Well inclination azimuth measurement system based on high-precision compensation algorithm

Abstract

The present invention belongs to the field of data processing technology, and specifically relates to a well inclination azimuth measurement system based on a high-precision compensation algorithm. The system includes: a reference vector construction part, a compensation part and a Kalman filter measurement part; the reference vector construction part is used to determine the reference reference of the geomagnetic field and gravity field in the downhole environment; the compensation part compensates the original sensor data measured by the downhole sensor group, and estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data; calculates the attitude change data; constructs a local interference field model according to the attitude change data, and obtains the corrected magnetometer data; the Kalman filter measurement part is used to model all error sources as state vectors, and correct the measurement errors in real time through the Kalman filter algorithm to output the final well inclination azimuth. The present invention significantly improves the accuracy, stability and robustness of well inclination azimuth measurement.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a well inclination azimuth measurement system based on a high-precision compensation algorithm. Background Art

[0002] Well inclination and azimuth measurement is an indispensable technology in the drilling process of oil and gas wells, geothermal wells and mines. Its purpose is to ensure that the design of the wellbore trajectory is consistent with the construction requirements by monitoring the posture and orientation of the drilling tool, thereby improving the safety and efficiency of the drilling operation. In the underground environment, the complex geological structure, strong temperature and pressure changes, and local magnetic field interference bring huge technical challenges to well inclination and azimuth measurement. In the existing technology, the data fusion method based on inertial sensors and magnetic sensors is generally used to solve the well inclination and azimuth, but due to the influence of error accumulation, dynamic interference and environmental factors, there are still some problems that need to be solved.

[0003] At present, most well inclination and azimuth measurement systems are based on inertial navigation technology (INS) and geomagnetic measurement technology. Accelerometers, gyroscopes, magnetometers and other sensors are used to obtain the acceleration, angular velocity and geomagnetic field data of downhole tools, and then the gravity field and geomagnetic field reference are combined to solve the well inclination and azimuth. Inertial sensors (such as gyroscopes and accelerometers) can provide linear acceleration and angular velocity information of downhole tools. Through time integration, the attitude and position of downhole tools can be obtained in theory. However, due to the common noise, zero bias and drift problems of inertial sensors, long-term operation will lead to rapid amplification of cumulative errors, which will eventually affect the measurement accuracy. Some studies have proposed methods of using low-cost micro-electromechanical system (MEMS) inertial sensors, but this type of sensor is greatly affected by drift and noise and is only suitable for short-term high-precision measurements. Magnetometers can sense the three-axis geomagnetic field components of downhole tools and calculate the azimuth of the tool by comparing with the geomagnetic field reference vector. However, the geomagnetic field is significantly affected by the local environment downhole, and the local interference field may be caused by ferromagnetic materials around the wellbore, electromagnetic devices or magnetic anomalies in geological structures. These interferences can cause significant deviations between the magnetometer measurement data and the actual geomagnetic field distribution. Although the prior art has proposed some interference field compensation methods, such as using geomagnetic models or local static field correction, these methods cannot process the interference field in real time in a dynamic and complex environment, especially when the downhole tool posture changes rapidly, the correction accuracy will be significantly reduced. Summary of the invention

[0004] The main purpose of the present invention is to provide a well inclination azimuth measurement system based on a high-precision compensation algorithm, which has achieved significant improvements in the accuracy, stability and robustness of the well inclination azimuth measurement system.

[0005] In order to solve the above technical problems, the present invention provides a well inclination azimuth measurement system based on a high-precision compensation algorithm, the system comprising: a reference vector construction part, a compensation part and a Kalman filter measurement part; the reference vector construction part is used to construct a geomagnetic field reference vector and a gravity field reference vector according to a geomagnetic model and characteristics of the earth's gravity field, and determine the reference reference of the geomagnetic field and gravity field in the downhole environment; the compensation part calculates the temperature and pressure compensation coefficient according to the coupling relationship between temperature and pressure, and compensates the original sensor data obtained by the downhole sensor group in combination with the geomagnetic field reference vector to obtain compensated data, and estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data; according to the downhole initial attitude angle, the accumulated error of the gyroscope data is corrected, and the attitude change data is calculated; according to the attitude change data, a local interference field model is constructed, and the interference field component is eliminated from the compensated magnetometer data to obtain the corrected magnetometer data; the Kalman filter measurement part is used to model all error sources as state vectors, and correct the measurement error in real time through the Kalman filter algorithm to output the final well inclination azimuth.

[0006] Furthermore, the raw sensor data is a vector, whose vector elements include: raw accelerometer data, raw magnetometer data, raw gyroscope data, raw downhole depth, raw temperature and raw pressure; the well inclination azimuth includes: well inclination angle and azimuth angle; the downhole sensor group includes: accelerometer, magnetometer, gyroscope, depth sensor, temperature sensor and pressure sensor; the downhole initial attitude angle includes: initial well inclination angle and initial azimuth angle.

[0007] Furthermore, the reference vector construction part uses the following formula to construct the geomagnetic field reference vector according to the geomagnetic model and the characteristics of the earth's gravity field: and the gravity field reference vector :

[0008] ;

[0009] ;

[0010] in, is the surface geomagnetic field strength; is the geomagnetic inclination, which indicates the angle between the geomagnetic field and the horizontal plane; is the magnetic declination, which represents the angle between magnetic north and geographic north; is the geomagnetic field attenuation coefficient, which indicates the linear rate of change of geomagnetic field intensity with depth, and is the set value; is the original downhole depth; is the characteristic attenuation depth, which indicates the characteristic depth of the geomagnetic field exponentially decaying with depth, and is the set value; is the acceleration of gravity on the surface; is the radius of the Earth.

[0011] Furthermore, the compensation part calculates the temperature and pressure compensation coefficient according to the coupling relationship between temperature and pressure through the following formula: :

[0012] ;

[0013] in, is the original temperature; is the original pressure; is the average temperature, which is the mean of the temperature measured historically; is the average pressure, which is the mean of the pressure measured historically; is the pressure coefficient, ranging from 0.5 to 0.9; is the temperature-pressure coupling coefficient, ranging from 1 to 1.5; is the Boltzmann constant; is the temperature characteristic energy of the sensor, which is determined by calibrating the temperature sensor. Then, the raw sensor data measured by the downhole sensor group is compensated by combining the geomagnetic field reference vector through the following formula to obtain the compensated data:

[0014] ;

[0015] in, Raw sensor data; is the data after compensation; is the reference vector of the Earth's magnetic field Strength; is the raw magnetometer data strength.

[0016] Furthermore, the compensation part estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data by the following formula:

[0017] ;

[0018] ;

[0019] ;

[0020] ;

[0021] in, is the compensated accelerometer data; is the normalized acceleration vector; is the initial well inclination angle in the initial downhole attitude angle; is the horizontal component; is the compensated magnetometer data; The initial azimuth in the initial attitude angle downhole.

[0022] Furthermore, the compensation part uses the following formula to correct the accumulated error of the gyroscope data according to the initial attitude angle downhole and calculate the attitude change data:

[0023] ;

[0024] ;

[0025] in, is the raw gyroscope data, which is an angular velocity; To compensate the gyro data; is the posture change data; is the initial attitude angle underground; ; is the angular velocity of the Earth's rotation; is the geographical latitude of the measuring point where the gyroscope is located; is the current time; is the first time integration variable, ranging from 0 to ; is the vector cross product; is the second time integral variable, ranging from 0 to ; is the matrix transpose operation.

