the same time, the saturated function boundary layer technology is used to perform chattering elimination processing on the sliding mode switching control quantity to obtain a smooth sliding mode control signal. The drilling vibration interference component in the geographic absolute azimuth data is compensated for based on the adaptive switching gain parameter and the smooth sliding mode control signal, and the geographic absolute azimuth signal is output.

[0012] Optionally, in the eighth implementation of the first aspect of the present invention, the step of inputting the sliding mode surface state quantity into an adaptive gain regulator for drilling vibration interference amplitude detection and dynamic control gain adjustment to obtain adaptive switching gain parameters, and simultaneously using saturated function boundary layer technology to perform chattering elimination processing on the sliding mode switching control quantity to obtain a smooth sliding mode control signal, includes: The sliding surface state quantity is input into an adaptive gain regulator to extract the amplitude and separate the frequency components of the drilling vibration signal, thereby obtaining the vibration interference amplitude signal and vibration frequency characteristic parameters. Based on the vibration interference amplitude signal, an adaptive gain law calculation and gain boundary constraint are performed to obtain a dynamically switching gain adjustment factor. The sliding mode switching control variable is subjected to saturation function boundary layer thickness setting and sign function smoothing replacement processing to obtain boundary layer thickness parameters and saturation function control law; The gain parameters are fused according to the dynamic switching gain adjustment factor and the vibration frequency characteristic parameters to obtain the adaptive switching gain parameters. The control signal is then smoothed according to the saturation function control law and the boundary layer thickness parameters to obtain the smooth sliding mode control signal.

[0013] Secondly, the present invention provides a real-time monitoring device for drill pipe azimuth angle signals, the real-time monitoring device for drill pipe azimuth angle signals comprising: The data acquisition module is used to acquire relative azimuth angle data of the drill pipe and rotational motion characteristic data of the drill pipe; The analysis module is used to input the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model to perform angular velocity integration and rotation state analysis, so as to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. The calculation module is used to calculate the rotation transformation relationship between the orientation coordinate system and the geographic coordinate system based on the real-time rotation deviation angle, and obtain the azimuth angle transformation compensation parameters. The conversion module is used to perform rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth conversion compensation parameters to obtain the geographic absolute azimuth data. The output module is used to perform sliding mode control on the geographic absolute azimuth data and output the geographic absolute azimuth signal.

[0014] The technical solution provided by this invention establishes a multi-sensor synchronous acquisition system by integrating the orientation instrument's magnetic field sensor and the drill pipe's accelerometer and gyroscope. This system accurately captures the six-degree-of-freedom motion state of the drill pipe during drilling. A drill pipe rotation angle prediction model combined with a rotation angle recognition neural network is used to accurately predict the rotation deviation angle of the drill pipe relative to the geographic coordinate system. A dynamic rotation transformation matrix is ​​constructed based on the real-time rotation deviation angle to achieve accurate conversion from the orientation instrument coordinate system to the geographic coordinate system, effectively eliminating the influence of drill pipe rotation on azimuth measurement. The combination of fuzzy PID control and an adaptive compensation controller enables intelligent adjustment of compensation parameters, automatically optimizing the compensation strategy according to real-time changes in drilling conditions, significantly improving the accuracy and stability of azimuth compensation. A sliding mode control algorithm combined with adaptive gain adjustment and saturated function boundary layer technology effectively suppresses the interference of drilling vibration on the azimuth signal, ensuring high-precision azimuth monitoring even in complex drilling environments. By establishing a precise mapping relationship between the drill pipe rotation state and the geographic coordinate system, accurate real-time conversion of the relative azimuth angle measured by the orientation instrument to the absolute azimuth angle required for drilling guidance is achieved. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of an embodiment of the real-time monitoring method for drill pipe azimuth signal in this invention; Figure 2 This is a schematic diagram of one embodiment of the drill pipe azimuth signal real-time monitoring device in this invention. Detailed Implementation

[0017] This invention provides a method and apparatus for real-time monitoring of drill pipe azimuth signals. The terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the real-time monitoring method for drill pipe azimuth signal in this invention includes: Step S101: Collect relative azimuth angle data and rotational motion characteristic data of the drill pipe; It is understood that the executing entity of this invention can be a real-time monitoring device for drill pipe azimuth signals, or it can be a terminal or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example.

[0019] Specifically, a directional magnetic field sensor module and an inertial measurement module are integrated and installed inside or near the drill pipe. The directional magnetic field sensor detects the geomagnetic field intensity components in the X, Y, and Z directions, respectively. Simultaneously, a triaxial accelerometer and a triaxial gyroscope acquire the linear acceleration and angular velocity data of the drill pipe in the three axes, respectively. These data are synchronously acquired using a consistent time sampling benchmark, forming a set of original magnetic field orientation signals and inertial motion signals of the drill pipe corresponding to a timestamp. These two types of signals are then input into a structurally optimized Kalman filter model. Based on its recursive state estimation characteristics, random noise in the signals is dynamically modeled and suppressed, outputting denoised magnetic field orientation signals and denoised inertial motion signals. The magnetic field azimuth signal is transformed from the orientation instrument's own coordinate system to the drill pipe's coordinate system. Considering the fixed deviations or attitude angles between the magnetic field sensor and the drill pipe body in terms of structural installation, a coordinate system mapping matrix is ​​constructed based on the sensor's installation geometry. The denoised magnetic field signal is then subjected to spatial coordinate transformation to achieve coordinate system alignment. Simultaneously, to compensate for magnetic declination errors caused by local geomagnetic anomalies, magnetic interference, or geographical location changes, orientation correction is applied to the signal using a geomagnetic model or measured magnetic declination data, resulting in a standardized, drift-free magnetic field azimuth signal in the drill pipe's coordinate system. Based on the magnetic field azimuth calculation algorithm, attitude estimation is performed using triaxial magnetic induction intensity data and the calibrated coordinate system, combined with acceleration information, to accurately determine the drill pipe's rotation angle around the vertical axis, i.e., the real-time relative azimuth angle data. To fully extract the dynamic rotation behavior characteristics of the drill pipe, the denoised inertial motion signal is transformed from the sensor coordinate system to the drill pipe body coordinate system. Based on the initially established geometric installation relationship, the six-axis signal is synchronously transformed using a coordinate rotation matrix or a quaternion rotation model. The rotation trend around each axis is extracted from the angular velocity component, thereby forming the drill pipe rotation motion characteristic data corresponding to the magnetic field orientation data.

[0020] Step S102: Input the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model to perform angular velocity integration and rotation state analysis, and obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. Specifically, the drill pipe rotation motion characteristic data is structured. The raw inertial measurement data is input into the data preprocessing layer of the drill pipe rotation angle prediction model. The X, Y, and Z axis angular velocity components are separated sequentially, and the amplitude is normalized using the physical distribution characteristics of angular velocity under different drilling conditions, ensuring that the angular velocity data of each channel are within a uniform numerical range, resulting in a standardized three-axis angular velocity sequence. The normalized three-axis angular velocity data is then used as time series samples and input into the time series analysis layer of the prediction model. Within this layer, a dynamic modeling algorithm based on autoregressive moving average theory is used to establish an ARMA(p,q) time series model for the angular velocity sequence of each axis. A sliding window mechanism is used to continuously track the changing trend of drill pipe rotation behavior, thereby extracting prediction parameters characterizing rotational inertia and disturbance feedback patterns, including autoregressive coefficients, moving average coefficients, and time delay terms, forming a dynamic parameter set for drill pipe rotation trend prediction. The predicted parameters of the drill pipe's rotational motion are input into the angular velocity integral calculation layer. Combined with real-time angular velocity values, a numerical integration algorithm is used to perform discrete-time domain integration, thereby gradually accumulating the rotation angle values ​​of each axis. The output three-axis cumulative rotation angle values ​​reflect the overall spatial rotational behavior of the drill pipe since its initial reference state. To improve the physical interpretation and engineering adaptability of the rotational state parameters, a nonlinear feature extraction algorithm is applied to the cumulative rotation angle values. This algorithm identifies the drill pipe's current instantaneous rotational mode from multiple aspects, including the rate of change of angle, fluctuation pattern, and periodic behavior. These modes include continuous uniform rotation, low-frequency oscillation, and rapid deflection. The rotational mode parameters are then used as dynamic control variables input into the coordinate transformation module. In this module, the model uses a rotation matrix or quaternion algorithm to spatially map and calculate the instantaneous rotational state parameters based on the preset transformation rules between the drill pipe's body coordinate system and the geographic coordinate system. Finally, geographic alignment is performed using geographic reference data to obtain the real-time rotational deviation angle of the drill pipe relative to the geographic coordinate system.

