A high-precision navigation pointing method and system based on satellite cooperation

Abstract

This invention discloses a high-precision navigation pointing method and system based on satellite collaboration, comprising: acquiring satellite observation data and IMU data of a preset carrier; generating carrier phase double-difference observation values; dividing the satellite inflow levels by combining the IMU data; predicting and updating the three-dimensional attitude error accumulation state of the preset carrier using a nonlinear Muskingan model; correcting the observation noise covariance based on the anomaly detection results of the Melbourne-Wübbena combination to obtain observation weights; performing a tight combination solution of attitude position based on the three-dimensional attitude error accumulation state and observation weights using Kalman filtering to obtain carrier attitude data and high-precision navigation position; constructing a northeast-sky coordinate system based on the high-precision navigation position to calculate the pointing angle deviation; and determining the navigation state confidence level based on the joint decision result. This method utilizes inflow level division and observation weight updates to improve the adaptability and accuracy of navigation pointing in complex environments, while also exhibiting good interpretability.

Description

Technical Field

[0001] This invention relates to the field of navigation, and in particular to a high-precision navigation pointing method and system based on satellite coordination. Background Technology

[0002] High-precision navigation and attitude pointing technology is widely used in fields such as autonomous driving, precision drone operations, satellite communication antenna alignment, and engineering surveying. Its performance directly determines the accuracy of operations and the reliability of missions. The current mainstream high-precision attitude determination scheme takes dual-antenna GNSS carrier phase differential technology as its core. It eliminates systematic errors such as ionospheric and clock errors through dual-difference observations between antennas and satellites. Combined with baseline geometric calculations, it obtains the heading, pitch, and roll attitude of the carrier. It improves the redundancy of calculations through collaborative observations of multiple systems and constellations such as BeiDou, GPS, and Galileo. With the carrier's own IMU inertial measurement unit, it compensates for attitude output under dynamic scenes and signal obstruction. It outputs centimeter-level positioning and attitude accuracy in open environments, providing continuous and reliable position and attitude perception capabilities for complex operations.

[0003] Existing dual-antenna GNSS attitude determination schemes still face scene limitations in complex scenarios such as urban canyons and tree-shaded areas. Dual-difference observations are susceptible to multipath interference and cycle slips, and conventional fixed-threshold cycle slip detection is prone to misjudgment or omission, leading to the accumulation and divergence of attitude errors. When modeling the carrier attitude error, linear assumptions are often used, which have limited adaptability to complex scenarios. Weak signal interference such as multipath interference in real-world scenarios cannot be effectively suppressed. Satellite collaboration is limited to the multi-system observation level, resulting in insufficient coupling between attitude calculation and pointing output, and the inability to provide reliable degraded pointing services in occluded scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a high-precision navigation pointing method and system based on satellite cooperation.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution: The first aspect of this invention provides a high-precision navigation pointing method based on satellite cooperation, comprising: Acquire dual-frequency carrier phase observations, pseudorange observations, and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; Based on the carrier phase double-difference observations and IMU inertial data, the inflow levels of each satellite data are divided using satellite cooperative geometric parameters to obtain the main inflow term and the secondary inflow term; The three-dimensional attitude error accumulation state of the preset carrier is predicted and updated based on the main inflow term and the secondary inflow term using a nonlinear Muskingan model. Anomaly detection is performed based on the pseudorange observations and dual-frequency carrier phase observations. A joint decision is made based on the detection results. The attitude calculation weights of each satellite are then corrected based on the decision results to obtain the observation weights. Based on the three-dimensional attitude error accumulation state and observation weights, attitude position is calculated by Kalman filtering to obtain carrier attitude data and high-precision navigation position. The pointing calculation is performed based on the carrier attitude data and the preset target position, and the navigation state confidence is determined by the joint decision result to obtain the navigation position and pointing angle information of the preset carrier.

[0006] Furthermore, the method for obtaining the carrier phase double-difference observations includes: A dual-antenna GNSS receiver based on a pre-defined carrier receives dual-frequency carrier phase observations and pseudorange observations from each cooperating satellite, measures the carrier-to-noise ratio of each satellite signal, and performs inter-antenna differential analysis on the dual-frequency carrier phase observations based on the baseline length and baseline direction of the dual antennas to obtain single-difference carrier phase observations. The elevation angles of each visible satellite are calculated based on the satellite ephemeris and the approximate location of the receiver, and the GDOP contribution of each satellite is calculated. The satellites are sorted in descending order based on the elevation angles. If the elevation angles are the same, they are sorted in descending order based on the carrier-to-noise ratio (CNR). If the CNRs are the same, they are sorted in descending order based on the GDOP contribution. The satellite ranked first is taken as the reference satellite. Based on the reference satellite, the single-difference carrier phase observations are differentially analyzed between satellites to generate double-difference carrier phase observations. The baseline direction is determined according to the preset carrier and the installation position of the dual antennas. The visible satellites are determined according to the health indicators of the cooperating satellites.

[0007] Further, it is characterized by comprising: The dual-frequency carrier phase observations include the first carrier phase observation and the second carrier phase observation of antenna 1, and the third carrier phase observation and the fourth carrier phase observation of antenna 2. The pseudorange observations include the first pseudorange observation and the second pseudorange observation of antenna 1, and the third pseudorange observation and the fourth pseudorange observation of antenna 2. The IMU inertial data includes the three-axis acceleration data and the three-axis angular velocity data of the preset carrier. The satellite cooperative geometric parameters include the elevation angle and the GDOP contribution.