[0026] Furthermore, the compensation part uses the following formula to construct a local interference field model based on the attitude change data, remove the interference field component from the compensated magnetometer data, and obtain the corrected magnetometer data: :

[0027] ;

[0028] ;

[0029] in, is the interference field gradient vector; is the volume of the local space underground; is the vector position of any point in the local space underground; express The rotation matrix of .

[0030] Furthermore, the Kalman filter measurement part models all error sources as state vectors using the following formula: :

[0031] ;

[0032] ;

[0033] in, for The first element of for The second element of , , , and Both elements.

[0034] Furthermore, the Kalman filter measurement part corrects the measurement error in real time based on the state vector through the Kalman filter algorithm:

[0035] ;

[0036] ;

[0037] ;

[0038] ;

[0039] in, The time step index of the filter, ranging from 1 to ; The upper limit of the time step index of the set filter; is the state transfer matrix of the Kalman filter; is process noise, which satisfies Gaussian distribution; For the The state vector of the Kalman filter time step; For the The state vector of the Kalman filter time step; For the The measurement error vector for each Kalman filter time step; For the The measurement matrix for Kalman filter time steps; is the measurement noise, which satisfies the Gaussian distribution and has a covariance of ; is the forgetting factor, ranging from 0.9 to 1; is the measurement matrix of the initial Kalman filter time step; is the updated state covariance matrix; is the predicted state covariance matrix; For the The state estimate for each Kalman filter time step; No. The state estimate for each Kalman filter time step; is the Kalman gain.

[0040] Furthermore, the Kalman filter measurement part outputs the final well inclination azimuth through the following formula: :

[0041] ;

[0042] ;

[0043] in, for The fourth element of for The first element of is the final first intermediate variable; for The fifth element of for The second element of for The third element of is the final second intermediate variable.

[0044] The well inclination azimuth measurement system based on the high-precision compensation algorithm of the present invention has the following beneficial effects:

[0045] The present invention adopts a combination of dynamic modeling and integral correction in interference field correction, which significantly improves the adaptability of the system in complex geomagnetic environments. The distribution of the magnetic field downhole is affected by many factors such as local ferromagnetic materials, electromagnetic equipment and magnetization anomalies in geological structures. Traditional static models are difficult to adapt to the dynamic changes of the magnetic field. The present invention constructs a local interference field gradient model by using compensated sensor data and attitude change information, and calculates the cumulative effect of the interference field by integrating the local space, thereby eliminating the interference component in the magnetometer measurement value. This method can track and correct dynamic magnetic field interference in real time, so that the corrected magnetometer data is closer to the real geomagnetic field characteristics, providing a reliable benchmark for downhole orientation measurement. This dynamic interference field correction technology overcomes the limitations of the static compensation method in the prior art and shows significant advantages in dynamic and complex environments.

[0046] The present invention significantly reduces the cumulative error of the system by dynamically correcting the initial attitude angle. In traditional methods, the initial attitude angle is usually obtained by a one-time solution. As time goes by, the degree of deviation from the true value will continue to increase, especially in the case of long-term operation or frequent changes in attitude. The present invention corrects the initial attitude angle by using the error component in the state vector in real time, and dynamically adjusts the calculation results of the well inclination angle and azimuth angle, so that the system can effectively suppress the cumulative deviation during long-term operation. In particular, the joint correction mechanism of the magnetic field error and the acceleration error further improves the accuracy of the attitude solution, so that the final calculated well inclination azimuth is closer to the actual value.

[0047] The present invention is significantly innovative in error compensation. By introducing a temperature and pressure coupling compensation model, it comprehensively solves the impact of high temperature and high pressure environment on sensor data accuracy. In an underground environment, temperature and pressure fluctuations will cause nonlinear effects on the measurement results of accelerometers, magnetometers and gyroscopes, while the single temperature or pressure compensation model in the prior art usually ignores the coupling effect of the two. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0049] Figure 1 A schematic diagram of the system structure of a well inclination azimuth measurement system based on a high-precision compensation algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The method of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments of the present invention.

[0051] Example 1, reference Figure 1:A well inclination azimuth measurement system based on a high-precision compensation algorithm, the system comprising: a reference vector construction part, a compensation part and a Kalman filter measurement part; the reference vector construction part is used to construct a geomagnetic field reference vector and a gravity field reference vector according to a geomagnetic model and characteristics of the earth's gravity field, and determine the reference reference of the geomagnetic field and gravity field in the downhole environment; the compensation part calculates the temperature and pressure compensation coefficient according to the coupling relationship between temperature and pressure, and compensates the original sensor data obtained by the downhole sensor group in combination with the geomagnetic field reference vector to obtain compensated data, and estimates the downhole initial attitude angle according to the compensated accelerometer data and compensated magnetometer data in the compensated data; according to the downhole initial attitude angle, the accumulated error of the gyroscope data is corrected, and the attitude change data is calculated; according to the attitude change data, a local interference field model is constructed, and the interference field component is eliminated from the compensated magnetometer data to obtain the corrected magnetometer data; the Kalman filter measurement part is used to model all error sources as state vectors, and correct the measurement error in real time through the Kalman filter algorithm to output the final well inclination azimuth.

[0052] Specifically, first of all, the construction of the reference vectors of the geomagnetic field and gravity field is the theoretical basis of this technology. The geomagnetic field is a vector field generated by the movement of molten metal in the core of the earth, which exhibits stable distribution characteristics on the surface and near the surface of the earth. Through the international geomagnetic reference field model, the strength and direction of the geomagnetic field at a certain geographical location can be accurately calculated. The geomagnetic field reference vector plays a reference role in sensor measurement calibration in this system. At the same time, the characteristics of the earth's gravity field can be characterized by the distribution of gravitational acceleration, which is perpendicular to the surface in an ideal state and close to the standard gravity value. The significance of the reference vector is that they provide a reference frame under ideal conditions, which can be used to verify the output data of the downhole sensor. In the complex downhole environment, due to geological characteristics and equipment disturbances, the geomagnetic field and gravity field may deviate, and the existence of the reference vector enables the system to effectively identify and eliminate these abnormal deviations. Secondly, the temperature and pressure compensation algorithm is a key link in achieving measurement accuracy. In the downhole environment, the working conditions of the sensor are often accompanied by significant temperature and pressure changes, which will have a significant impact on the measurement results of the sensor. For example, accelerometers and magnetometers will experience sensitivity changes and zero bias drift under high temperature and high pressure conditions. This source of error has certain nonlinear characteristics and a complex coupling relationship with temperature and pressure. The compensation algorithm establishes a mathematical model of the impact of temperature and pressure on sensor output data through experimental calibration, thereby dynamically adjusting the sensor data during real-time measurement. The core of this model is to capture the system errors caused by temperature and pressure, and through the application of compensation coefficients, the compensated data is closer to the true value. This compensation not only enhances the accuracy of the original sensor data, but also provides high-quality input for the subsequent calculation of attitude angles.