[0021] Step S103: Calculate the rotation transformation relationship between the orientation instrument coordinate system and the geographic coordinate system based on the real-time rotation deviation angle to obtain the azimuth angle transformation compensation parameters; Specifically, the real-time rotational deviation angle undergoes three-axis decoupling processing, separating the angle from its overall attitude change state into independent rotational components around the X, Y, and Z axes. This process is performed using a three-dimensional attitude analysis algorithm, including Euler angle calculation or quaternion transformation, to resolve the overall rotational information into angle values ​​in three independent directions, characterizing the spatial deflection behavior of the drill pipe in the pitch, roll, and yaw directions, respectively. Using the independent rotational angles of the drill pipe around the three axes as input, a three-dimensional rotational transformation matrix is ​​constructed according to the Euler angle rotation order. This matrix is ​​formed by multiplying and superimposing the basic rotational matrices that rotate around the X, Y, and Z axes, respectively, to form a three-dimensional direction cosine matrix, where each element is a trigonometric function combination of the input angle, used to describe the spatial rotation state of the orientation instrument's body coordinate system relative to the geographic coordinate system under the influence of the drill pipe's real-time attitude change. Since the drill pipe is prone to directional ambiguity and attitude drift under high-speed rotation and nonlinear disturbances, this rotational transformation matrix needs to be dynamically updated at a high frequency, and matrix orthogonality checks and condition number constraints are used to ensure numerical stability and reversibility. To improve the accuracy of coordinate transformation within the geographic reference frame, a geographic north reference alignment mechanism is introduced based on the initial rotation matrix. By referencing a geomagnetic model or measured magnetic declination data, the Z-axis direction (heading axis) in the rotation matrix is ​​corrected to ensure it always points to true north in the geographic coordinate system. Simultaneously, considering the variations in magnetic declination across different regions, a magnetic declination compensation module is introduced in the later stages of matrix construction to incrementally correct the rotation component of the Z-axis direction. This eliminates interference from magnetic field distortion and regional errors in azimuth measurements, resulting in a corrected rotation transformation matrix. Based on this matrix, azimuth offset extraction is performed. The rotation matrix is ​​applied to the relative azimuth vector output by the orientation instrument. Matrix multiplication is used to calculate the new direction of this vector in the geographic coordinate system. Combined with the initial attitude state, vector difference calculation is performed to extract the azimuth offset caused by drill pipe movement. This offset serves as the core input for azimuth compensation parameters. It is also processed using multivariate weighting based on current drill pipe rotation rate, vibration state, and other characteristics, ultimately outputting a set of azimuth transformation compensation parameters adapted to the current attitude state.

[0022] Step S104: Perform rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth transformation compensation parameters to obtain the geographic absolute azimuth data; Specifically, the azimuth angle conversion compensation parameters are input into the adaptive compensation controller. This controller integrates a fuzzy logic inference unit and a gain adjustment mechanism. It analyzes the compensation coefficients under the current attitude error state and performs fuzzy rule inference operations based on historical compensation behavior and current disturbance trends to dynamically calculate the rotation compensation coefficients matching the current drill pipe state and the adjustment parameter set of the fuzzy PID controller, which includes proportional, integral, and derivative coefficients. Based on the real-time rotation compensation coefficients output by the adaptive compensation controller, the relative azimuth angle data of the drill pipe is corrected for angle offset and the rotation error is eliminated. Angle offset correction deducts the azimuth angle offset caused by the rotation deviation of the drill pipe from the relative angle data according to its direction and magnitude, obtaining an intermediate angle value that approximates the true azimuth. The rotation error elimination process suppresses the influence of residual disturbances and instantaneous angle fluctuations on the compensation results through a smoothing adjustment mechanism and fine-tunes the compensation magnitude using dynamic weighting, thereby achieving more refined and responsive rotation compensation correction and obtaining the azimuth angle data after rotation compensation. The azimuth data after rotation compensation is used as the input vector. Combined with the direction cosine rotation matrix constructed by the fuzzy PID control parameters, matrix multiplication is performed to map the relative angle to the geographic coordinate system. At the same time, the transformation result is aligned with the preset geographic reference direction to obtain the geographic absolute azimuth data.

[0023] Step S105: Perform sliding mode control on the geographic absolute azimuth data and output the geographic absolute azimuth signal.

[0024] Specifically, the expected trajectory tracking error is calculated using geographic absolute azimuth data. This process uses the ideal azimuth trajectory or target guidance value set during drilling operations as a reference input, and performs time-by-time difference calculations with the current geographic absolute azimuth data to obtain an azimuth tracking error signal reflecting the current degree of guidance deviation of the system. The error signal is then processed using its time derivative to capture the rate of error change over time, thereby obtaining an error change rate signal reflecting the dynamic response of the error. Based on the azimuth tracking error signal and the error change rate signal, a sliding surface function is constructed and a switching control law is designed. By combining the error signal and its change rate signal according to a specified proportional coefficient, a sliding surface function is constructed, which is a weighted sum of the error change rate and the error itself. This forces the system state to be projected near the sliding surface once it deviates from the target trajectory. Based on this sliding surface state, a corresponding switching control law is designed, using a typical signed function structure to generate the sliding switching control quantity, ensuring that the system can quickly enter and remain near the sliding surface for error suppression. However, the chattering problem introduced by traditional sign functions is particularly severe in the high-frequency vibration environment of drilling. Therefore, a saturated function boundary layer technique is introduced in the control law implementation process, replacing the sign function with a continuously differentiable function to avoid system oscillations and actuator fatigue caused by high-frequency control switching, and to obtain a smooth and stable sliding mode control signal. To enhance the adaptive capability of sliding mode control to drilling disturbances, the sliding surface state variable is input to an adaptive gain regulator. This regulator detects the vibration disturbance amplitude under the working conditions of the drill pipe in real time and dynamically adjusts the switching gain parameter in the controller based on the current disturbance characteristics. When an increase in drilling disturbance is detected, the regulator increases the switching gain of the controller to enhance the system's ability to suppress errors; conversely, it appropriately reduces the gain to reduce system output fluctuations and avoid overcompensation. At the same time, this mechanism can effectively prevent controller instability or output oscillations caused by drastic changes in environmental disturbances. The adaptive switching gain parameter and the sliding mode control signal with chatter suppression are applied together to the geographic absolute azimuth data to perform real-time interference compensation and dynamic correction on the high-frequency abnormal components caused by drilling machinery disturbances, electromagnetic interference, or inertial sensor fluctuations, thereby outputting the geographic absolute azimuth signal.

