A Smart Land Acquisition and Relocation High-Precision Mapping and Positioning Method Based on BeiDou and UAV Fusion

Abstract

This invention relates to the field of satellite navigation and positioning technology, and discloses a smart land acquisition and resettlement high-precision mapping and positioning method based on the fusion of Beidou and UAV. The method includes: acquiring satellite observation data and UAV motion parameters; establishing a carrier phase double-difference observation equation and calculating a position constraint model based on the motion parameters; using the phase deviation value generated by multipath error modeling to perform reverse compensation on the carrier phase double-difference observation equation to maintain the full-rank characteristic of the variance-covariance matrix; and using the position constraint model in the search space to eliminate candidate integer ambiguity combinations that exceed displacement constraints, and outputting the mapping coordinates. This invention solves the matrix ill-conditioning problem caused by the scarcity of observation values ​​under strong reflection environments through the synergy of residual reverse compensation and physical motion envelope constraints, ensuring the robustness of the positioning benchmark and the topological consistency of the mapping model under complex occlusion conditions.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation and positioning technology, specifically to a smart land acquisition and resettlement high-precision mapping and positioning method based on the integration of BeiDou and UAVs. Background Technology

[0002] The current BeiDou Navigation Satellite System, combined with airborne carrier phase differential technology, can provide centimeter-level positioning references for UAV mapping tasks. By constructing a double-difference observation model, it can quickly fix integer ambiguities. Under normal open conditions, the system mainly deals with systematic errors caused by tropospheric delay or ionospheric delay to ensure the convergence of the observation equations.

[0003] However, in situations involving large amounts of building debris and exposed metal frames, such as smart land acquisition and resettlement, the presence of irregular reflective surfaces leads to multipath bias in the signal. Existing solutions typically use signal-to-noise ratio or satellite elevation angle as weighting criteria to deweight or eliminate interfered satellites. This defensive processing logic targeting interfering satellites results in a bottleneck in the solution system in heavily obscured areas due to insufficient effective observations, leading to rank starvation in the observation equations and causing the covariance matrix to become ill-conditioned. In addition to limitations in observation geometry, there are also shortcomings at the control algorithm level. For example, Chinese invention patent application CN121432493A discloses a collaborative positioning method for BeiDou / GNSS+5G UAVs based on multimodal neural networks, which utilizes 5G signals and visual features through a neural network. Nonlinear modeling-assisted positioning; industry attempts have focused on mitigating signal obstruction by increasing the number of receiving satellite constellations or improving receiver sensitivity. However, such linear improvement approaches cannot eliminate the deterministic phase delay introduced by the reflection path. This strategy, which relies solely on eliminating damaged observations, creates an irreconcilable contradiction between maintaining accuracy and preserving the system's spatial geometric strength. Specifically, existing technologies suffer from the following shortcomings: 1. Weight adjustment logic can only suppress random noise and cannot correct the deterministic phase offset caused by signal reflection; 2. Eliminating damaged satellites causes the geometric strength of the solution system to degrade in the confined space, leading to distortion of the covariance matrix; 3. The ambiguity search algorithm lacks deep coupling with physical motion constraints, resulting in slow reconvergence after signal loss.

[0004] Therefore, the technical problem to be solved by this invention is how to reconstruct carrier phase observations using motion data collected by airborne inertial measurement units and achieve reverse cancellation of multipath residuals without sacrificing spatial geometric strength, thereby maintaining the convergence of the ambiguity resolution space. Summary of the Invention