[0053] Based on the compensated data, the system calculates the initial attitude angle of the downhole equipment. The attitude angle is a key parameter that describes the orientation of the equipment in three-dimensional space, and its calculation depends on the directional information of the gravity field and the geomagnetic field. Specifically, the angle between the gravity acceleration vector measured by the sensor and the reference vector of the earth's gravity field can be used to estimate the pitch angle and roll angle of the equipment. Similarly, by comparing the geomagnetic vector output by the sensor with the geomagnetic reference vector, the heading angle of the equipment can be calculated. Here, the gravity field provides a reference for the vertical direction, while the geomagnetic field provides a constraint on the horizontal direction. The calculation process of the attitude angle involves mathematical methods of vector projection and angle solution. Through this calculation method based on vector relations, the system can accurately estimate the initial attitude angle of the equipment, providing initial conditions for subsequent dynamic attitude tracking. However, due to the cumulative error of the gyroscope during the measurement process, the attitude change data output by it may drift over time. This drift is caused by the zero bias error and random noise of the gyroscope, which is particularly significant in dynamic measurement. This system effectively reduces the impact of cumulative errors by correcting the gyroscope output using the initial attitude angle. Specifically, when the gyroscope's angular velocity data is integrated over time to obtain attitude changes, the result will gradually deviate from the true value. The system uses the initial attitude angle as a reference point and corrects the gyroscope's output through real-time attitude constraints. This correction method based on the reference vector can significantly improve the long-term stability and dynamic accuracy of attitude angle measurement.

[0054] Furthermore, the system models and eliminates the magnetic field interference problem in the underground environment. The complexity of the underground environment means that the output of the magnetometer not only contains geomagnetic field information, but may also be affected by local interference fields. For example, underground metal equipment and magnetic minerals in the formation will introduce additional magnetic field components. In order to eliminate these interference field components from the sensor data, the system first estimates the interference field characteristics at the location of the equipment through the attitude angle calculation results. Then, combined with the compensated magnetometer data, a local interference field model is constructed to separate the interference component from the actual measurement value. The core of this process is to accurately identify the difference between the interference field and the geomagnetic field and eliminate the component through mathematical methods. After this stage of processing, the corrected magnetometer data is closer to the real geomagnetic field, providing a more reliable input for azimuth calculation. Finally, the system uses the Kalman filter algorithm to dynamically model and correct the error sources in the measurement process in real time. Kalman filtering is a recursive optimization algorithm that can establish a dynamic balance between the system state and the measurement data to achieve the optimal estimation of the error. In this system, Kalman filtering models the device's attitude, orientation, and various errors as state vectors, and continuously corrects the measurement results by fusing sensor data and state predictions. The core idea of ​​Kalman filtering is to incorporate the statistical characteristics of system noise and measurement noise into the optimization target, and readjust the estimated value of the state vector at each measurement update. This dynamic filtering method can not only correct system errors in real time, but also adapt to environmental changes, providing a guarantee for the system's high-precision output.

[0055] Embodiment 2: The original sensor data is a vector, whose vector elements include: original accelerometer data, original magnetometer data, original gyroscope data, original downhole depth, original temperature and original pressure; the well inclination azimuth includes: well inclination angle and azimuth angle; the downhole sensor group includes: accelerometer, magnetometer, gyroscope, depth sensor, temperature sensor and pressure sensor; the downhole initial attitude angle includes: initial well inclination angle and initial azimuth angle.

[0056] Specifically, the downhole sensor group is composed of a variety of sensors that measure acceleration, magnetic field strength, angular velocity, depth, temperature and pressure. The fusion of these data is the core of the system's precise measurement. The raw accelerometer data is used to sense the projection of gravity acceleration in different directions. Combined with the characteristics of the earth's gravity field, it can be used to derive the inclination angle information of the device in space. The raw magnetometer data is used to measure the strength and direction of the geomagnetic field and is a key input for calculating the azimuth of the device. The gyroscope measures the angular velocity of the device and can capture the dynamic changes of the device's attitude over time. In addition, the depth sensor provides vertical depth information of the device's location, while temperature and pressure data are important parameters for evaluating and compensating for the impact of the environment on the sensor's measurement accuracy.

[0057] The initial attitude angle downhole includes the initial well inclination angle and the initial azimuth angle, which are two key parameters describing the direction of the device in the three-dimensional space downhole. The initial well inclination angle is derived from the relationship between the accelerometer data and the direction of the earth's gravity. The accelerometer measures the total acceleration of the device, which includes the component of the earth's gravity. By solving the gravity component, the inclination angle of the device relative to the vertical direction of the earth can be obtained. The initial azimuth angle is calculated by the relationship between the magnetometer data and the direction of the geomagnetic field. Although the distribution of the geomagnetic field downhole may be subject to local interference, its overall trend can still provide a reliable direction reference. By combining the reference vector of the geomagnetic field, the system can deduce the rotation angle of the device on the horizontal plane. The high temperature and high pressure conditions in the complex environment downhole have a significant impact on the performance of the sensor, such as sensitivity changes and zero bias drift of the accelerometer and magnetometer. Temperature and pressure data are used to evaluate this impact in real time. The system constructs a coupled model of temperature and pressure through pre-calibration experiments and introduces it into the compensation algorithm. Specifically, the system dynamically adjusts the data output by the sensor based on the real-time collected temperature and pressure information, so that these data are closer to their true values. The data from the depth sensor also provides important support for compensation and correction. The geomagnetic field characteristics and environmental conditions at different depths may vary. Depth information can be used as a variable in the compensation model to further improve the compensation accuracy. All the data output by the sensor is integrated into a vector, and each element corresponds to the measurement value of a sensor. This data organization method helps to uniformly process different types of sensor data, enabling the system to vectorize the data through algorithms and perform compensation and correction on this basis.

[0058] The well inclination azimuth includes the well inclination angle and azimuth angle, which are the key parameters output during the dynamic measurement process. Based on the initial attitude angle, the system updates the attitude angle of the equipment in real time through the angular velocity data captured by the gyroscope. However, the angular velocity measurement of the gyroscope is inevitably affected by zero drift and cumulative errors. Therefore, the system combines the data of the magnetometer and accelerometer to constrain and correct the gyroscope output. This process relies on mathematical filtering algorithms, which can effectively reduce the impact of cumulative errors in dynamic measurements. In addition, the output data of the magnetometer may be interfered by local magnetic fields. The system obtains corrected magnetometer data by eliminating the interference field components. This process involves the decomposition and reconstruction of the magnetic field components, and uses the spatial distribution characteristics of the magnetometer data to separate the geomagnetic field and the interference field. The corrected magnetometer data provides a reliable basis for the calculation of the azimuth angle.

[0059] Embodiment 3: The reference vector construction part uses the following formula to construct the geomagnetic field reference vector according to the geomagnetic model and the characteristics of the earth's gravity field: and the gravity field reference vector :

[0060] ;

[0061] ;

[0062] in, is the surface geomagnetic field strength; is the geomagnetic inclination, which indicates the angle between the geomagnetic field and the horizontal plane; is the magnetic declination, which represents the angle between magnetic north and geographic north; is the geomagnetic field attenuation coefficient, which indicates the linear rate of change of geomagnetic field intensity with depth, and is the set value; is the original downhole depth; is the characteristic attenuation depth, which indicates the characteristic depth of the geomagnetic field exponentially decaying with depth, and is the set value; is the acceleration of gravity on the surface; is the radius of the Earth.