[0008] Furthermore, the method for obtaining the primary inflow item and the secondary inflow item includes: Based on the carrier attitude quaternion from the previous epoch and the IMU inertial data, the attitude mechanical orchestration algorithm is used to extrapolate the predicted attitude angles for the current epoch. Based on the predicted attitude angles, a rotation matrix from the carrier coordinate system to the geocentric-ground-fixed coordinate system is constructed. The baseline vector in the carrier coordinate system is calculated based on the baseline length and direction of the dual antennas. The baseline vector is then transformed to the geocentric-ground-fixed coordinate system using the rotation matrix. The geocentric-ground-fixed coordinates of each antenna in the dual antennas are calculated based on the approximate location of the receiver. The geocentric-ground-fixed coordinates of each visible satellite are calculated based on the satellite ephemeris. The geometric distance from each antenna to each satellite is calculated. The inter-station single difference and inter-satellite double difference are calculated for the geometric distances to obtain the predicted geometric distances. The attitude angles include heading angle, pitch angle, and roll angle. The carrier phase double-difference observation, elevation angle, and carrier-to-noise ratio of the preset carrier at the current epoch are obtained. The carrier phase double-difference observation is subtracted from the geometric distance prediction to obtain the double-difference residual of each satellite. The cooperative weight of each satellite is calculated by normalizing the product of the square of the elevation angle and the carrier-to-noise ratio. The satellites are sorted according to the cooperative weight and divided into inflow levels according to a preset number to obtain the primary inflow satellites and secondary inflow satellites. The double-difference residuals of the primary inflow satellites and secondary inflow satellites are weighted and summed according to the cooperative weight to obtain the primary inflow term and secondary inflow term of the current epoch. The preset number is greater than or equal to 3 and less than the total number of visible satellites.

[0009] Furthermore, the method for obtaining the accumulated state of the three-dimensional attitude error includes: The update formula for the nonlinear Muskingan model is: ; in This represents the cumulative state of the three-dimensional attitude error at the current epoch, where is the three-dimensional rotation vector, and represents the preset heading angle error, pitch angle error, and roll angle error of the carrier. Here is the state transition matrix, and here is a diagonal matrix. This represents the accumulated state of the three-dimensional attitude error from the previous epoch. For the control matrix, Assign coefficients to inflow terms. The main inflow term for the current epoch. It is a hyperbolic tangent nonlinear saturation function. For the next inflow term in the current epoch, As a scale factor, It is a non-linear exponent; The three-dimensional attitude error is the rotational vector attitude error between the actual carrier attitude and the reference attitude in the current epoch, and the reference attitude is the state prediction value in the current epoch.

[0010] Furthermore, the method for obtaining the observation weights includes: The Melbourne-Wübbena combination is calculated based on dual-antenna dual-frequency carrier phase observations and pseudorange observations. The MW combination difference between adjacent epochs of each satellite is obtained. The standard deviation of the dual-antenna MW combination within the sliding window is corrected by weighting the satellite GDOP to obtain an adaptive threshold. If the MW combination difference of each dual antenna for the same satellite exceeds the adaptive threshold, the corresponding satellite is marked as an anomaly. The adaptive threshold is the product of the standard deviation and the GDOP weighting coefficient. For anomalous satellites, the observation noise covariance sub-block is corrected using a nonlinear leakage formula, which is: ; in To correct the observation noise covariance sub-block, for For the first The observation noise covariance sub-block of each satellite The preset discharge coefficient is a positive real number. For the first The MW combination difference of adjacent epochs of satellites, An adaptive threshold; Based on the modified observation noise covariance sub-blocks of each satellite, a block-diagonal global observation noise covariance matrix is ​​constructed. The global observation noise covariance matrix is ​​inverted to obtain the attitude calculation weight matrix. The diagonal elements of the inverse matrix corresponding to the modified observation noise covariance sub-block are used as observation weights according to the attitude calculation weight matrix.

[0011] Furthermore, the method for obtaining the carrier attitude data and high-precision navigation position includes: The compact combination observation residuals are constructed based on the double-difference residuals of each satellite. The Kalman gain matrix is ​​calculated based on the observation weights. The three-dimensional attitude error cumulative state is used as the prior estimate of the attitude error sub-vector of the current epoch, and the approximate position of the receiver is used as the prior estimate of the position error sub-vector. The attitude error subvector and position error subvector are updated by observation based on the Kalman gain matrix and the compact combination observation residuals to obtain the corrected attitude error subvector and position error subvector. The attitude quaternion is adjusted based on the corrected attitude error subvector to obtain the carrier attitude data. The approximate position of the receiver is compensated based on the corrected position error subvector to obtain the high-precision navigation position.

[0012] Furthermore, the method for obtaining the navigation position and pointing angle information includes: A northeast-sky coordinate system is constructed based on the high-precision navigation position. The difference between the preset target position and the high-precision navigation position is calculated. The target's geographic azimuth and geographic pitch angle are calculated based on the difference in the northeast-sky coordinates. The pointing angle deviation is calculated by combining the heading angle and pitch angle in the carrier's attitude data. If the number of satellite anomaly markers in the judgment result is less than or equal to one, the high-precision navigation position and pointing angle deviation are output and marked as high confidence. The positioning accuracy of the high-precision navigation position is greater than or equal to centimeters, and the pointing angle deviation accuracy of the high confidence is less than or equal to 0.1 degrees. If the number of satellite anomaly markers in the judgment result is greater than one, the carrier attitude data is retained, and based on the navigation position of the previous epoch, the navigation position and pointing angle deviation with decimeter-level positioning accuracy are calculated and output through INS position using IMU inertial data. The position error ellipse is extracted based on the prior covariance matrix of Kalman filtering, and the boundary is calculated using the two-dimensional covariance ellipse formula. The boundary of the position error ellipse is expanded by 1.5 times to obtain the navigation position safety boundary. The safety boundary is marked as low confidence, and the navigation position and pointing angle information of the preset carrier is obtained. The pointing angle deviation accuracy of the low confidence is less than or equal to 0.5 degrees.