[0005] This invention proposes a smart land acquisition and relocation high-precision mapping and positioning method based on the fusion of BeiDou and UAVs, comprising the following steps: Step S1: Obtain raw satellite observation data including BeiDou satellite carrier phase observations, pseudorange observations, and Doppler frequency shift, and obtain three-dimensional motion parameters including the real-time velocity and acceleration of the UAV. Step S2: Construct carrier phase double-difference observation equations based on the original satellite observation data, and calculate the maximum physical displacement threshold between adjacent epochs based on the three-dimensional motion parameters to generate the position constraint model of the carrier phase double-difference observation equations. Step S3: The multipath reflection signal is modeled using signal power characteristics to generate phase delay deviation value, and the phase delay deviation value is substituted as a reverse compensation term into the variance-covariance matrix of the carrier phase double difference observation equation to maintain the full rank characteristic of the variance-covariance matrix without removing the observation values ​​of the interfered satellite. Step S4: Using the position constraint model as the search boundary, within the ambiguity space determined by the compensated variance-covariance matrix, candidate integer ambiguities exceeding the maximum physical displacement threshold are eliminated, and dimensionality reduction search is performed on the remaining candidate solutions to obtain fixed solutions for integer ambiguities. Step S5: Calculate and output centimeter-level mapping coordinates based on the integer ambiguity fixed solution.

[0006] Preferably, the step of generating the phase delay deviation value in step S3 includes: monitoring the signal power characteristic values ​​of the in-phase branch and the quadrature branch in the carrier tracking loop; calculating the amplitude ratio between the signal power characteristic values; performing a function mapping between the carrier phase observation value and the amplitude ratio to separate the phase difference between the direct wave signal and the reflected wave signal, so as to generate the phase delay deviation value.

[0007] Preferably, the step of compensating the variance-covariance matrix in step S3 includes: converting the phase delay deviation value into an equivalent variance correction amount of the phase observation value of the corresponding satellite channel; adjusting the diagonal variance elements of the corresponding satellite channel in the variance-covariance matrix based on the equivalent variance correction amount; and reconstructing the ambiguity decorrelation transformation matrix using the adjusted variance-covariance matrix.

[0008] Preferably, step S4, which involves eliminating candidate integer ambiguities, includes: determining the maximum displacement envelope interval of the current epoch relative to the previous epoch based on the preset flight path envelope and dynamic acceleration limit in the three-dimensional motion parameters; converting each candidate integer ambiguity to the spatial position domain to generate candidate displacement increments; and determining that the corresponding candidate integer ambiguity is invalid and performing search branch truncation if the candidate displacement increment is outside the maximum displacement envelope interval.

[0009] Preferably, the maximum physical displacement threshold ΔL satisfies the following determination logic: Where ΔL is the theoretical displacement magnitude of the UAV between adjacent epochs, v is the real-time velocity in the three-dimensional motion parameters, a is the acceleration in the three-dimensional motion parameters, and Δt is the data sampling time interval.

[0010] Preferably, the method further includes the following steps: Step S61, real-time monitoring of the carrier phase jump status in the original satellite observation data; Step S62, when a cycle slip is determined to occur, using inertial navigation data to correct the Doppler frequency shift in the original satellite observation data to compensate for the phase center angular velocity deviation of the UAV antenna; Step S63, using the corrected Doppler frequency shift to predict the carrier phase increment of the current epoch, and performing phase value bridging on the satellite channel where the cycle slip has occurred through the carrier phase increment.

[0011] Preferably, the method further includes the following steps: Step S71, calculating the fluctuation frequency of the phase delay deviation value on the time axis; Step S72, when the fluctuation frequency exceeds the preset jitter threshold, reducing the weight factor of the corresponding satellite channel in the variance-covariance matrix to suppress the intrusion of high-frequency multipath interference on the accuracy of centimeter-level mapping coordinates.

[0012] Preferably, the method further includes the following steps: Step S81, while acquiring centimeter-level mapping coordinates, recording the original three-dimensional point cloud data collected by the UAV payload; Step S82, using the posterior variance factor of the integer ambiguity fixed solution to perform spatial reference alignment on the original three-dimensional point cloud data, so as to correct the local model distortion caused by signal occlusion.

[0013] Preferably, the step of maintaining the full-rank characteristic of the variance-covariance matrix in step S3 includes: eliminating residuals by substituting the phase delay deviation value into the carrier phase double-difference observation equation, so that the multipath damaged signal returns to the equivalent direct wave observation state, thereby ensuring that the number of independent double-difference observation equations involved in the solution is not less than the minimum rank requirement required for integer ambiguity calculation.