[0063] Specifically, the geomagnetic field reference vector The construction is based on the strength of the Earth's magnetic field at the surface. , geomagnetic inclination and magnetic declination The three components of the geomagnetic field are distributed along the local coordinate axes of the well deviation tool. Considering the nonlinear change of the geomagnetic field in the depth direction, the present invention combines the linear attenuation coefficient of the geomagnetic field and an exponential decay factor , establishing a depth Mathematical description of the Earth's magnetic field at . The linear part of the formula Characterizes the gradual attenuation of the geomagnetic field component with depth, and the exponential factor The nonlinear characteristics of attenuation are further corrected, and the complexity of the change of the geomagnetic field with depth is simulated. This modeling method significantly improves the accuracy of the geomagnetic field reference benchmark under deep well conditions. For the three directional components of the geomagnetic field vector, the present invention ensures that its directionality is consistent with the geomagnetic inclination and declination through the construction of three-dimensional components. The magnetic declination represents the angle of the geomagnetic field vector relative to the horizontal plane. describes the deviation angle of magnetic north relative to geographic north. , and Combining the depth variation characteristics of the geomagnetic field, the formula constructs a three-dimensional reference vector that changes dynamically with depth but always conforms to the geomagnetic model.

[0064] Secondly, the gravity field reference vector The construction is based on the surface gravity acceleration Assume that the direction of the gravity field is always perpendicular to the horizontal plane, that is, along Axis distribution. In order to simulate the gradual change characteristics of the earth's gravitational field, the present invention introduces a Related correction factors , describing the phenomenon that the gravitational acceleration is slightly enhanced with increasing depth. This correction not only conforms to the basic law of the earth's gravitational distribution, but also provides a stable gravity reference for the measurement of well inclination and azimuth. Unlike the geomagnetic field vector, the directionality of the gravity field vector remains unchanged, which provides a fixed gravity field reference for subsequent attitude solution and error correction. By combining the geomagnetic field reference vector and the gravity field reference vector, the present invention realizes the dynamic description and modeling of two key physical fields in the downhole environment. Compared with the simple linear model or the static model that ignores the depth change commonly used in the prior art, the vector construction formula of the present invention significantly improves the accuracy and applicability of the reference vector in different depth environments by introducing depth-related linear and nonlinear corrections. This high-precision reference vector lays a solid theoretical foundation for the subsequent steps of the compensation algorithm, including temperature and pressure compensation of the original sensor data, solution of the attitude angle, and elimination of magnetic field interference. More importantly, the construction of the reference vector is not only a static description of the geomagnetic field and gravity field, but also indirectly provides a reliable error modeling basis for the Kalman filter part. In actual measurement, the Kalman filter algorithm relies on the difference between the state vector and the observed value for error correction. The present invention greatly reduces the accumulation of initial errors through accurate modeling of the dynamic reference vector, providing support for the real-time and stability of the Kalman filter. Therefore, the construction of the geomagnetic field reference vector and the gravity field reference vector is an indispensable part of the entire system. Its theoretical design and parameterized description not only reflect the innovativeness of the technical solution, but also provide important guarantees for high-precision well inclination and azimuth measurement in practical applications.

[0065] Embodiment 4: The compensation part calculates the temperature and pressure compensation coefficient according to the coupling relationship between temperature and pressure through the following formula: :

[0066] ;

[0067] in, is the original temperature; is the original pressure; is the average temperature, which is the mean of the temperature measured historically; is the average pressure, which is the mean of the pressure measured historically; is the pressure coefficient, ranging from 0.5 to 0.9; is the temperature-pressure coupling coefficient, ranging from 1 to 1.5; is the Boltzmann constant; is the temperature characteristic energy of the sensor, which is determined by calibrating the temperature sensor. Then, the raw sensor data measured by the downhole sensor group is compensated by combining the geomagnetic field reference vector through the following formula to obtain the compensated data:

[0068] ;

[0069] in, Raw sensor data; is the data after compensation; is the reference vector of the Earth's magnetic field Strength; is the raw magnetometer data strength.

[0070] Specifically, the temperature compensation model of Boltzmann distribution theory is first adopted in the formula, and the temperature characteristic energy parameter is introduced and the Boltzmann constant , a nonlinear response relationship between temperature and sensor output is established. Specifically, the sensor output drift caused by temperature change can be regarded as a response error caused by changes in molecular energy distribution, and this drift follows an exponential change law. The terms included in the formula are This relationship is reflected in is the average temperature obtained from historical measurements and is used as a system reference value to quantify the magnitude of temperature changes. , the present invention can perform experimental calibration for different types of sensors and accurately characterize their response characteristics under different temperature conditions. Compared with the traditional linear temperature compensation model, this nonlinear model can more accurately describe the behavior of the sensor under extreme temperature conditions. Secondly, the present invention takes into account the impact of pressure changes on the sensor output and further combines the coupling between temperature and pressure. A gain correction term is used in the formula, , which is used to describe the linear and nonlinear effects of pressure on sensor output. is the pressure coefficient, which is used to quantify the direct impact of pressure changes. Its value range is experimentally determined to be 0.5 to 0.9, indicating the intensity of the pressure effect on sensor drift. At the same time, in order to further compensate for the interaction between temperature and pressure, the temperature-pressure coupling coefficient is introduced into the formula , which ranges from 1 to 1.5. This design can dynamically adjust the amplitude of pressure compensation so that it can automatically adapt according to temperature changes, thus overcoming the problem of accuracy loss caused by ignoring the temperature-pressure coupling effect in traditional compensation models. This nonlinear dynamic compensation combined with environmental parameters makes Become a highly adaptive compensating factor.

[0071] After calculating the temperature and pressure compensation coefficient, the present invention further combines the geomagnetic field reference vector To normalize the sensor output. Sensor raw data After compensation, the product term in the formula A two-layer correction effect is achieved: on the one hand, the temperature and pressure compensation coefficient eliminates the measurement error caused by environmental changes; on the other hand, the normalization correction based on the geomagnetic field reference eliminates the interference component of the geomagnetic field, so that the sensor data can more accurately reflect the real physical field state underground. This dual correction method effectively solves the problem of superposition of multi-source errors in the complex environment underground and significantly improves the robustness of the system. It is worth noting that the introduction of the geomagnetic field reference vector not only provides a reference standard for the compensation formula, but also reflects the technical advantages of the present invention in multi-physical field fusion compensation. During the compensation process, by calculating the strength of the geomagnetic field reference vector Intensity with sensor measurements The change of geomagnetic field strength is usually affected by geomagnetic field interference and the local environment of the well. Through this normalization process, the system can effectively reduce the negative impact of magnetic field interference on orientation measurement, thereby providing more accurate initial data for the subsequent Kalman filter algorithm.

[0072] Embodiment 5: The compensation part estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data by the following formula:

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] in, is the compensated accelerometer data; is the normalized acceleration vector; is the initial well inclination angle in the initial downhole attitude angle; is the horizontal component; is the compensated magnetometer data; The initial azimuth in the initial attitude angle downhole.