[0025] In one specific embodiment, the process of performing step S101 may specifically include the following steps: The three-axis magnetic field strength components detected by the orientation instrument's magnetic field sensor and the six-axis inertial signals detected by the drill pipe's accelerometer gyroscope are simultaneously acquired to obtain the original magnetic field orientation signal and the original inertial motion signal of the drill pipe. The original magnetic field orientation signal and the original inertial motion signal of the drill pipe are respectively input into a Kalman filter for noise cancellation and signal smoothing to obtain the denoised magnetic field orientation signal and the denoised inertial motion signal. Based on the sensor installation geometry, coordinate system alignment and magnetic declination correction are performed on the denoised magnetic field orientation signal to obtain the standardized magnetic field orientation signal in the drill pipe body coordinate system. The azimuth angle of the standardized magnetic field azimuth signal is calculated based on the magnetic field azimuth solution algorithm to obtain the relative azimuth angle data of the drill pipe. The coordinate system is then transformed on the denoised inertial motion signal to obtain the rotational motion characteristic data of the drill pipe.

[0026] Specifically, a directional magnetic field sensor and an inertial measurement unit (IMU) are arranged inside or near the drill pipe body to collect geomagnetic field information and drill pipe dynamic attitude information, respectively. The directional magnetic field sensor measures the vector components of the geomagnetic field in the X, Y, and Z axes in real time, outputting three-axis magnetic field strength data and recording the magnetic field vector projection of the drill pipe's location. Simultaneously, the IMU integrated into the drill pipe body acquires the linear acceleration and angular velocity signals of the drill pipe through internal three-axis accelerometers and three-axis gyroscopes, forming a six-axis inertial signal that characterizes the drill pipe's motion state and spatial rotation characteristics during drilling. To ensure the temporal consistency and time-domain alignment of all sensor data, the system employs a unified timestamp mechanism and sets up sampling synchronization logic, collecting and encoding magnetic field data and inertial data according to the same sampling period to avoid information misalignment or processing deviations introduced by multi-source asynchronous data. The resulting three-axis magnetic field data constitutes the original magnetic field orientation signal of the drill pipe, while the combined data of three-axis acceleration and three-axis angular velocity constitute the original inertial motion signal of the drill pipe. Noise suppression and signal smoothing are then performed on the above original signals. An extended Kalman filter is used to independently filter the raw magnetic field orientation signal and the raw inertial motion signal. The Kalman filter algorithm iteratively corrects the sensor output using the covariance relationship between the current observation value and the system state prediction value, and constructs a state estimation model through the state transition matrix and the observation matrix, effectively filtering out high-frequency noise and non-stationary abrupt components. After filtering, the magnetic field signal yields triaxial geomagnetic intensity data, removing magnetic interference from the drill pipe itself and occasional external magnetic field disturbances. After filtering, the inertial signal yields smooth acceleration and angular velocity curves, effectively suppressing angle drift and velocity anomalies caused by oscillations and abrupt changes. To accurately map the sensor measurement results to the physical coordinate system of the drill pipe body, coordinate system alignment processing is performed on the magnetic field orientation signal. Since the magnetic field sensor is installed inside the orientation instrument, there is a fixed but non-collinear geometric installation deviation between its body coordinate system and the drill pipe structural coordinate system. During the system initialization phase, the installation rotation matrix or Euler angle parameters are obtained through mechanical calibration or attitude testing, and the magnetic field data is transformed from the sensor coordinate system to the drill pipe body coordinate system during data processing. To improve the geographic reference accuracy of magnetic field direction determination, the system performs magnetic declination correction processing. The Earth's magnetic field does not point strictly to true north in geographic space; different regions exhibit fixed or gradually varying magnetic declination angles. Therefore, by combining a geomagnetic model, geographical location, and real-time magnetic declination reference data, the magnetic field vector is rotated and adjusted to achieve geographic north magnetic alignment. The magnetic field signal after coordinate system transformation and magnetic declination correction is the standardized magnetic field azimuth signal in the drill pipe body coordinate system. Its three-axis data reflects the accurate spatial direction of the Earth's magnetic field in the drill pipe body reference system.The system analyzes the standardized magnetic field signal using a magnetic field orientation calculation algorithm. Based on the classic three-axis geomagnetic vector calculation model, it calculates the heading angle (azimuth angle) using the XY components in the horizontal plane. The system then extracts the direction projection angle of the magnetic vector in the horizontal plane using inverse trigonometric functions such as the atan2 function. Combined with previous magnetic declination correction results, the output value is geographically corrected to obtain the relative azimuth angle data corresponding to the drill pipe's current attitude. Simultaneously, the denoised inertial motion signal undergoes coordinate system transformation. Based on the physical installation attitude of the accelerometer and gyroscope, their output acceleration and angular velocity are transformed from the sensor reference frame to the drill pipe's body coordinate system. A rotation matrix or quaternion interpolation method is used to ensure the continuity and accuracy of the transformation. The transformed inertial signal possesses a geometric meaning consistent with the drill pipe's structural orientation, accurately representing the linear motion trend and rotational behavior of the drill pipe around each axis. This forms structured drill pipe rotational motion characteristic data, which, together with the aforementioned relative azimuth angle data, constitutes the spatial attitude input set.

[0027] In one specific embodiment, the process of performing step S102 may specifically include the following steps: The drill pipe rotation motion characteristic data are input into the data preprocessing layer of the drill pipe rotation angle prediction model for triaxial angular velocity separation and amplitude normalization to obtain a standardized triaxial angular velocity sequence. The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain the drill pipe rotation motion trend prediction parameters. The drill pipe rotation motion trend prediction parameters are input into the angular velocity integral calculation layer of the drill pipe rotation angle prediction model for numerical integration and cumulative rotation angle calculation to obtain the cumulative rotation angle value of the drill pipe's three axes. Nonlinear feature extraction and rotation pattern recognition are performed on the cumulative rotation angle values ​​of the drill pipe's three axes to obtain the instantaneous rotation state parameters of the drill pipe. Coordinate system transformation and geographic reference correction are then performed on the instantaneous rotation state parameters of the drill pipe to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system.