[0013] A second aspect of the present invention provides a high-precision navigation and pointing system based on satellite cooperation, comprising: Data acquisition module: used to acquire dual-frequency carrier phase observations, pseudorange observations and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; Inflow hierarchy segmentation module: used to segment the inflow hierarchy of each satellite data based on the carrier phase double difference observation value and IMU inertial data through satellite cooperative geometric parameters, and obtain the main inflow term and the secondary inflow term; Three-dimensional attitude error accumulation state update module: used to predict and update the three-dimensional attitude error accumulation state of the preset carrier based on the main inflow term and the secondary inflow term through a nonlinear Muskingan model; Observation weight correction module: used to perform anomaly detection based on the pseudorange observation value and dual-frequency carrier phase observation value, make joint decisions based on the detection results, and correct the attitude calculation weights of each satellite based on the decision results to obtain the observation weights; Attitude position calculation module: used to calculate the attitude position based on the three-dimensional attitude error accumulation state and observation weights through Kalman filtering, to obtain the carrier attitude data and high-precision navigation position; Navigation pointing module: used to perform pointing calculation based on the carrier attitude data and preset target position, and determine the navigation state confidence through joint decision results to obtain the navigation position and pointing angle information of the preset carrier.

[0014] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: This invention divides satellite data into primary and secondary inflows, and processes observation data of different quality differently. It retains the dominant role of high-geometric-precision satellites, nonlinearly saturates the secondary inflows to suppress the abnormal influence of low-quality data, and predicts and updates the cumulative state of three-dimensional attitude error using a nonlinear Muskingan model. By utilizing the linear superposition of the primary inflow and the hyperbolic tangent saturation characteristics of the secondary inflow, it balances the contributions of dominant and redundant information, improving the adaptability of carrier attitude prediction in complex environments. Anomaly detection is performed based on the Melbourne-Wübbena combination to correct the observation noise covariance, and abnormal satellites are softly isolated to maintain the integrity of the observation structure. Attitude and position are tightly combined and solved using Kalman filtering. The Muskingan predicted attitude is used as a priori estimate for observation updates to obtain navigation position and high-precision attitude. When anomalies are detected, the decimeter-level degradation output is achieved through INS position estimation, enabling the system to provide continuous and reliable navigation pointing information when GNSS fails. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the steps of a high-precision navigation and pointing method based on satellite cooperation in an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0017] Reference Figure 1 As shown, this invention provides a high-precision navigation pointing method based on satellite cooperation, comprising: Acquire dual-frequency carrier phase observations, pseudorange observations, and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; In the actual assessment, the pre-set carrier was a quadcopter agricultural drone, equipped with a dual-frequency, dual-antenna GNSS receiver with a baseline length of 1.2 meters, installed along the nose direction. The MEMS-IMU sampling rate was 200Hz. The pre-set target position was 10m east, 5m north, and -3m above the landing point in the ENU coordinate system. There was a risk of obstruction from low-rise buildings. Under GPS L1 / L2 observations, 8 satellites were visible at the current epoch. The dual-frequency carrier phase observations were obtained from the PRN3 satellite, and the first carrier of antenna 1... The phase observation value L11 is 1234567.892m, which is the equivalent distance value of the L1 band. L21 is 963852.741m. The third carrier phase observation value L12 of antenna 2 is 1234568.245m, and the fourth carrier phase observation value L22 is 963853.102m. The first pseudorange observation value is 20230000.5m, and the second pseudorange observation value is 20230005.2m. The triaxial acceleration in the IMU inertial data is [0.2, -0.1, 9.8] m / s². 2 The three-axis angular velocity is [0.05, -0.03, 0.02] rad / s, the inter-station single difference of the carrier phase observation in the L1 band is 0.353m, the elevation angle of PRN1 is the highest 78 degrees, and PRN1 is taken as the reference satellite. Then the carrier phase double difference observation value of antenna 1 in the inter-satellite double difference is 0.125m, and that of antenna 2 is 0.098m. Based on the carrier phase double-difference observations and IMU inertial data, the inflow levels of each satellite data are divided using satellite cooperative geometric parameters to obtain the main inflow term and the secondary inflow term; In the actual assessment, based on the previous epoch, the corresponding heading angle is approximately 23°, pitch angle -2°, and roll angle 1°. The GNSS epoch interval is 0.05 seconds. Using the attitude quaternions [0.98, 0.01, -0.02, 0.20] and IMU angular velocity from the previous epoch, extrapolated by 0.05 seconds at 200Hz, the predicted attitude angles for the current epoch are obtained as: heading 23.1°, pitch -2.1°, and roll 0.9°. A rotation matrix is ​​constructed, and the baseline vector of the carrier coordinate system is set to b = [1.2, 0, 0]. TTransforming to the ECEF coordinate system, the single-point positioning accuracy of the approximate receiver position is approximately 2.5m. At longitude 116.391°, latitude 39.904°, and elevation 50.2m, the geocentric and geofixed coordinates of the PRN3 satellite are calculated based on the approximate receiver position and broadcast ephemeris. The predicted geometric distance is calculated to be 20230015.8m. Subtracting the carrier phase double-difference observation from the predicted geometric distance, the double-difference residual of PRN3 is 0.012m, which includes noise and multipath error. The cooperative weights of each satellite are calculated by normalizing the product of the square of the elevation angle and the carrier-to-noise ratio. PRN3 has an elevation angle of 35°, a carrier-to-noise ratio of 38dB-Hz, and a GDOP contribution of 0.15. PRN5 has an elevation angle of 65°. With a carrier-to-noise ratio of 48 dB-Hz and a cooperative weight of 12.5, the normalized value is 0.087. The preset number is 4, which is less than the number of visible satellites (8). Therefore, the primary inflow satellites are the first 4 satellites with an elevation angle > 60°, namely PRN1, PRN5, PRN7, and PRN9, with residuals of [-0.005, 0.003, 0.008, -0.002] m respectively. The secondary inflow satellites are the remaining 4 satellites, namely PRN3, PRN11, PRN13, and PRN15, with residuals of [0.012, -0.015, 0.020, -0.018] m respectively. Including noise, the double-difference residuals of the primary and secondary inflow satellites are weighted and summed according to the cooperative weight. The primary inflow term is 0.004 m, and the secondary inflow term is 0.025 m. The three-dimensional attitude error accumulation state of the preset carrier is predicted and updated based on the main inflow term and the secondary inflow term