[0078] Specifically, the compensated accelerometer data It reflects the directional component of the gravity field downhole, and its essence is the projection value of gravity acceleration perceived by the downhole tool. However, due to the inherent error of the sensor and the interference field in the complex downhole environment, directly using the accelerometer data will lead to deviation in attitude angle calculation. Therefore, the present invention normalizes the accelerometer data and converts it into a unit vector , retaining the direction information while eliminating the amplitude error. The normalization process divides the compensated accelerometer data by its modulus so that the length of the vector is always 1. This ensures the accuracy of the directional calculation and lays the foundation for the subsequent calculation of the geometric relationship with the gravity field reference vector. The normalized acceleration vector Represents the direction of the gravity field perceived by the downhole tool. The dot product operation can be used to obtain the angle between the two, that is, the initial well inclination angle . Geometrically, the dot product reflects the projection relationship between two vectors, and by taking the inverse cosine function, the dot product result can be converted into an angle value. This calculation process utilizes the stability characteristics of the gravity field, that is, the direction of the gravity field is always perpendicular to the earth's surface, and aligns the acceleration direction sensed by the sensor with this stable reference, thereby accurately determining the well inclination angle. The calculation of the initial well inclination angle not only depends on the compensation accuracy of the sensor data, but also benefits from the high-precision modeling of the reference vector. This multi-link precise design significantly improves the reliability of the attitude angle solution.

[0079] After determining the initial well inclination, the magnetometer data is used to solve the initial azimuth . Magnetometer measurements After compensation, the geomagnetic field component at the location of the downhole tool can be accurately reflected. However, the distribution of the geomagnetic field is complex, and its direction is constrained by the geomagnetic field model and may be affected by local interference in the well. , eliminating the interference of the projection of the geomagnetic field in the vertical direction of the tool. The calculation of the horizontal component is based on the principle of vector orthogonal projection. By subtracting the component along the normalized acceleration vector direction from the magnetometer data, the pure horizontal geomagnetic field component is extracted, thereby avoiding the influence of the gravity field direction on the magnetometer data. The calculation of the initial azimuth is based on the horizontal component With the geomagnetic field reference vector By using the combination of cross product and dot product, the orthogonal and parallel components of the magnetic field direction can be considered at the same time, ensuring that the calculation result of the azimuth angle has directionality. The result of the cross product reflects the rotation relationship between the two vectors in three-dimensional space, while the dot product provides information on their directional consistency. The present invention uses these geometric properties to accurately convert the angle between the horizontal component and the reference vector into the initial azimuth angle. . In this process, the compensation processing of magnetometer data and the high-precision modeling of reference vectors once again played an important role to ensure the robustness of the solution results. The attitude angle solution method proposed in Example 5 of the present invention realizes high-precision estimation of the initial well inclination angle and the initial azimuth angle by fusion modeling of compensated sensor data and reference vectors. Through normalization processing and horizontal component extraction, the amplitude error and vertical interference in the sensor data are eliminated; secondly, the directionality and accuracy of the attitude angle calculation are ensured through the geometric relationship of the dot product and the cross product; finally, the introduction of the compensation algorithm provides a solid data foundation for the entire calculation process and significantly improves the applicability in complex downhole environments. Compared with the traditional attitude angle solution method based on static models, the present invention takes the physical field reference as the core, and through multi-level error elimination and compensation, it greatly reduces the influence of sensor errors and environmental interference on the accuracy of attitude angle solution, thereby providing more stable and accurate initial conditions for well inclination and azimuth measurement.

[0080] Embodiment 6: The compensation part uses the following formula to correct the accumulated error of the gyroscope data according to the initial attitude angle downhole and calculate the attitude change data:

[0081] ;

[0082] ;

[0083] in, is the raw gyroscope data, which is an angular velocity; To compensate the gyro data; is the posture change data; is the initial attitude angle underground; ; is the angular velocity of the Earth's rotation; is the geographical latitude of the measuring point where the gyroscope is located; is the current time; is the first time integration variable, ranging from 0 to ; is the vector cross product; is the second time integral variable, ranging from 0 to ; is the matrix transpose operation.

[0084] Specifically, the interference of the earth's rotation on the gyroscope data is particularly significant in the underground environment, especially in deep well conditions, where the earth's rotation angular velocity The projection of the gyroscope measurement value will change dynamically with the change of geographic latitude and attitude angle. Therefore, the present invention theoretically establishes a compensation model that includes the influence of the earth's rotation, and expresses the projection component of the earth's rotation angular velocity in the downhole tool coordinate system in a vectorized manner. and geographic latitude , initial well inclination , initial azimuth A three-dimensional vector is constructed by combining the two, which describes the dynamic influence of the earth's rotation on the gyroscope measurement. Specifically, this model decomposes the influence of the earth's rotation angular velocity into three components in the tool coordinate system, projects them along the three directions of the tool's local coordinates, and dynamically adjusts them according to the changes in attitude angle and latitude. Through this modeling method, the compensation model can not only accurately eliminate the errors caused by the earth's rotation, but also ensure its applicability at different depths and geographical locations. In order to further improve the accuracy of the compensation, the present invention also incorporates the influence of temperature and pressure into the correction process, using the temperature and pressure compensation coefficient The correction value of the earth's rotation projection is dynamically adjusted. In the downhole environment, changes in temperature and pressure can cause fluctuations in the sensitivity of the gyroscope, and this fluctuation may lead to the accumulation of errors in the compensation results. By combining the temperature and pressure compensation coefficient with the influence of the earth's rotation, the present invention effectively solves this problem, making the correction of the gyroscope data more accurate and robust. After compensating for the errors in the original gyroscope data, the calculation of the attitude change data becomes another core link of the present invention. The attitude change data is the accumulation of attitude changes of the downhole tool within a certain period of time, and contains angular change information in three-dimensional space. Since angular velocity is the differential of the attitude angle change, the calculation of the attitude change needs to be completed through integration. The present invention calculates the attitude change by calculating the compensated angular velocity data. Perform two-level integration to accurately solve the posture change data . First, the first layer of time integration is performed to calculate the cumulative effect of angular velocity on time, and the main attitude changes are obtained. Subsequently, in order to further compensate for the nonlinear errors that may be introduced in the process of angular velocity changes, the present invention introduces a second-order time integral term based on vector cross product. This term greatly improves the accuracy of attitude change data calculation by simulating the rate of change of angular velocity in the time dimension and its secondary effect on the overall attitude change. By combining the first-order and second-order integrations, the present invention can not only accurately describe attitude changes, but also effectively compensate for the cumulative deviations caused by ignoring high-order errors in traditional methods. In addition, the calculation of attitude change data also needs to be based on the initial attitude angle downhole. The initial attitude angle is accurately calculated by the method of Example 5, including the initial well inclination angle and the initial azimuth . By combining the initial conditions, the present invention can ensure the precise alignment of the attitude change data with the actual tool attitude downhole, thereby avoiding the result deviation caused by the propagation of initial errors. Compared with the traditional attitude solution method, the present invention combines dynamic compensation and multi-layer integration, which not only solves the drift error and environmental impact problems in the gyroscope data, but also significantly improves the accuracy of attitude change calculation through high-order integral models. In practical applications, this method can adapt to a variety of changing conditions in complex downhole environments, including temperature, pressure and geomagnetic field interference. Traditional methods are usually difficult to deal with the comprehensive impact of environmental factors on gyroscope data, and the present invention provides a comprehensive and accurate solution for the well inclination and azimuth measurement system by organically integrating the earth's rotation compensation, temperature-pressure coupling correction and initial attitude angle solution.