[0028] Specifically, the drill pipe rotation motion feature data is input into the data preprocessing layer of the drill pipe rotation angle prediction model. The original drill pipe rotation motion feature data includes triaxial angular velocity signals processed by a Kalman filter, i.e., instantaneous angular velocity measurements around the X, Y, and Z axes. The preprocessing layer performs component separation processing on this triaxial angular velocity signal, decomposing the angular velocity data matrix into three independent one-dimensional sequences, each corresponding to the rotational dynamic trajectory along one axis. To address the differences in amplitude, inconsistencies in dimensions, or unbalanced data distribution across different axes during actual rotation, the system employs amplitude normalization strategies such as maximum-minimum normalization or mean-standard-deviation normalization to compress the triaxial angular velocity data into a uniform scale interval, such as [-1, 1] or a normal interval with a mean of 0 and a standard deviation of 1. This ensures the consistency of the contribution weights of each channel to model training and avoids numerical explosion or convergence oscillations during numerical computation, outputting a standardized triaxial angular velocity sequence. The standardized angular velocity data is input into the time series analysis layer of the drill pipe rotation angle prediction model. Within this layer, the system performs short-term trend modeling and noise dynamic suppression analysis on the input sequence based on autoregressive moving average (ARMA) modeling theory. The system constructs an ARMA(p,q) time series model using the standardized angular velocity sequence for each axis as the target variable, where p is the autoregressive order and q is the moving average order. A sample set is constructed using data samples from a historical sliding window to fit the predicted value of the current angular velocity. During the fitting process, the system identifies the ARMA model parameters based on the least squares method or maximum likelihood estimation method. Simultaneously, it introduces the AIC (Akaike Information Criterion) and BIC (Bayesian Information Criterion) to optimize the model order, ensuring that the prediction model has sufficient expressive power while avoiding overfitting risks. The output of this time series analysis layer is the drill pipe rotation motion trend prediction parameter, which describes the autocorrelation structure, periodic trend, and fluctuation intensity of the angular velocity of each axis in the time domain. The predicted parameters for the drill pipe's rotational motion are input into the angular velocity integral calculation layer of the drill pipe rotation angle prediction model. In this layer, the system superimposes the actual angular velocity of each axis at the current moment with the predicted parameters to form a dynamically corrected angular velocity input, and performs high-precision numerical integration on this value. The integration employs highly stable discrete integration algorithms such as the trapezoidal rule or symplectic integral method, integrating and accumulating the continuous angular velocities at a set time sampling step. The rotation angle increment around each axis is calculated sequentially at each moment, and the integrals are superimposed over time to finally obtain the cumulative rotation angle values ​​of the drill pipe around the X, Y, and Z axes. Nonlinear feature extraction and pattern recognition processing are then performed on the cumulative rotation angle values ​​of the drill pipe's three axes. This process is used to extract discriminative motion pattern indicators from the original angle trajectory and determine the current rotational state of the drill pipe.During feature extraction, the system calculates multiple features such as the rate of change of angle, the frequency of angle fluctuations, and the energy spectral density. It then employs methods such as principal component analysis, t-SNE dimensionality reduction, or convolutional encoders to compress features and fit the spatial distribution of high-dimensional angle trajectories. Pattern recognition, based on typical rotational state labels in the training samples (e.g., continuous rotation, intermittent oscillation, hysteresis), performs classifier operations such as K-means, SVM, or neural networks in the extracted feature space, outputting the drill pipe's current instantaneous rotational state parameters, including key information such as the current state category, typical rotation period, and fluctuation intensity level. Coordinate system transformation and georeferenced correction are then applied to the drill pipe's instantaneous rotational state parameters. This process is based on the established mapping relationship between the drill pipe body coordinate system and the geographic coordinate system. The rotation angles of the X, Y, and Z axes are projected onto the standard geographic coordinate system based on the geographic north, east and gravity vertical lines according to the rotation matrix multiplication or quaternion rotation formula. The final direction correction is performed based on the measured magnetic declination data to eliminate the attitude offset introduced by geomagnetic differences or local interference, and output the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system with actual geographic meaning.

[0029] In one specific embodiment, the process of inputting the standardized triaxial angular velocity sequence into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain the drill pipe rotation motion trend prediction parameters can specifically include the following steps: The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model. In the time series analysis layer, the standardized triaxial angular velocity sequence is divided into time windows and historical data is sampled to obtain the historical angular velocity data sequence. The historical angular velocity data sequence is input into the autoregressive moving average model parameter recognizer to optimize the model order and search for the optimal parameters, thereby obtaining the autoregressive order p value and the moving average order q value. Least squares fitting and coefficient solving are performed on the historical angular velocity data sequence based on the autoregressive order p-value and the moving average order q-value to obtain the autoregressive coefficient vector and the moving average coefficient vector. Based on the autoregressive coefficient vector and the moving average coefficient vector, angular velocity prediction and trend extrapolation are performed to obtain the drill pipe rotation motion trend prediction parameters.

[0030] Specifically, the standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model to perform dynamic time modeling and trend prediction calculations. Based on the set analysis period and response accuracy requirements, the system performs time window segmentation on the standardized angular velocity sequence. This operation uses a sliding window mechanism to divide continuous angular velocity data into multiple sub-intervals with temporal characteristics. Each time window contains a fixed length of historical sample points to capture the inertial evolution characteristics of drill pipe rotation behavior in the short term. Simultaneously, to improve the model's time adaptability, the system performs multi-step historical data sampling within each window, extracting representative points from the sequence at fixed sampling intervals to construct a representative historical angular velocity data sequence. The historical angular velocity data sequence is then input into the autoregressive moving average model parameter recognizer in the time series analysis layer. This recognizer establishes an optimal fitting model for the historical angular velocity sequence and determines the key structural parameters of the model, namely the autoregressive order p-value and the moving average order q-value. During the model structure search process, the system executes a step-by-step testing strategy, constructing multiple candidate sets of ARMA models with different p and q combinations. For each combination, historical data sequences are fitted, and corresponding model evaluation metrics are calculated, such as the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and the Predictive Residual Sum of Squares (RSS). To ensure the model possesses accurate predictive capabilities without overfitting under complex rotational modes, the system employs a multi-criteria integrated decision strategy to rank and score different order combinations. By minimizing the comprehensive evaluation metrics, the optimal autoregressive order p and the optimal moving average order q are ultimately determined, thus completing the adaptive order optimization process of the entire model structure. After determining the optimal order parameters, the system fixes the p and q values ​​and performs numerical modeling on the historical angular velocity data sequences based on this structure, entering the least squares fitting and coefficient solving stage. In this stage, the system uses the least squares method to construct an error minimization function, setting the objective function as the sum of squared errors between the actual and predicted angular velocity values. The derivative of this error function yields a set of linear equations concerning the model parameters. The system numerically solves the equations using normal equations or QR decomposition to calculate the autoregressive coefficient vector and moving average coefficient vector with optimal fitting accuracy. The autoregressive coefficient vector represents how the current angular velocity is affected by a linear combination of its own values ​​over several historical time points, while the moving average coefficient vector describes how the current angular velocity is affected by the weighted smoothing effect of historical white noise disturbances. The system applies the autoregressive and moving average coefficient vectors to the data points at the end of the current time window to perform multi-step trend extrapolation and prediction calculations for the drill pipe angular velocity. During prediction, the system uses the known historical angular velocity samples and disturbance residuals, employing the structural formula of the ARMA model for iterative calculations to progressively deduce the expected angular velocity values ​​for several future time steps.To improve the robustness of extrapolation, the system simultaneously introduces confidence interval calculation and dynamic error correction mechanisms. During the prediction process, the confidence level of the estimated values ​​is dynamically adjusted to ensure stable trend direction estimates can be obtained even under high-frequency disturbances or sudden changes in state. The system constructs a structured output by combining the predicted angular velocity values ​​of each axis with their corresponding confidence levels, forming a set of parameters for predicting the rotational motion trend of the drill pipe under its current attitude.

[0031] In one specific embodiment, the process of executing step S103 may specifically include the following steps: The real-time rotational deviation angle is separated into three rotational components and mapped to the coordinate axes to obtain the independent rotational angle components of the drill pipe around the X-axis, Y-axis and Z-axis; Based on independent rotation angle components, a three-dimensional rotation transformation matrix is ​​constructed and matrix elements are calculated to obtain the real-time rotation transformation matrix from the orientation coordinate system to the geographic coordinate system. Geographic north reference alignment and magnetic declination compensation are performed based on the real-time rotation transformation matrix to obtain the corrected rotation transformation matrix. Based on the corrected rotation transformation matrix, the azimuth offset is extracted and the compensation coefficient is calculated to obtain the azimuth transformation compensation parameters.