using a nonlinear Muskingan model. In the actual evaluation, the previous epoch state was [0.01, 0.005, -0.008]. T rad, corresponding to roll error, pitch error and yaw error, the state transition matrix is ​​a 3×3 matrix, the control matrix is ​​a 3×3 matrix, the corresponding coefficient is 0.1, the time interval is 0.05s, the inflow allocation coefficient is 0.7, the value range is 0.6~0.8, the scale factor is 0.05m, the value range is 0.03~0.05, the nonlinear exponent is 2. The predicted updated secondary inflow term contains a large residual from the low-elevation angle satellite PRN3, which is suppressed to 0.215 by tanh to prevent abnormal errors from being directly injected into the attitude error estimation. The current epoch prior state is obtained as [0.012,0.007,-0.005]; Anomaly detection is performed based on the pseudorange observations and dual-frequency carrier phase observations. A joint decision is made based on the detection results. The attitude calculation weights of each satellite are then corrected based on the decision results to obtain the observation weights. In practical evaluation, the Melbourne-Wübbena combination is calculated based on dual-antenna dual-frequency carrier phase observations and pseudorange observations to obtain the MW combination difference between adjacent epochs for each satellite. The L value of antenna 1 of the PRN3 satellite is also considered. MW The current value is 1.245m, the previous epoch was 0.395m, and the combination difference is 0.85m; the L of antenna 2 MW The current noise level is 1.238m, compared to 0.402m in the previous epoch. The combined noise difference is 0.836m, exceeding the normal noise range. The standard deviation of the dual-antenna MW combination within the sliding window is corrected using satellite GDOP weighting. The sliding window is 20 epochs, with a standard deviation of 0.03. The current GDOP is 2.5. Preset GDOP weighting coefficients are obtained based on the different satellite geometric distributions corresponding to the GDOP. The GDOP weighting coefficient ranges from 0.1 to 0.3, with the current value being 0.1. An adaptive threshold of 0.1125m is calculated. Since the combined noise difference of the dual-antenna MW exceeds the adaptive threshold, PRN3 is marked as an anomalous satellite. The observation noise covariance sub-block of the anomalous satellite is corrected using a nonlinear leakage formula. The observation noise covariance of PRN3 is (0.003m). 2 Converted to 9×10 -6 m 2 The discharge coefficient was set to 1.0, and the nonlinear exponent, consistent with the prediction update process, was kept at 2, resulting in a corrected observation noise covariance sub-block of 5.1 × 10⁻⁶. -4 m 2 In the Kalman filter, the observation weight of PRN3 is reduced to 1 / 57 of that of normal satellites for soft noise isolation. Based on the three-dimensional attitude error accumulation state and observation weights, attitude position is calculated by Kalman filtering to obtain carrier attitude data and high-precision navigation position. In the actual evaluation, a compact combination observation residual vector is constructed based on the double-difference residuals of each satellite. The double-difference pseudorange residuals after removing the reference satellite are [0.15, -0.08, 0.22, 0.05, -0.12, 0.18, -0.03]. T m, the double-difference carrier phase residuals are [0.012, -0.005, 0.008, 0.003, -0.002, 0.015, -0.008]. T With a wavelength of 19cm, the observation residual vector dimension is 14×1 for 7 satellites × 2 types of observations. The design matrix is ​​constructed based on the satellite ephemeris position, the approximate receiver position (longitude 116.391°, latitude 39.904°), and the rotation matrix. The attitude error submatrix is ​​shown below. Based on the projection relationship between the dual-antenna baseline vector and the satellite line-of-sight vector, PRN3's... The position error submatrix is ​​[0, -0.7, 0.98] m / rad. The satellite line-of-sight direction vector is used, the ambiguity submatrix is ​​the identity matrix, the carrier phase observation pair has an ambiguity coefficient of 1, it inherits the posterior covariance of the previous epoch, and is expressed through the state transition matrix. Predict the current epoch. The process noise matrix is ​​among them. Based on the performance settings of the MEMS-IMU, the gyroscope bias instability is 0.01° / h, the accelerometer bias is 50μg, and Q is a diagonal matrix whose diagonal elements correspond to the process noise variance of each component in the state vector. The state vector includes 3D attitude error, 3D velocity error, 3D position error, 3D gyroscope bias, 3D accelerometer bias, and 7D double-difference integer ambiguity, for a total of 22 dimensions. Since the integer ambiguity is constant when no cycle slip occurs, the process noise is determined only by the cycle slip probability. In the high confidence mode, it is considered constant, so the process noise variance corresponding to the ambiguity subvector is 0. In the actual evaluation, the Kalman gain matrix is ​​calculated based on the observation weights. The corrected observation noise covariance matrix is ​​block-diagonal, and the pseudorange is independent of the carrier phase. The calculated gain matrix has a dimension of 22×14. The attitude error sub-vector and position error sub-vector are updated based on the Kalman gain matrix and the tightly combined observation residuals. The prior estimate of the attitude error sub-vector is set to directly use the output Muskingan state [0.012, 0.007, -0.005]. T The position error subvector is set to [0.5, 0.3, 0.8] based on the uncertainty of the receiver's approximate position. T m, the velocity error subvector is set to [0.1, 0.1, 0.15] based on the IMU integral drift. T m / s, double-difference integer ambiguity is a fixed value of [12,-5,8,3,-2,7,-4] from the previous epoch. T (Zhou), after the observation update, the corrected attitude error subvector is [0.003, 0.002, -0.001]. T The corrected position error subvector is [-0.015, 0.012, -0.008] rad. T m, the ambiguity subvector, after integer constraints, remains [12, -5, 8, 3, -2, 7, -4]. T No cycle slip occurred. The corrected state vector includes position, velocity, attitude error, IMU zero bias estimate, and ambiguity. The posterior covariance matrix is ​​calculated and stored based on the corrected state vector for prediction in the next epoch. In actual evaluation, the corrected attitude error subvector [0.003, 0.002, -0.001] is used. T Construct an antisymmetric matrix using rad, and then convert the rotation vector into a rotation matrix using the Rodriguez formula, where the rotation matrix is... ; To obtain an antisymmetric matrix, the attitude quaternion is adjusted based on the calculation results. The adjustment formula is as follows: ;,in The attitude quaternions output by the mechanical programming correspond to 23.1° heading, -2.1° pitch, and 0.9° roll. The attitude correction quaternion is derived from the rotation vector. The transformation yielded the carrier attitude data as follows: heading 23.08°, pitch -2.08°, and roll 0.92°. The attitude error was reduced from the prior 0.8° to the posterior 0.23°, based on the corrected position error subvector [-0.015, 0.012, -0.008]. T The approximate location of the compensation receiver, and the high-precision navigation location after compensation: longitude 116.3909998°, latitude 39.9040001°, elevation 50.192m, horizontal accuracy ±0.02m, and elevation accuracy ±0.03m. The pointing calculation is performed based on the carrier attitude data and the preset target position, and the navigation state confidence is determined by the joint decision result to obtain the navigation position and pointing angle information of the preset carrier.