[0085] Embodiment 7: The compensation part uses the following formula to construct a local interference field model based on the attitude change data, remove the interference field component from the compensated magnetometer data, and obtain the corrected magnetometer data :

[0086] ;

[0087] ;

[0088] in, is the interference field gradient vector; is the volume of the local space underground; is the vector position of any point in the local space underground; express The rotation matrix of .

[0089] Specifically, local magnetic field interference in the downhole environment is usually caused by a variety of factors, such as ferromagnetic materials inside downhole tools, interference fields generated by electromagnetic equipment, non-uniform magnetization in the formation, and magnetic anomalies adjacent to the wellbore. These interferences will be superimposed on the measurement results of the magnetometer, causing its output data to be unable to accurately reflect the true distribution of the geomagnetic field. In order to eliminate the influence of these interferences on the measurement accuracy, the present invention first describes the spatial distribution characteristics of the interference field through the interference field gradient vector. Specifically, the interference field gradient vector is based on the difference between the compensated magnetometer data and the geomagnetic field reference vector, and the partial derivative is calculated in the spatial coordinate direction, thereby obtaining the interference field in , , The rate of change of direction. This gradient description can not only reveal the spatial distribution characteristics of the interference field, but also provide key parameters for subsequent integral compensation. The gradient calculation formula is: ; shows how the interference field gradient is constructed based on the difference between the actual measurement value and the reference vector. However, since the posture of the downhole tool is always in dynamic change during the measurement process, the manifestation of the magnetic field interference also changes accordingly, so the static gradient description cannot fully capture these changes. , constructs a dynamic rotation matrix , converting the interference field gradient from the global coordinate system to the tool's current reference coordinate system. The role of this dynamic mapping is that it can adjust the direction of the interference field gradient in real time so that it is always consistent with the tool's posture change. The core of this posture adjustment method is to compensate for the magnetic field measurement error introduced by the tool rotation through the time dynamic change relationship of the tool posture angle, thereby ensuring the accuracy of the gradient calculation. The application of the rotation matrix not only enhances the adaptability of the interference model to the dynamic environment, but also enables the interference compensation algorithm to handle extreme working conditions where the tool posture changes frequently. After calculating the interference field gradient vector, the present invention further adopts an integration method to eliminate the influence of the interference field on the magnetometer data. This process traverses all points in the local space, integrates the interference field gradient, calculates the cumulative effect of the interference field, and eliminates it from the compensated magnetometer data. In the integral formula, the local space volume The definition of reflects the magnetic field distribution range in a limited area around the tool, while the vector position It is used to represent the position of any point in space. The core idea of ​​this integral compensation process is revealed: by integrating the distribution of the gradient field in space, the overall effect of the interference field can be effectively quantified and removed from the magnetometer data. This integral operation is mathematically equivalent to calculating the total impact of the local interference field on the entire tool reference system, and the result is a global interference field correction value. When this correction value is removed from the compensated magnetometer data, the remaining data only contains the true geomagnetic field component, thus forming the corrected magnetometer data. . The significant advantage of this process is that it can not only eliminate the direct impact of the interference field, but also capture the spatial variation trend of the interference field through gradient description, thereby improving the accuracy of interference compensation. Compared with the traditional magnetic field interference elimination method, the innovation of the present invention lies in the combination of dynamic compensation of attitude changes and gradient integration method of local space, which realizes the accurate modeling and elimination of dynamic and non-uniform magnetic field interference. In traditional methods, the interference field is usually simply modeled as a static or linear distribution. The present invention not only improves the flexibility of the interference field description by introducing gradient and rotation matrices, but also calculates the global interference effect by integrating the local space, avoiding the problem of error accumulation under the static model. In addition, the introduction of dynamic attitude data enables the system to maintain the ability to accurately compensate for the interference field under the conditions of the tool's constantly changing attitude and position. This multi-level and multi-dimensional interference compensation algorithm significantly improves the reliability of magnetometer data and provides a purer magnetic field benchmark for well inclination and azimuth measurement.

[0090] Embodiment 8: The Kalman filter measurement part models all error sources as state vectors using the following formula: :

[0091] ;

[0092] ;

[0093] in, for The first element of for The second element of , , , and Both elements.

[0094] Specifically, the state vector It is the core modeling object of the Kalman filter algorithm and contains five error components, corresponding to the initial well inclination error , initial azimuth error , gyroscope angular velocity error , magnetic field error and acceleration error These error components are derived from different physical fields and measurement equipment, and fully describe the possible sources of error in the wellbore position measurement process. The Kalman filter algorithm dynamically estimates and corrects these error components, making the final output wellbore position more accurate. Initial wellbore angle error and the initial azimuth error It is a key component of the state vector, reflecting the deviation between the attitude angle calculated by the gyroscope and the initial attitude angle. The calculation of these two errors is based on the attitude change data accumulated by the gyroscope. , which is the attitude change obtained by the gyroscope through time integration. and They are The well inclination and azimuth in represent the attitude angle calculated by gyroscope integration, while and is the initial attitude angle. Since the gyroscope measurement data is easily affected by drift errors and cumulative deviations, these errors will cause the calculated attitude angle to gradually deviate from the true value through the accumulation of time. This deviation is modeled as an independent error component in the state vector, which is convenient for Kalman filtering to track and correct in real time, thereby avoiding the impact of cumulative errors on the final output of the system.

[0095] Gyroscope angular velocity error It is another important component of the state vector, reflecting the actual angular velocity after compensation. The raw measurement value of the gyroscope The angular velocity measurement data of the gyroscope is not only affected by its own noise, but also by environmental parameters such as the earth's rotation, temperature, and pressure. The present invention corrects the gyroscope measurement data through the compensation formula of Example 6 to obtain a value closer to the true value. However, even after compensation, the angular velocity data may still have some residual errors, which are reflected in the state vector The Kalman filter can further optimize the accuracy of the data by describing and tracking the magnetic field error. describes the corrected magnetometer data With the geomagnetic field reference vector Because the local magnetic field interference in the downhole environment may have complex nonlinear characteristics, even after the interference field elimination process of Example 7, there may still be certain residual errors in the magnetometer data. These errors reflect the dynamic changes of the local magnetic field model and the limitations of the magnetometer compensation accuracy. The present invention models these residual errors through the magnetic field error components in the state vector, and adjusts them in real time in combination with the Kalman filter algorithm, thereby improving the adaptability of the magnetometer data in a dynamic environment. Acceleration error is the last component of the state vector, reflecting the compensated accelerometer data With the gravity field reference vector The main function of accelerometer data is to provide the acceleration component of the downhole tool relative to the gravity field, which is used to solve the attitude and motion state of the tool. However, the accelerometer data will be affected by tool vibration, temperature drift and sensor noise. These interferences will cause the measured values ​​to deviate from the true gravity field distribution. By modeling these deviations as acceleration errors , the state vector can capture and track these error sources in real time, providing more complete error information for the Kalman filter. The definition of the state vector is the basis of the Kalman filter error modeling, and the core of the Kalman filter is to correct the error in real time through state prediction and observation update. The error state transition model describes how the error component evolves over time based on the dynamic change characteristics of the downhole environment. The observation update combines the sensor data and the physical field reference vector, and dynamically adjusts the error estimate by measuring the noise and the weight distribution of the system model. This process enables the Kalman filter to not only effectively correct the current error, but also predict the future error evolution trend, providing a guarantee for the long-term stability of the system. Compared with the traditional error compensation method, the present invention overcomes the limitations of single error source processing through the comprehensive definition and dynamic modeling of the state vector, and realizes a systematic description of the complex error sources in the well. The introduction of the Kalman filter algorithm enables the system to update the error model in real time in a dynamically changing environment and maintain high-precision attitude and azimuth measurement capabilities. This multi-level, multi-source error modeling method not only improves the accuracy and stability of well inclination and azimuth measurement, but also significantly enhances the applicability of the system in complex environments, providing comprehensive technical support for high-precision downhole measurement. By real-time correction of each error component in the state vector, the present invention effectively solves the problem of accuracy degradation caused by error accumulation and dynamic interference in well inclination and azimuth measurement, further consolidating the technical advantages of the present invention in complex downhole environments.