[0032] Specifically, the system extracts spatial three-dimensional attitude change information from the real-time rotational deviation angle of the drill pipe relative to the geographic coordinate system, and analyzes it into rotational angle components in three independent directions: the X-axis (roll axis), Y-axis (pitch axis), and Z-axis (yaw axis). This process is achieved through an attitude decoupling algorithm, using an Euler angle transformation model or a quaternion-to-Euler angle method to separate the overall rotational state into three orthogonal upward-pointing angle components, representing the instantaneous attitude change amplitude of the drill pipe in the three principal directions in space. Based on the independent rotational angle components around the X, Y, and Z axes, the system constructs a three-dimensional rotation transformation matrix axis by axis. During the construction process, basic rotation matrices are established for each of the three axes: the rotation matrix around the X-axis represents spatial rotation in the roll direction, the rotation matrix around the Y-axis describes attitude changes in the pitch direction, and the rotation matrix around the Z-axis captures rotational behavior in the yaw direction. Each basic rotation matrix is ​​composed of trigonometric functions (sine and cosine), with the rotation angle corresponding to the axis as the input parameter. The three basic rotation matrices are cascaded and multiplied in a fixed order (e.g., ZYX or XYZ) through matrix multiplication to form a three-dimensional rotation transformation matrix, used to describe the real-time spatial mapping relationship between the orientation system and the geographic coordinate system. Geographic north alignment and magnetic declination compensation are performed based on the real-time rotation transformation matrix. Geographic north alignment mainly addresses the reference deviation between the coordinate system upon which the orientation system output depends and the actual geographic direction, correcting and aligning the heading direction represented by the Z-axis in the rotation matrix with geographic north. The system queries the global geomagnetic model based on the current geographic location to obtain the projection direction of standard north in the local spatial coordinate system. Through rotation matrix multiplication or vector rotation compensation, the projection value of the corresponding direction in the matrix is ​​made to approximate the standard geographic north, thus achieving geographic reference alignment. Simultaneously, due to the angular offset between geomagnetic north and geographic north, i.e., magnetic declination, a compensation term is introduced after the rotation matrix is ​​constructed. This compensation is achieved through an additional rotation around the vertical axis. The original transformation matrix is ​​multiplied and superimposed with a Z-axis rotation matrix containing only magnetic declination, thus constructing a corrected rotational transformation matrix that includes magnetic deviation correction. The azimuth offset is extracted based on the structural relationship of the corrected rotational transformation matrix. The extraction process uses the relative azimuth vector currently output by the orientation instrument as the transformation input, and uses the corrected rotational matrix to spatially map this vector, obtaining its projection result in the geographic coordinate system. By comparing the difference in direction on the horizontal plane before and after mapping, the azimuth offset caused by the change in drill pipe attitude is calculated. This offset reflects the magnetic field projection deviation caused by the rotation of the drill pipe and is a dynamic error compensation parameter for the original orientation instrument data. To enhance the adaptability of the compensation process under different working conditions, factors such as vibration intensity, rate of change of angular velocity, and rotational inertia are introduced. A multi-factor function is constructed to dynamically adjust the offset, and the compensation coefficient is determined through a weighted strategy or fuzzy rule mechanism.These compensation coefficients are used to correct the orientation instrument's output angle data in real time, and ultimately form a set of azimuth conversion compensation parameters that include angle increment, vector direction rotation gain, and attitude disturbance tolerance.

[0033] In one specific embodiment, the process of executing step S104 may specifically include the following steps: The azimuth angle conversion compensation parameters are input into the adaptive compensation controller to analyze the compensation coefficient and calculate the dynamic adjustment factor, thereby obtaining the real-time rotation compensation coefficient and fuzzy PID control parameters. Based on the real-time rotation compensation coefficient, the relative azimuth angle data of the drill pipe is corrected for angular offset and rotation error is eliminated to obtain the azimuth angle data after rotation compensation. Geographic absolute azimuth data is obtained by performing matrix multiplication and coordinate system mapping based on the azimuth data after rotation compensation and fuzzy PID control parameters.

[0034] Specifically, the azimuth angle conversion compensation parameters are input into the adaptive compensation controller, which integrates a compensation coefficient analysis module and a fuzzy PID control unit. Based on the current spatial rotation state of the drill pipe and historical disturbance feedback, it automatically completes the analysis of the compensation model and the optimization scheduling of control parameters. In this process, the compensation controller receives a multi-dimensional compensation parameter vector including azimuth offset, attitude disturbance components, magnetic deviation correction parameters, and rotation trend weights. According to the rotation state mapping function set internally by the controller, these input parameters are mapped into several structured variables, including key indicators such as attitude instability, azimuth error intensity, and direction drift rate. The controller performs logical partitioning and attribution analysis on these variables, combined with existing rotation behavior classification rules, to determine the current dynamic state of the drill pipe. Based on the above analysis results, the system constructs a compensation strategy structure corresponding to the rotation state and extracts real-time rotation compensation coefficients that conform to the current working conditions. These coefficients are used to control the correction magnitude and response speed of the angle correction. Simultaneously, the fuzzy PID control unit within the controller is activated, taking the current azimuth error e(t) and error change rate de(t) / dt as inputs. Fuzzy inference is performed using membership functions and a rule base, outputting three dynamic control factors: proportional gain coefficient Kp, integral gain coefficient Ki, and derivative gain coefficient Kd. To prevent controller overshoot or instability under conditions of frequent vibration or rapid state changes, the fuzzy PID unit simultaneously executes boundary constraints and fine-tuning strategies when outputting control parameters, ensuring that each compensation action is adaptive, robust, and continuous, outputting a set of highly targeted fuzzy PID control parameters. The controller applies the currently output rotation compensation coefficient to the original relative azimuth data measured by the orientation instrument, correcting angle drift and measurement errors caused by drill pipe attitude changes, sensor installation deviations, or geomagnetic interference through weighted superposition. The system performs mathematical operations on the original azimuth value and compensation coefficient, performing angle offset correction calculations. Simultaneously, a differential dynamic elimination term is introduced to suppress short-term instability characteristics in the angle-time change curve, addressing the micro-fluctuations caused by periodic disturbances. The output of this process is the rotationally compensated azimuth data. This azimuth data, along with fuzzy PID control parameters, is input into the matrix control module to initiate the three-dimensional coordinate transformation calculation. This module uses the rotationally compensated azimuth data as the spatial vector input source, combined with a response adjustment matrix constructed from the PID parameters, to achieve the mapping of angles between coordinate systems through matrix multiplication. Specifically, the system takes the representation of the current angle vector in the orientation instrument coordinate system as input and dynamically adjusts the cosine elements of each direction of the rotation matrix based on the PID adjustment, forming a real-time changing direction correction matrix. After the rotation matrix is ​​constructed, the system applies it to the compensated azimuth vector, completing the coordinate mapping from the local sensor coordinate system to the global geographic coordinate system.To improve the accuracy of the transformation, a geographic reference correction process is introduced after the matrix operation is completed. The final output angle is adjusted in the heading direction by using preset magnetic declination correction parameters, and finally the geographic absolute azimuth data is obtained.

[0035] In one specific embodiment, the process of inputting the azimuth angle conversion compensation parameters into the adaptive compensation controller for compensation coefficient analysis and dynamic adjustment factor calculation to obtain the real-time rotation compensation coefficient and fuzzy PID control parameters can specifically include the following steps: Numerical analysis and error feature extraction were performed on the azimuth conversion compensation parameters to obtain the azimuth compensation error amount and error change rate; The azimuth angle compensation error and the error change rate are input into the adaptive compensation controller for membership function calculation and fuzzy rule matching to obtain the fuzzy control output. Based on the fuzzy control output, the centroid method is used to defuzzify and the PID parameter adjustment is calculated to obtain the proportional coefficient adjustment, integral coefficient adjustment, and derivative coefficient adjustment. The reference PID parameters are dynamically corrected and the compensation coefficients are updated based on the proportional coefficient adjustment, integral coefficient adjustment, and derivative coefficient adjustment to obtain the real-time rotational compensation coefficients and fuzzy PID control parameters.