[0018] In the actual evaluation, a Northeast-Sky ENU coordinate system was constructed based on the high-precision navigation position. The Northeast-Sky coordinate difference between the preset target (10m East, 5m North, -3m Sky) and the carrier was calculated. The Northeast-Sky coordinate difference is [9.85, 4.92, -2.95]m. Based on the Northeast-Sky coordinate difference, the target's geographic azimuth and geographic pitch angles were calculated. The target's geographic azimuth angle is 63.5°, and the target's geographic pitch angle is -15.2°. Combining the heading and pitch angles from the carrier's attitude data, the pointing angle deviation was calculated. The heading deviation is 63.5° - 23.08°. =40.42°, then the nose needs to turn 40.4° to the right to align with the target. The pitch deviation is -15.2° - (-2.08°) = -13.12°, and it needs to pitch down 13.1°. PRN3 has triggered anomaly detection during the current navigation process. The number of satellite anomaly markers is less than or equal to one. The high-precision navigation position is output as longitude 116.3909998°, latitude 39.9040001°, elevation 50.192m, horizontal accuracy ±0.02m, elevation accuracy ±0.03m, and pointing angle deviation, and is marked as high confidence. In the actual evaluation, PRN3 and PRN5 triggered anomaly detection during another navigation process, resulting in nonlinear leakage. This led to the output of a navigation position in low-confidence mode. Based on the previous epoch, the decimeter-level navigation position was calculated using INS position estimation: longitude 116.390995°, latitude 39.904003°, and elevation 50.25m. Compared with the high-precision navigation position, the horizontal error was 0.28m, the elevation error was 0.058m, and the corresponding pointing angles were 23.12° heading, -2.05° pitch, and 0.95° roll. The position error ellipse was extracted based on the prior covariance matrix of Kalman filtering. The semi-major axis of 0.8m and the semi-minor axis of 0.4m were enlarged by a factor of 1.5 and used as a safety boundary, which was then marked as low confidence. The agricultural drone reduced its descent speed and waited for the GNSS signal to recover or switched to visual-assisted landing.

[0019] In this embodiment, the method for obtaining the carrier phase double-difference observation value includes: A dual-antenna GNSS receiver based on a pre-defined carrier receives dual-frequency carrier phase observations and pseudorange observations from each cooperating satellite, measures the carrier-to-noise ratio of each satellite signal, and performs inter-antenna differential analysis on the dual-frequency carrier phase observations based on the baseline length and baseline direction of the dual antennas to obtain single-difference carrier phase observations. The elevation angles of each visible satellite are calculated based on the satellite ephemeris and the approximate location of the receiver, and the GDOP contribution of each satellite is calculated. The satellites are sorted in descending order based on the elevation angles. If the elevation angles are the same, they are sorted in descending order based on the carrier-to-noise ratio (CNR). If the CNRs are the same, they are sorted in descending order based on the GDOP contribution. The satellite ranked first is taken as the reference satellite. Based on the reference satellite, the single-difference carrier phase observations are differentially analyzed between satellites to generate double-difference carrier phase observations. The baseline direction is determined according to the preset carrier and the installation position of the dual antennas. The visible satellites are determined according to the health indicators of the cooperating satellites.

[0020] In this embodiment, the feature is that it includes: The dual-frequency carrier phase observations include the first carrier phase observation and the second carrier phase observation of antenna 1, and the third carrier phase observation and the fourth carrier phase observation of antenna 2. The pseudorange observations include the first pseudorange observation and the second pseudorange observation of antenna 1, and the third pseudorange observation and the fourth pseudorange observation of antenna 2. The IMU inertial data includes the three-axis acceleration data and the three-axis angular velocity data of the preset carrier. The satellite cooperative geometric parameters include the elevation angle and the GDOP contribution.