[0096] Embodiment 9: The Kalman filter measurement part corrects the measurement error in real time based on the state vector through the Kalman filter algorithm:

[0097] ;

[0098] ;

[0099] ;

[0100] ;

[0101] in, The time step index of the filter, ranging from 1 to ; The upper limit of the time step index of the set filter; is the state transfer matrix of the Kalman filter; is process noise, which satisfies Gaussian distribution; For the The state vector of the Kalman filter time step; For the The state vector of the Kalman filter time step; For the The measurement error vector for each Kalman filter time step; For the The measurement matrix for Kalman filter time steps; is the measurement noise, which satisfies the Gaussian distribution and has a covariance of ; is the forgetting factor, ranging from 0.9 to 1; is the measurement matrix of the initial Kalman filter time step; is the updated state covariance matrix; is the predicted state covariance matrix; For the The state estimate for each Kalman filter time step; No. The state estimate for each Kalman filter time step; is the Kalman gain.

[0102] Specifically, the core of Kalman filtering is to predict the current state of the system through the state transition equation, and update and correct the predicted value in combination with the observed data. The main error components of downhole tools are described, including the deviation of the initial attitude angle, the residual error after sensor compensation, and the influence of dynamic disturbance. The state prediction part is based on the state transfer matrix and the state vector at the previous moment To estimate the current state, the state transfer matrix is ​​used to describe the evolution of the state over time, and the process noise This reflects the unpredictable interference factors in the dynamic changes of the system. This state prediction mechanism provides the filter with a preliminary estimate based on the dynamic model, enabling the system to predict the next state change based on the current state trend. Based on the state prediction, the Kalman filter further uses the observation equation to convert the measurement error into With the state vector The observation equation is connected through the measurement matrix Transform the observed data into state space and incorporate the measurement noise Describes the uncertainty of the observation value. Since the observation data in the underground environment is easily affected by interference and noise, the measurement noise covariance matrix is introduced to quantify the unreliability of observations. This observation update mechanism corrects the predicted state by the actual measurement value, thereby dynamically eliminating the deviation caused by process model approximation or measurement error. The key lies in how the filter weighs the contribution between state prediction and observation update, which is determined by the Kalman gain The calculation of the Kalman gain takes into account the predicted state covariance and the measurement noise covariance , whose size directly reflects the degree of trust the system has in the observed data. If the covariance of the predicted state is large, it means that the system has a high uncertainty about the current state. At this time, the Kalman gain will rely more on the observed data, thus giving a greater weight to the observation update; on the contrary, when the measurement noise is large, the filter will reduce the influence of the observed data to avoid the influence of noise interference on the state estimation. The dynamic adjustment of the Kalman gain is the key to the filter's ability to maintain high efficiency and robustness in a dynamic environment.

[0103] By introducing the Kalman gain, the filter updates and corrects the state vector, and uses the difference between the predicted value and the observed value (i.e., the innovation amount) to adjust the state estimate. The innovation amount represents the deviation between the predicted state and the actual observation, and its size reflects the degree of deviation between the model and the measurement. The filter adjusts the state vector so that the updated state estimate can be closer to the true value. At the same time, the state covariance matrix is ​​also dynamically adjusted during the update process to ensure that the filter's uncertainty prediction of the future state is more reasonable. In practical applications, this Kalman filter algorithm provides higher accuracy and reliability for well inclination azimuth measurement by dynamically correcting the initial attitude angle error of the downhole tool, the sensor drift error, and the deviation caused by the interference field. Compared with the traditional error compensation method, the Kalman filter algorithm in the present invention can not only capture dynamic errors in real time, but also adjust the degree of dependence on the measurement data according to environmental changes, so that the system can still maintain stable measurement performance in a complex downhole environment. This multi-level and multi-dimensional error correction mechanism effectively solves the threat of cumulative errors and dynamic interference to the accuracy of well inclination azimuth measurement, and is an important innovation of the present invention in high-precision measurement technology.

[0104] Embodiment 10: The Kalman filter measurement part outputs the final well deviation azimuth through the following formula: :

[0105] ;

[0106] ;

[0107] in, for The fourth element of for The first element of is the final first intermediate variable; for The fifth element of for The second element of for The third element of is the final second intermediate variable.

[0108] Specifically, in downhole measurement, the initial attitude angle includes the initial well inclination angle and the initial azimuth , which are calculated from the compensated accelerometer and magnetometer data. However, due to the complex and changeable downhole environment, the initial angle will gradually deviate from the true value due to the cumulative effects of sensor drift, environmental interference and tool movement. Example 10: The error component updated in real time in the state vector Dynamically correct the initial attitude angle to calculate the final correction angle and These correction angles reflect the actual changes in tool posture in three-dimensional space after error compensation. The acceleration error component is introduced into the correction formula and magnetic field error components By combining the modulus length ratio of these error components and the sine value of the initial well inclination angle, the formula describes how the error affects the well inclination angle in a dynamic environment and dynamically adjusts the angle value through the correction formula. Part of the correction formula The contribution of the interaction between magnetic field and acceleration to the correction of well inclination is emphasized. The magnetic field error provides a high-precision description of the local interference field through the compensated state quantity, while the acceleration error further corrects the vertical deviation of the attitude angle in combination with the change in the gravity field reference vector. This interactive correction mechanism effectively reduces the measurement error in complex environments, making the corrected well inclination angle closer to the true value. Azimuth The correction process also relies on the estimated error component in the state vector, and pays special attention to the gyroscope angular velocity error The gyroscope is prone to cumulative deviation of the attitude angle due to drift effect in long-term operation. The present invention couples the gyroscope error component with the acceleration error and uses a correction formula to dynamically adjust the azimuth angle value. Part of the correction formula The coupling relationship is shown in Figure 1. The acceleration error provides information about the tool's movement in the direction of the gravity field, while the gyroscope error component dynamically corrects the horizontal component of the azimuth angle in combination with the tool's rotation rate change. This design can simultaneously eliminate the dual effects of accumulated drift and local disturbance on the azimuth angle, so that the corrected azimuth angle still maintains high accuracy in a dynamic environment. Corrected well inclination angle and azimuth It is further converted into quaternion form to represent the three-dimensional attitude of the tool. The output in quaternion form has significant mathematical advantages over the traditional Euler angle, which not only avoids the singularity problem, but also can represent the rotation information in a more compact way. This conversion process maps the attitude information of the tool from the angle domain to the quaternion domain, which not only provides a standardized representation for subsequent data processing and application, but also improves the robustness of attitude fusion. The present invention realizes the gradual correction of the attitude angle, and the design of the correction formula makes full use of the interactive information of the magnetic field and acceleration to dynamically adjust the angle deviation. At the same time, the output in quaternion form not only simplifies the attitude representation, but also provides a powerful mathematical tool for subsequent attitude optimization and expansion. Compared with the traditional method, the present invention has significant advantages in applicability and measurement accuracy in complex underground environments, especially under long-term operation and strong interference conditions, it can still maintain the stability and reliability of the system. By expressing the three-dimensional attitude information in the form of quaternion, the singularity problem that may occur in the traditional Euler angle representation is solved. Quaternion has mathematical superiority and can represent three-dimensional rotation in a compact way, while avoiding the computational instability problem that may occur when the attitude angle is close to 90 degrees or 0 degrees. The output in the form of quaternions not only improves the robustness of the attitude representation, but also provides a more efficient mathematical tool for subsequent attitude fusion and dynamic compensation. In addition, the standardized output in the form of quaternions is applicable to a variety of attitude analysis algorithms, making the measurement results of the present invention more flexible and reliable in extended applications.