[0036] Specifically, numerical analysis and error modeling are performed on the azimuth conversion compensation parameters to extract quantitative indicators reflecting the fluctuation characteristics of the system's compensation performance. In this process, the compensation parameter analysis module takes the currently calculated azimuth conversion compensation amount as input and uses a combination of difference calculation and filtering smoothing to compare and analyze this value with the compensation parameters from the previous moment. This reveals the compensation inconsistencies introduced during azimuth correction due to drastic changes in drill pipe attitude, magnetic field disturbances, or system response lag, thus forming the azimuth compensation error amount at the current moment. To enhance the error's ability to reflect dynamic trends, the system introduces a differential derivative mechanism, calculating the error change rate based on the time difference between the current error amount and historical error amounts. This value effectively quantifies the growth or decay rate of the azimuth error. The azimuth compensation error amount and error change rate are input into the adaptive compensation controller, triggering the membership function calculation and rule matching mechanism within the fuzzy controller. This fuzzy controller divides the numerical range of error quantity and rate of change into multiple fuzzy linguistic variables, such as "large positive error," "medium error," and "small negative change," and configures a triangular or trapezoidal membership function for each linguistic variable to ensure sufficient overlap and transition zones between fuzzy variables, thereby enhancing the coherence and robustness of the fuzzy inference process. The controller calculates the membership values ​​of the input error quantity and rate of change on each fuzzy linguistic set, forming a two-dimensional fuzzy input space. Based on this space, it performs fuzzy rule matching processing in a built-in fuzzy rule base. This rule base, established by combining empirical summaries and control theory, includes 49 If-Then type rules. Each rule defines the control output direction corresponding to a specific set of input combinations, such as "if the error is large positive and the rate of change is medium positive, then the output is forced compensation." The system generates a set of multi-dimensional fuzzy control output results by activating, weighting, fusing, and aggregating all rules that meet the conditions. These output results indicate the relative strength and directional trend of the current control behavior at the fuzzy semantic level. The system inputs the fuzzy control output to the centroid-based defuzzification module for precise numerical processing. During this process, the system fuses all activation rule outputs using a weighted average method, based on the control value range and activation membership degree corresponding to each output linguistic variable, to obtain a precise numerical output, i.e., the defuzzified result. This value reflects the overall adjustment direction and intensity suggested by the controller under the current error state. Based on this defuzzified value, the system calculates three specific adjustment quantities: proportional coefficient adjustment, integral coefficient adjustment, and derivative coefficient adjustment, corresponding to the three types of gain parameters in the PID controller, used to adjust the system's control sensitivity and stability in fast response, steady-state tracking, and error change trends.To improve adjustment accuracy, the system introduces adjustment limiting and nonlinear scaling mechanisms. When outputting the adjustment value, upper and lower thresholds are set based on the current operating conditions, and the adjustment value is compressed or amplified using a nonlinear scaling formula for the absolute value of the error. This prevents uncontrolled output under strong disturbances or slow response under weak disturbances. These three adjustment values ​​are applied to the reference PID parameter set used by the current controller, performing a dynamic weighted correction operation to form a new parameter set: the real-time rotational compensation coefficient and the updated fuzzy PID control parameters. This correction process is performed through incremental superposition, with Kp, Ki, and Kd being added to the corresponding proportional, integral, and derivative coefficient adjustments, respectively, thus achieving adaptive parameter adjustment within each control cycle.

[0037] In one specific embodiment, the process of executing step S105 may specifically include the following steps: The expected trajectory tracking error is calculated and the error is differentiated from the geographic absolute azimuth data to obtain the azimuth tracking error signal and the error rate of change signal. Based on the azimuth tracking error signal and the error change rate signal, the sliding surface function is constructed and the switching control law is designed to obtain the sliding surface state quantity and the sliding switching control quantity. The sliding mode surface state quantity is input into the adaptive gain regulator to detect the drilling vibration interference amplitude and dynamically adjust the control gain to obtain the adaptive switching gain parameter. At the same time, the saturated function boundary layer technology is used to perform chattering elimination processing on the sliding mode switching control quantity to obtain a smooth sliding mode control signal. The drilling vibration interference component in the geographic absolute azimuth data is compensated for by adaptive switching gain parameters and smooth sliding mode control signals, and the geographic absolute azimuth signal is output.

[0038] Specifically, the system compares the current absolute azimuth data with the preset desired trajectory, performing continuous trajectory tracking error calculation. This involves extracting the difference between the current output and the ideal guidance through angle differences, forming a real-time tracking error signal. The system performs time-difference processing on this error value to obtain an error rate of change signal, reflecting the speed of error change and revealing the growth or convergence trend of the error. A sliding surface function is constructed based on the azimuth tracking error signal and the error rate of change signal. A sliding surface state variable is constructed based on the error signal and its derivative. This state variable represents the distance of the current system from the desired state and serves as the adjustment target for control behavior. Based on the fundamental principles of sliding mode control, the system designs a switching control variable to correct deviations from the sliding surface, ensuring rapid response when the error is large and stability when the error approaches zero. This switching control variable possesses disturbance suppression capabilities, effectively addressing nonlinear disturbances and uncertainties caused by high-speed rotation and strong vibration of the drill pipe. To enhance the controller's adaptability in actual drilling scenarios, an adaptive gain regulator is introduced. The controller takes the sliding surface state variable as input and dynamically adjusts the control gain parameter by detecting the vibration amplitude during the current drill pipe movement. When an increased external disturbance is detected, the system automatically increases the control gain to enhance control strength, while decreasing the gain when the disturbance weakens to avoid overcompensation, thus ensuring the controller has adaptive suppression capabilities against various disturbances. Simultaneously, to avoid chattering caused by frequent switching actions in traditional sliding mode control, the system employs saturated function boundary layer technology to smooth the control signal. When the sliding surface state variable is small, the change in control gain is gradually reduced, causing the control signal to gradually stabilize as it approaches a steady state, resulting in a smooth sliding mode control signal without high-frequency jumps. Based on the adaptive switching gain parameter and the smoothed sliding mode control signal, interference compensation is performed on the drilling vibration interference component in the geographic absolute azimuth data. The compensation process separates the effective information and high-frequency interference components in the geographic azimuth signal and superimposes the control signal as a correction value onto the disturbance component, achieving dynamic correction of the azimuth data. To ensure the stability and continuity of this correction process, output limiting protection and time averaging filtering mechanisms are implemented to avoid sudden angle jumps or misadjustments in the control signal when local disturbances change drastically. The output is the corrected and vibration-suppressed absolute geographic azimuth signal.

[0039] In one specific embodiment, the process of inputting the sliding mode surface state quantity into an adaptive gain regulator to detect drilling vibration interference amplitude and dynamically adjust the control gain to obtain adaptive switching gain parameters, and simultaneously using saturated function boundary layer technology to perform chattering elimination processing on the sliding mode switching control quantity to obtain a smooth sliding mode control signal, can specifically include the following steps: The sliding surface state variables are input into an adaptive gain regulator to extract the amplitude and separate the frequency components of the drilling vibration signal, thereby obtaining the vibration interference amplitude signal and vibration frequency characteristic parameters. Based on the vibration disturbance amplitude signal, an adaptive gain law calculation and gain boundary constraint are performed to obtain a dynamically switching gain adjustment factor. The boundary layer thickness parameter and the saturation function control law are obtained by setting the boundary layer thickness using the saturation function and smoothing the sign function on the sliding mode switching control variable. Gain parameters are fused based on the dynamic switching gain adjustment factor and vibration frequency characteristic parameters to obtain adaptive switching gain parameters. The control signal is then smoothed based on the saturation function control law and boundary layer thickness parameters to obtain a smooth sliding mode control signal.