[0021] In this embodiment, the method for obtaining the main inflow item and the secondary inflow item includes: Based on the carrier attitude quaternion from the previous epoch and the IMU inertial data, the attitude mechanical orchestration algorithm is used to extrapolate the predicted attitude angles for the current epoch. Based on the predicted attitude angles, a rotation matrix from the carrier coordinate system to the geocentric-ground-fixed coordinate system is constructed. The baseline vector in the carrier coordinate system is calculated based on the baseline length and direction of the dual antennas. The baseline vector is then transformed to the geocentric-ground-fixed coordinate system using the rotation matrix. The geocentric-ground-fixed coordinates of each antenna in the dual antennas are calculated based on the approximate location of the receiver. The geocentric-ground-fixed coordinates of each visible satellite are calculated based on the satellite ephemeris. The geometric distance from each antenna to each satellite is calculated. The inter-station single difference and inter-satellite double difference are calculated for the geometric distances to obtain the predicted geometric distances. The attitude angles include heading angle, pitch angle, and roll angle. The carrier phase double-difference observation, elevation angle, and carrier-to-noise ratio of the preset carrier at the current epoch are obtained. The carrier phase double-difference observation is subtracted from the geometric distance prediction to obtain the double-difference residual of each satellite. The cooperative weight of each satellite is calculated by normalizing the product of the square of the elevation angle and the carrier-to-noise ratio. The satellites are sorted according to the cooperative weight and divided into inflow levels according to a preset number to obtain the primary inflow satellites and secondary inflow satellites. The double-difference residuals of the primary inflow satellites and secondary inflow satellites are weighted and summed according to the cooperative weight to obtain the primary inflow term and secondary inflow term of the current epoch. The preset number is greater than or equal to 3 and less than the total number of visible satellites.

[0022] In this embodiment, the method for obtaining the three-dimensional attitude error accumulation state includes: The update formula for the nonlinear Muskingan model is: ; in This represents the cumulative state of the three-dimensional attitude error at the current epoch, where is the three-dimensional rotation vector, and represents the preset heading angle error, pitch angle error, and roll angle error of the carrier. Here is the state transition matrix, and here is a diagonal matrix. This represents the accumulated state of the three-dimensional attitude error from the previous epoch. For the control matrix, Assign coefficients to inflow terms. The main inflow term for the current epoch. It is a hyperbolic tangent nonlinear saturation function. For the next inflow term in the current epoch, As a scale factor, It is a non-linear exponent; The three-dimensional attitude error is the rotational vector attitude error between the actual carrier attitude and the reference attitude in the current epoch, and the reference attitude is the state prediction value in the current epoch.

[0023] In this embodiment, the method for obtaining the observation weights includes: The Melbourne-Wübbena combination is calculated based on dual-antenna dual-frequency carrier phase observations and pseudorange observations. The MW combination difference between adjacent epochs of each satellite is obtained. The standard deviation of the dual-antenna MW combination within the sliding window is corrected by weighting the satellite GDOP to obtain an adaptive threshold. If the MW combination difference of each dual antenna for the same satellite exceeds the adaptive threshold, the corresponding satellite is marked as an anomaly. The adaptive threshold is the product of the standard deviation and the GDOP weighting coefficient. For anomalous satellites, the observation noise covariance sub-block is corrected using a nonlinear leakage formula, which is: ; in To correct the observation noise covariance sub-block, for For the first The observation noise covariance sub-block of each satellite The preset discharge coefficient is a positive real number. For the first The MW combination difference of adjacent epochs of satellites, An adaptive threshold; Based on the modified observation noise covariance sub-blocks of each satellite, a block-diagonal global observation noise covariance matrix is ​​constructed. The global observation noise covariance matrix is ​​inverted to obtain the attitude calculation weight matrix. The diagonal elements of the inverse matrix corresponding to the modified observation noise covariance sub-block are used as observation weights according to the attitude calculation weight matrix.

[0024] In this embodiment, the method for obtaining the carrier attitude data and high-precision navigation position includes: The compact combination observation residuals are constructed based on the double-difference residuals of each satellite. The Kalman gain matrix is ​​calculated based on the observation weights. The three-dimensional attitude error cumulative state is used as the prior estimate of the attitude error sub-vector of the current epoch, and the approximate position of the receiver is used as the prior estimate of the position error sub-vector. The attitude error subvector and position error subvector are updated by observation based on the Kalman gain matrix and the compact combination observation residuals to obtain the corrected attitude error subvector and position error subvector. The attitude quaternion is adjusted based on the corrected attitude error subvector to obtain the carrier attitude data. The approximate position of the receiver is compensated based on the corrected position error subvector to obtain the high-precision navigation position.

[0025] In this embodiment, the method for obtaining the navigation position and pointing angle information includes: A northeast-sky coordinate system is constructed based on the high-precision navigation position. The difference between the preset target position and the high-precision navigation position is calculated. The target's geographic azimuth and geographic pitch angle are calculated based on the difference in the northeast-sky coordinates. The pointing angle deviation is calculated by combining the heading angle and pitch angle in the carrier's attitude data. If the number of satellite anomaly markers in the judgment result is less than or equal to one, the high-precision navigation position and pointing angle deviation are output and marked as high confidence. The positioning accuracy of the high-precision navigation position is greater than or equal to centimeters, and the pointing angle deviation accuracy of the high confidence is less than or equal to 0.1 degrees. If the number of satellite anomaly markers in the judgment result is greater than one, the carrier attitude data is retained, and based on the navigation position of the previous epoch, the navigation position and pointing angle deviation with decimeter-level positioning accuracy are calculated and output through INS position using IMU inertial data. The position error ellipse is extracted based on the prior covariance matrix of Kalman filtering, and the boundary is calculated using the two-dimensional covariance ellipse formula. The boundary of the position error ellipse is expanded by 1.5 times to obtain the navigation position safety boundary. The safety boundary is marked as low confidence, and the navigation position and pointing angle information of the preset carrier is obtained. The pointing angle deviation accuracy of the low confidence is less than or equal to 0.5 degrees.