[0109] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. A well inclination azimuth measurement system based on a high-precision compensation algorithm, characterized in that: The system comprises: a reference vector construction part, a compensation part and a Kalman filter measurement part; the reference vector construction part is used to construct a geomagnetic field reference vector and a gravity field reference vector according to the geomagnetic model and the characteristics of the earth's gravity field, and determine the reference reference of the geomagnetic field and the gravity field in the downhole environment; the compensation part calculates the temperature pressure compensation coefficient according to the coupling relationship between temperature and pressure, and compensates the original sensor data measured by the downhole sensor group in combination with the geomagnetic field reference vector to obtain compensated data, and estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data; according to the downhole initial attitude angle, the accumulated error of the gyroscope data is corrected, and the attitude change data is calculated; according to the attitude change data, a local interference field model is constructed, and the interference field component is eliminated from the compensated magnetometer data to obtain the corrected magnetometer data; the Kalman filter measurement part is used to model all error sources as state vectors, and correct the measurement error in real time through the Kalman filter algorithm to output the final well inclination azimuth; The reference vector construction part uses the following formula to construct the geomagnetic field reference vector according to the geomagnetic model and the characteristics of the earth's gravity field: and the gravity field reference vector : ； ； in, is the surface geomagnetic field strength; is the geomagnetic inclination, which indicates the angle between the geomagnetic field and the horizontal plane; is the magnetic declination, which represents the angle between magnetic north and geographic north; is the geomagnetic field attenuation coefficient, which indicates the linear rate of change of geomagnetic field intensity with depth and is the set value; is the original downhole depth; is the characteristic attenuation depth, which indicates the characteristic depth of the geomagnetic field exponentially decaying with depth, and is the set value; is the acceleration of gravity on the surface; is the radius of the Earth.

2. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 1, characterized in that: The original sensor data is a vector, whose vector elements include: original accelerometer data, original magnetometer data, original gyroscope data, original downhole depth, original temperature and original pressure; the well inclination azimuth includes: well inclination angle and azimuth angle; the downhole sensor group includes: accelerometer, magnetometer, gyroscope, depth sensor, temperature sensor and pressure sensor; the downhole initial attitude angle includes: initial well inclination angle and initial azimuth angle.

3. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 1, characterized in that: The compensation part calculates the temperature and pressure compensation coefficient according to the coupling relationship between temperature and pressure through the following formula: : ； in, is the original temperature; is the original pressure; is the average temperature, which is the mean of the temperature measured historically; is the average pressure, which is the mean of the pressure measured historically; is the pressure coefficient, ranging from 0.5 to 0.9; is the temperature-pressure coupling coefficient, ranging from 1 to 1.5; is the Boltzmann constant; is the temperature characteristic energy of the sensor, which is determined by calibrating the temperature sensor. Then, the raw sensor data measured by the downhole sensor group is compensated by combining the geomagnetic field reference vector through the following formula to obtain the compensated data: ； in, Raw sensor data; is the data after compensation; is the reference vector of the Earth's magnetic field Strength; is the raw magnetometer data strength.

4. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 3, characterized in that: The compensation part estimates the downhole initial attitude angle according to the compensated accelerometer data and the compensated magnetometer data in the compensated data through the following formula: ； ； ； ； in, is the compensated accelerometer data; is the normalized acceleration vector; is the initial well inclination angle in the initial downhole attitude angle; is the horizontal component; is the compensated magnetometer data; The initial azimuth in the initial attitude angle downhole.

5. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 4, characterized in that: The compensation part uses the following formula to correct the accumulated error of the gyroscope data according to the initial attitude angle downhole and calculate the attitude change data: ； ； in, is the raw gyroscope data, which is an angular velocity; To compensate the gyro data; is the posture change data; is the initial attitude angle underground; ; is the angular velocity of the Earth's rotation; is the geographical latitude of the measuring point where the gyroscope is located; is the current time; is the first time integration variable, ranging from 0 to ; is the vector cross product; is the second time integral variable, ranging from 0 to ; is the matrix transpose operation.

6. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 5, characterized in that: The compensation part uses the following formula to construct a local interference field model based on the attitude change data, remove the interference field component from the compensated magnetometer data, and obtain the corrected magnetometer data : ； ； in, is the interference field gradient vector; is the volume of the local space underground; is the vector position of any point in the local space underground; express The rotation matrix of .

7. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 6, characterized in that: The Kalman filter measurement part models all error sources as state vectors using the following formula: : ； ； in, for The first element of for The second element of , , , and Both elements.

8. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 7, characterized in that: The Kalman filter measurement part corrects the measurement error in real time based on the state vector through the Kalman filter algorithm: ； ； ； ； in, The time step index of the filter, ranging from 1 to ; The upper limit of the time step index of the set filter; is the state transfer matrix of the Kalman filter; is process noise, which satisfies Gaussian distribution; For the The state vector of the Kalman filter time step; For the The state vector of the Kalman filter time step; For the The measurement error vector for each Kalman filter time step; For the The measurement matrix for Kalman filter time steps; is the measurement noise, which satisfies the Gaussian distribution and has a covariance of ; is the forgetting factor, ranging from 0.9 to 1; is the measurement matrix of the initial Kalman filter time step; is the updated state covariance matrix; is the predicted state covariance matrix; For the The state estimate for each Kalman filter time step; No. The state estimate for each Kalman filter time step; is the Kalman gain.

9. The well inclination azimuth measurement system based on high-precision compensation algorithm according to claim 8, characterized in that: The Kalman filter measurement part outputs the final well deviation azimuth through the following formula: : ； ； in, for The fourth element of for The first element of is the final first intermediate variable; for The fifth element of for The second element of for The third element of is the final second intermediate variable.

Images