[0040] Specifically, the sliding surface state variables are input into an adaptive gain regulator to initiate the active identification and dynamic adjustment process for drilling disturbance characteristics. This regulator uses the sliding surface state variables as input, analyzes their time-series characteristics, and employs an envelope extraction algorithm and bandpass filter bank to perform amplitude extraction and frequency component separation of the vibration signal. This yields the vibration disturbance amplitude signal representing the disturbance intensity and frequency characteristic parameters reflecting the vibration source. The amplitude signal represents the vibration energy density experienced by the drill pipe structure, while the frequency characteristic parameters characterize the dominant frequency distribution of the disturbance signal. Based on the vibration disturbance amplitude signal, an adaptive gain law is calculated. Using the vibration disturbance amplitude as input, and according to a predefined disturbance response mapping relationship, high-amplitude vibrations are converted into high-response-rate gain adjustment factors, ensuring the controller has rapid response capabilities under strong disturbance scenarios. To avoid excessive gain response or exceeding the system's capacity, a gain boundary constraint mechanism is introduced. Based on historical system stability thresholds, upper and lower limits are set on the output gain value to ensure the controller does not output excessive corrective actions due to sudden disturbances. This improves the numerical stability and physical controllability of the overall control system, and dynamically switches the output gain adjustment factor. To suppress the high-frequency chattering problem caused by the use of hard-signed functions in traditional sliding mode control, the system performs nonlinear improvements on the sliding mode control structure. The original sign control function is replaced with a saturated control function with smooth transition characteristics. During this replacement, a transition region for control behavior change, namely the boundary layer thickness parameter, is defined. This parameter delineates the buffer region between linear adjustment and nonlinear forced switching of the control law. The existence of the boundary layer prevents abrupt changes in the control signal when the error is small, allowing it to adjust smoothly and gradually within a certain range, effectively reducing the frequency and amplitude of control signal jumps. Based on this thickness parameter, the system constructs a new saturated function control law. When the control signal amplitude is less than the boundary layer threshold, a linear adjustment strategy is activated; when the amplitude exceeds the boundary layer threshold, nonlinear control behavior is switched. The adaptive switching gain parameter is obtained by fusing the gain parameter based on the dynamic switching gain adjustment factor and the vibration frequency characteristic parameter. This parameter reflects the intensity level of the current drilling environment disturbance by fusing two disturbance dimensions: frequency and amplitude. It also includes the temporal variation trend of the vibration signal, thus possessing stronger disturbance adaptability and compensation sensitivity. The control signal is smoothed according to the saturation function control law and the boundary layer thickness parameter to obtain a smooth sliding mode control signal.

[0041] The above describes the real-time monitoring method for drill pipe azimuth signal in embodiments of the present invention. The following describes the real-time monitoring device for drill pipe azimuth signal in embodiments of the present invention. Please refer to [link to relevant documentation]. Figure 2 One embodiment of the drill pipe azimuth signal real-time monitoring device of the present invention includes: The data acquisition module is used to acquire relative azimuth angle data of the drill pipe and rotational motion characteristic data of the drill pipe; The analysis module is used to input the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model to perform angular velocity integration and rotation state analysis, and obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. The calculation module is used to calculate the rotational transformation relationship between the orientation coordinate system and the geographic coordinate system based on the real-time rotational deviation angle, and obtain the azimuth transformation compensation parameters. The conversion module is used to perform rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth conversion compensation parameters to obtain the geographic absolute azimuth data. The output module is used for sliding mode control of geographic absolute azimuth data and outputs geographic absolute azimuth signal.

[0042] Through the collaborative efforts of the aforementioned components, a multi-sensor synchronous acquisition system is established by integrating the orientation instrument's magnetic field sensor with the drill pipe's accelerometer and gyroscope. This system accurately captures the six-degree-of-freedom motion of the drill pipe during drilling. A drill pipe rotation angle prediction model combined with a rotation angle recognition neural network enables precise prediction of the drill pipe's rotation deviation angle relative to the geographic coordinate system. Based on the real-time rotation deviation angle, a dynamic rotation transformation matrix is ​​constructed to achieve accurate conversion from the orientation instrument coordinate system to the geographic coordinate system, effectively eliminating the influence of drill pipe rotation on azimuth measurement. The combination of fuzzy PID control and an adaptive compensation controller enables intelligent adjustment of compensation parameters, automatically optimizing the compensation strategy based on real-time changes in drilling conditions, significantly improving the accuracy and stability of azimuth compensation. A sliding mode control algorithm combined with adaptive gain adjustment and saturated function boundary layer technology effectively suppresses the interference of drilling vibration on the azimuth signal, ensuring high-precision azimuth monitoring even in complex drilling environments. By establishing a precise mapping relationship between the drill pipe rotation state and the geographic coordinate system, accurate real-time conversion of the relative azimuth measured by the orientation instrument to the absolute azimuth required for drilling guidance is achieved.

[0043] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus, devices, and units can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0044] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a drill pipe azimuth signal real-time monitoring device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0045] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for real-time monitoring of drill pipe azimuth signal, characterized in that, include: Collect relative azimuth angle data of the drill pipe and rotational motion characteristic data of the drill pipe; The drill pipe rotation motion characteristic data are input into the drill pipe rotation angle prediction model for angular velocity integration and rotation state analysis to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. Based on the real-time rotation deviation angle, the rotation transformation relationship between the orientation instrument coordinate system and the geographic coordinate system is calculated to obtain the azimuth angle transformation compensation parameters; Based on the azimuth transformation compensation parameters, the relative azimuth data of the drill pipe is rotated and the coordinate system is transformed to obtain the geographic absolute azimuth data. Sliding mode control is applied to the geographic absolute azimuth data to output the geographic absolute azimuth signal.

2. The method for real-time monitoring of drill pipe azimuth signal according to claim 1, characterized in that, The data collected include the relative azimuth angle of the drill pipe and the rotational motion characteristic data of the drill pipe, including: The three-axis magnetic field strength components detected by the orientation instrument's magnetic field sensor and the six-axis inertial signals detected by the drill pipe's accelerometer gyroscope are simultaneously acquired to obtain the original magnetic field orientation signal and the original inertial motion signal of the drill pipe. The original magnetic field orientation signal and the original inertial motion signal of the drill rod are respectively input into a Kalman filter for noise cancellation and signal smoothing to obtain the denoised magnetic field orientation signal and the denoised inertial motion signal. Based on the sensor installation geometry, coordinate system alignment and magnetic declination correction are performed on the denoised magnetic field orientation signal to obtain a standardized magnetic field orientation signal in the drill pipe body coordinate system. The azimuth angle of the standardized magnetic field azimuth signal is calculated according to the magnetic field azimuth calculation algorithm to obtain the relative azimuth angle data of the drill pipe, and the coordinate system transformation is performed on the denoised inertial motion signal to obtain the rotational motion characteristic data of the drill pipe.