[0026] A second aspect of the present invention also provides a high-precision navigation and pointing system based on satellite cooperation, comprising: Data acquisition module: used to acquire dual-frequency carrier phase observations, pseudorange observations and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; Inflow hierarchy segmentation module: used to segment the inflow hierarchy of each satellite data based on the carrier phase double difference observation value and IMU inertial data through satellite cooperative geometric parameters, and obtain the main inflow term and the secondary inflow term; Three-dimensional attitude error accumulation state update module: used to predict and update the three-dimensional attitude error accumulation state of the preset carrier based on the main inflow term and the secondary inflow term through a nonlinear Muskingan model; Observation weight correction module: used to perform anomaly detection based on the pseudorange observation value and dual-frequency carrier phase observation value, make joint decisions based on the detection results, and correct the attitude calculation weights of each satellite based on the decision results to obtain the observation weights; Attitude position calculation module: used to calculate the attitude position based on the three-dimensional attitude error accumulation state and observation weights through Kalman filtering, to obtain the carrier attitude data and high-precision navigation position; Navigation pointing module: used to perform pointing calculation based on the carrier attitude data and preset target position, and determine the navigation state confidence through joint decision results to obtain the navigation position and pointing angle information of the preset carrier.

[0027] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

Claims

1. A high-precision navigation and pointing method based on satellite cooperation, characterized in that, Includes the following steps: Acquire dual-frequency carrier phase observations, pseudorange observations, and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; Based on the carrier phase double-difference observations and IMU inertial data, the inflow levels of each satellite data are divided using satellite cooperative geometric parameters to obtain the main inflow term and the secondary inflow term; The three-dimensional attitude error accumulation state of the preset carrier is predicted and updated based on the main inflow term and the secondary inflow term using a nonlinear Muskingan model. Anomaly detection is performed based on the pseudorange observations and dual-frequency carrier phase observations. A joint decision is made based on the detection results. The attitude calculation weights of each satellite are then corrected based on the decision results to obtain the observation weights. Based on the three-dimensional attitude error accumulation state and observation weights, attitude position is calculated by Kalman filtering to obtain carrier attitude data and high-precision navigation position. Based on the carrier attitude data and the preset target position, a pointing calculation is performed, and the navigation state confidence is determined through a joint decision result to obtain the navigation position and pointing angle information of the preset carrier, including: A northeast-sky coordinate system is constructed based on the high-precision navigation position. The difference between the preset target position and the high-precision navigation position is calculated. The target's geographic azimuth and geographic pitch angle are calculated based on the difference in the northeast-sky coordinates. The pointing angle deviation is calculated by combining the heading angle and pitch angle in the carrier's attitude data. If the number of satellite anomaly markers in the judgment result is less than or equal to one, the high-precision navigation position and pointing angle deviation are output and marked as high confidence. The positioning accuracy of the high-precision navigation position is greater than or equal to centimeters, and the pointing angle deviation accuracy of the high confidence is less than or equal to 0.1 degrees. If the number of satellite anomaly markers in the judgment result is greater than one, the carrier attitude data is retained, and based on the navigation position of the previous epoch, the navigation position and pointing angle deviation with decimeter-level positioning accuracy are calculated and output through INS position using IMU inertial data. The position error ellipse is extracted based on the prior covariance matrix of Kalman filtering, and the boundary is calculated using the two-dimensional covariance ellipse formula. The boundary of the position error ellipse is expanded by 1.5 times to obtain the navigation position safety boundary. The safety boundary is marked as low confidence, and the navigation position and pointing angle information of the preset carrier is obtained. The pointing angle deviation accuracy of the low confidence is less than or equal to 0.5 degrees.

2. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, The method for obtaining the carrier phase double-difference observations includes: A dual-antenna GNSS receiver based on a pre-defined carrier receives dual-frequency carrier phase observations and pseudorange observations from each cooperating satellite, measures the carrier-to-noise ratio of each satellite signal, and performs inter-antenna differential analysis on the dual-frequency carrier phase observations based on the baseline length and baseline direction of the dual antennas to obtain single-difference carrier phase observations. The elevation angles of each visible satellite are calculated based on the satellite ephemeris and the approximate location of the receiver, and the GDOP contribution of each satellite is calculated. The satellites are sorted in descending order based on the elevation angles. If the elevation angles are the same, they are sorted in descending order based on the carrier-to-noise ratio (CNR). If the CNRs are the same, they are sorted in descending order based on the GDOP contribution. The satellite ranked first is taken as the reference satellite. Based on the reference satellite, the single-difference carrier phase observations are differentially analyzed between satellites to generate double-difference carrier phase observations. The baseline direction is determined according to the preset carrier and the installation position of the dual antennas. The visible satellites are determined according to the health indicators of the cooperating satellites.

3. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, include: The dual-frequency carrier phase observations include the first carrier phase observation and the second carrier phase observation of antenna 1, and the third carrier phase observation and the fourth carrier phase observation of antenna 2. The pseudorange observations include the first pseudorange observation and the second pseudorange observation of antenna 1, and the third pseudorange observation and the fourth pseudorange observation of antenna 2. The IMU inertial data includes the three-axis acceleration data and the three-axis angular velocity data of the preset carrier. The satellite cooperative geometric parameters include the elevation angle and the GDOP contribution.

4. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, The method for obtaining the main inflow item and the secondary inflow item includes: Based on the carrier attitude quaternion from the previous epoch and the IMU inertial data, the attitude mechanical orchestration algorithm is used to extrapolate the predicted attitude angles for the current epoch. Based on the predicted attitude angles, a rotation matrix from the carrier coordinate system to the geocentric-ground-fixed coordinate system is constructed. The baseline vector in the carrier coordinate system is calculated based on the baseline length and direction of the dual antennas. The baseline vector is then transformed to the geocentric-ground-fixed coordinate system using the rotation matrix. The geocentric-ground-fixed coordinates of each antenna in the dual antennas are calculated based on the approximate location of the receiver. The geocentric-ground-fixed coordinates of each visible satellite are calculated based on the satellite ephemeris. The geometric distance from each antenna to each satellite is calculated. The inter-station single difference and inter-satellite double difference are calculated for the geometric distances to obtain the predicted geometric distances. The attitude angles include heading angle, pitch angle, and roll angle. The carrier phase double-difference observation, elevation angle, and carrier-to-noise ratio of the preset carrier at the current epoch are obtained. The carrier phase double-difference observation is subtracted from the geometric distance prediction to obtain the double-difference residual of each satellite. The cooperative weight of each satellite is calculated by normalizing the product of the square of the elevation angle and the carrier-to-noise ratio. The satellites are sorted according to the cooperative weight and divided into inflow levels according to a preset number to obtain the primary inflow satellites and secondary inflow satellites. The double-difference residuals of the primary inflow satellites and secondary inflow satellites are weighted and summed according to the cooperative weight to obtain the primary inflow term and secondary inflow term of the current epoch. The preset number is greater than or equal to 3 and less than the total number of visible satellites.

5. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, The method for obtaining the three-dimensional attitude error cumulative state includes: The update formula for the nonlinear Muskingan model is: ； in This represents the cumulative state of the three-dimensional attitude error in the current epoch. It is a three-dimensional rotation vector. To preset the heading angle error, pitch angle error, and roll angle error of the carrier, Here is the state transition matrix. It is a diagonal matrix. This represents the accumulated state of the three-dimensional attitude error from the previous epoch. For the control matrix, Assign coefficients to inflow terms. The main inflow term for the current epoch. It is a hyperbolic tangent nonlinear saturation function. For the next inflow term in the current epoch, As a scale factor, It is a non-linear exponent; The three-dimensional attitude error is the rotational vector attitude error between the actual carrier attitude and the reference attitude in the current epoch, and the reference attitude is the state prediction value in the current epoch.

6. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, The method for obtaining the observation weights includes: The Melbourne-Wübbena combination is calculated based on dual-antenna dual-frequency carrier phase observations and pseudorange observations. The MW combination difference between adjacent epochs of each satellite is obtained. The standard deviation of the dual-antenna MW combination within the sliding window is corrected by weighting the satellite GDOP to obtain an adaptive threshold. If the MW combination difference of each dual antenna for the same satellite exceeds the adaptive threshold, the corresponding satellite is marked as an anomaly. The adaptive threshold is the product of the standard deviation and the GDOP weighting coefficient. For anomalous satellites, the observation noise covariance sub-block is corrected using a nonlinear leakage formula, which is: ； in To correct the observation noise covariance sub-block, For the first The observation noise covariance sub-block of each satellite To preset the discharge coefficient, It is a positive real number. For the first The MW combination difference of adjacent epochs of satellites, For adaptive threshold, It is a non-linear exponent; Based on the modified observation noise covariance sub-blocks of each satellite, a block-diagonal global observation noise covariance matrix is ​​constructed. The global observation noise covariance matrix is ​​inverted to obtain the attitude calculation weight matrix. The diagonal elements of the inverse matrix corresponding to the modified observation noise covariance sub-block are used as observation weights according to the attitude calculation weight matrix.

7. The high-precision navigation and pointing method based on satellite cooperation according to claim 1, characterized in that, The method for obtaining the carrier attitude data and high-precision navigation position includes: The compact combination observation residuals are constructed based on the double-difference residuals of each satellite. The Kalman gain matrix is ​​calculated based on the observation weights. The three-dimensional attitude error cumulative state is used as the prior estimate of the attitude error sub-vector of the current epoch, and the approximate position of the receiver is used as the prior estimate of the position error sub-vector. The attitude error subvector and position error subvector are updated by observation based on the Kalman gain matrix and the compact combination observation residuals to obtain the corrected attitude error subvector and position error subvector. The attitude quaternion is adjusted based on the corrected attitude error subvector to obtain the carrier attitude data. The approximate position of the receiver is compensated based on the corrected position error subvector to obtain the high-precision navigation position.

8. A high-precision navigation and pointing system based on satellite cooperation, used to execute the high-precision navigation and pointing method based on satellite cooperation as described in any one of claims 1 to 7, characterized in that, The system includes: Data acquisition module: used to acquire dual-frequency carrier phase observations, pseudorange observations and IMU inertial data of a preset carrier, and generate carrier phase double-difference observations based on the dual-frequency carrier phase observations; Inflow hierarchy segmentation module: used to segment the inflow hierarchy of each satellite data based on the carrier phase double difference observation value and IMU inertial data through satellite cooperative geometric parameters, and obtain the main inflow term and the secondary inflow term; Three-dimensional attitude error accumulation state update module: used to predict and update the three-dimensional attitude error accumulation state of the preset carrier based on the main inflow term and the secondary inflow term through a nonlinear Muskingan model; Observation weight correction module: used to perform anomaly detection based on the pseudorange observation value and dual-frequency carrier phase observation value, make joint decisions based on the detection results, and correct the attitude calculation weights of each satellite based on the decision results to obtain the observation weights; Attitude position calculation module: used to calculate the attitude position based on the three-dimensional attitude error accumulation state and observation weights through Kalman filtering, to obtain the carrier attitude data and high-precision navigation position; Navigation pointing module: used to perform pointing calculation based on the carrier attitude data and preset target position, and determine the navigation state confidence through joint decision results to obtain the navigation position and pointing angle information of the preset carrier.

Images