3. The method for real-time monitoring of drill pipe azimuth signal according to claim 1, characterized in that, The step of inputting the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model for angular velocity integration and rotation state analysis to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system includes: The drill pipe rotation motion characteristic data are input into the data preprocessing layer of the drill pipe rotation angle prediction model for triaxial angular velocity separation and amplitude normalization to obtain a standardized triaxial angular velocity sequence. The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain the drill pipe rotation motion trend prediction parameters. The drill pipe rotation motion trend prediction parameters are input into the angular velocity integral calculation layer of the drill pipe rotation angle prediction model for numerical integration and cumulative rotation angle calculation to obtain the cumulative rotation angle value of the drill pipe's three axes. Nonlinear feature extraction and rotation pattern recognition are performed on the cumulative rotation angle values ​​of the three axes of the drill pipe to obtain the instantaneous rotation state parameters of the drill pipe. Then, coordinate system transformation and geographic reference correction are performed on the instantaneous rotation state parameters of the drill pipe to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system.

4. The method for real-time monitoring of drill pipe azimuth signal according to claim 3, characterized in that, The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model for autoregressive moving average analysis to obtain drill pipe rotation motion trend prediction parameters, including: The standardized triaxial angular velocity sequence is input into the time series analysis layer of the drill pipe rotation angle prediction model. In the time series analysis layer, the standardized triaxial angular velocity sequence is segmented into time windows and historical data is sampled to obtain the historical angular velocity data sequence. The historical angular velocity data sequence is input into the autoregressive moving average model parameter recognizer for model order optimization and optimal parameter search to obtain the autoregressive order p value and the moving average order q value. Based on the autoregressive order p and the moving average order q, the historical angular velocity data sequence is fitted with least squares and the coefficients are solved to obtain the autoregressive coefficient vector and the moving average coefficient vector. Based on the autoregressive coefficient vector and the moving average coefficient vector, angular velocity prediction and trend extrapolation are performed to obtain the drill pipe rotation motion trend prediction parameters.

5. The method for real-time monitoring of drill pipe azimuth signal according to claim 1, characterized in that, The calculation of the rotation transformation relationship between the orientation instrument coordinate system and the geographic coordinate system based on the real-time rotation deviation angle yields the azimuth angle transformation compensation parameters, including: The real-time rotational deviation angle is separated into three rotational components and mapped to coordinate axes to obtain independent rotational angle components of the drill pipe around the X-axis, Y-axis and Z-axis; Based on the independent rotation angle components, a three-dimensional rotation transformation matrix is ​​constructed and matrix elements are calculated to obtain the real-time rotation transformation matrix from the orientation coordinate system to the geographic coordinate system. Based on the real-time rotation transformation matrix, geographic north reference alignment and magnetic declination compensation are performed to obtain the corrected rotation transformation matrix. Based on the corrected rotation transformation matrix, the azimuth offset is extracted and the compensation coefficient is calculated to obtain the azimuth transformation compensation parameters.

6. The method for real-time monitoring of drill pipe azimuth signal according to claim 1, characterized in that, The step of performing rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth transformation compensation parameters to obtain the geographic absolute azimuth data includes: The azimuth angle conversion compensation parameters are input into the adaptive compensation controller for compensation coefficient analysis and dynamic adjustment factor calculation to obtain the real-time rotation compensation coefficient and fuzzy PID control parameters. Based on the real-time rotation compensation coefficient, the relative azimuth angle data of the drill pipe is corrected for angle offset and rotation error is eliminated to obtain the azimuth angle data after rotation compensation. Based on the azimuth data after rotation compensation and the fuzzy PID control parameters, matrix multiplication and coordinate system mapping are performed to obtain the geographic absolute azimuth data.

7. The method for real-time monitoring of drill pipe azimuth signal according to claim 6, characterized in that, The step of inputting the azimuth angle conversion compensation parameters into the adaptive compensation controller for compensation coefficient analysis and dynamic adjustment factor calculation, to obtain the real-time rotation compensation coefficient and fuzzy PID control parameters, includes: Numerical analysis and error feature extraction are performed on the azimuth conversion compensation parameters to obtain the azimuth compensation error amount and error change rate; The azimuth angle compensation error and the error change rate are input into the adaptive compensation controller for membership function calculation and fuzzy rule matching to obtain the fuzzy control output. Based on the fuzzy control output, centroid method defuzzification and PID parameter adjustment are performed to obtain proportional coefficient adjustment, integral coefficient adjustment and derivative coefficient adjustment. The reference PID parameters are dynamically corrected and the compensation coefficients are updated based on the proportional coefficient adjustment, the integral coefficient adjustment, and the derivative coefficient adjustment to obtain the real-time rotational compensation coefficient and fuzzy PID control parameters.

8. The method for real-time monitoring of drill pipe azimuth signal according to claim 1, characterized in that, The step of performing sliding mode control on the geographic absolute azimuth data and outputting the geographic absolute azimuth signal includes: The expected trajectory tracking error is calculated and the error is differentiated on the geographic absolute azimuth data to obtain the azimuth tracking error signal and the error rate of change signal. Based on the azimuth tracking error signal and the error change rate signal, the sliding surface function is constructed and the switching control law is designed to obtain the sliding surface state quantity and the sliding switching control quantity. The sliding mode surface state quantity is input into an adaptive gain regulator to detect drilling vibration interference amplitude and dynamically adjust the control gain to obtain adaptive switching gain parameters. At the same time, the saturated function boundary layer technology is used to perform chattering elimination processing on the sliding mode switching control quantity to obtain a smooth sliding mode control signal. The drilling vibration interference component in the geographic absolute azimuth data is compensated for based on the adaptive switching gain parameter and the smooth sliding mode control signal, and the geographic absolute azimuth signal is output.

9. The method for real-time monitoring of drill pipe azimuth signal according to claim 8, characterized in that, The process involves inputting the sliding mode surface state quantity into an adaptive gain regulator to detect drilling vibration interference amplitude and dynamically adjust the control gain, thereby obtaining adaptive switching gain parameters. Simultaneously, saturated function boundary layer technology is used to perform chattering elimination processing on the sliding mode switching control quantity, resulting in a smooth sliding mode control signal. This includes: The sliding surface state quantity is input into an adaptive gain regulator to extract the amplitude and separate the frequency components of the drilling vibration signal, thereby obtaining the vibration interference amplitude signal and vibration frequency characteristic parameters. Based on the vibration interference amplitude signal, an adaptive gain law calculation and gain boundary constraint are performed to obtain a dynamically switching gain adjustment factor. The sliding mode switching control variable is subjected to saturation function boundary layer thickness setting and sign function smoothing replacement processing to obtain boundary layer thickness parameters and saturation function control law; The gain parameters are fused according to the dynamic switching gain adjustment factor and the vibration frequency characteristic parameters to obtain the adaptive switching gain parameters. The control signal is then smoothed according to the saturation function control law and the boundary layer thickness parameters to obtain the smooth sliding mode control signal.

10. A real-time monitoring device for drill pipe azimuth signal, characterized in that, A method for real-time monitoring of drill pipe azimuth signal as described in any one of claims 1-9, comprising: The data acquisition module is used to acquire relative azimuth angle data of the drill pipe and rotational motion characteristic data of the drill pipe; The analysis module is used to input the drill pipe rotation motion characteristic data into the drill pipe rotation angle prediction model to perform angular velocity integration and rotation state analysis, so as to obtain the real-time rotation deviation angle of the drill pipe relative to the geographic coordinate system. The calculation module is used to calculate the rotation transformation relationship between the orientation coordinate system and the geographic coordinate system based on the real-time rotation deviation angle, and obtain the azimuth angle transformation compensation parameters. The conversion module is used to perform rotation compensation and coordinate system transformation on the relative azimuth data of the drill pipe according to the azimuth conversion compensation parameters to obtain the geographic absolute azimuth data. The output module is used to perform sliding mode control on the geographic absolute azimuth data and output the geographic absolute azimuth signal.

Images