[0014] Preferably, step S5, which outputs centimeter-level mapping coordinates, includes: converting the solution coordinates corresponding to the integer ambiguity fixed solution to the projected coordinate system preset by the task; if the calculated standard deviation of the positioning accuracy is less than 30mm, then marking the current centimeter-level mapping coordinates as valid results and storing them in the spatial database.

[0015] The beneficial effects of this invention are: 1. In high-precision mapping and positioning, a carrier phase deterministic reconstruction mechanism based on inertial anchoring is adopted. This changes the conventional approach of maintaining model purity by reducing weights or removing damaged satellites when facing multipath interference. By extracting the signal-to-noise ratio change characteristics in real time and combining the strapdown integral displacement increment output by the inertial measurement unit, the multipath signal, which is a random deviation, is converted into a quantifiable phase delay residual and then canceled in reverse. This processing method allows the contaminated satellite channel to return to the equivalent direct wave state while maintaining full observation weights, maintaining the full-rank characteristic of the double-difference observation equation system, avoiding ill-conditioned covariance matrix caused by the scarcity of effective observations, and ensuring the robustness of the positioning reference in complex obstruction environments.

[0016] 2. By deeply coupling the physical kinematic envelope of the UAV with the ambiguity dimensionality reduction search algorithm, a deterministic mapping between the three-dimensional spatial limit displacement and the theoretical ambiguity boundary is established. The search space is limited by the maximum physical maneuverability of the UAV between epochs, and invalid candidate branches that exceed the displacement sphere envelope are truncated, reducing the search radius of the algorithm. This physical constraint mechanism forms a closed-loop collaboration with the compensated and corrected carrier phase observation value. Even in strong reflection environments such as building shadow areas or metal frames, it still maintains rapid convergence of the integer ambiguity solution space, eliminating the technical defect of frequent jumps between meter-level floating-point solutions and centimeter-level fixed solutions in the positioning results.

[0017] 3. By introducing cycle slip transient bridging logic based on Doppler prediction and a dynamic adjustment strategy for carrier tracking loop bandwidth, precise compensation for phase increments during signal loss of lock is achieved. The Doppler frequency shift is corrected based on the antenna phase center angular velocity offset calculated from inertial data. This maintains the logical continuity of carrier phase in the event of short-term signal interruption. This cross-dimensional information coordination mechanism, combined with real-time suppression of tracking loop bandwidth, effectively isolates the intrusion of high-frequency multipath jitter into carrier phase observations, reduces the frequency of cold start initialization triggered by the system under heavy interference conditions, and ensures the topological consistency of the 3D point cloud model benchmark during dynamic mapping. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the high-precision mapping and positioning calculation process of the BeiDou and UAV fusion invention. Figure 2 This is a diagram showing the precise line-of-sight geometric displacement increment solution of the inertial navigation and ephemeris data of this invention. Detailed Implementation

[0020] 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.

[0021] A high-precision mapping and positioning method for intelligent land acquisition and relocation based on the fusion of BeiDou and UAVs includes the following steps: Step S1: Obtain raw satellite observation data including BeiDou satellite carrier phase observations, pseudorange observations, and Doppler frequency shift, and obtain three-dimensional motion parameters including the real-time velocity and acceleration of the UAV. Step S2: Construct carrier phase double-difference observation equations based on the original satellite observation data, and calculate the maximum physical displacement threshold between adjacent epochs based on the three-dimensional motion parameters to generate the position constraint model of the carrier phase double-difference observation equations. Step S3: The multipath reflection signal is modeled using signal power characteristics to generate phase delay deviation value, and the phase delay deviation value is substituted as a reverse compensation term into the variance-covariance matrix of the carrier phase double difference observation equation to maintain the full rank characteristic of the variance-covariance matrix without removing the observation values ​​of the interfered satellite. Step S4: Using the position constraint model as the search boundary, within the ambiguity space determined by the compensated variance-covariance matrix, candidate integer ambiguities exceeding the maximum physical displacement threshold are eliminated, and dimensionality reduction search is performed on the remaining candidate solutions to obtain fixed solutions for integer ambiguities. Step S5: Calculate and output centimeter-level mapping coordinates based on the integer ambiguity fixed solution.

[0022] Preferably, the step of generating the phase delay deviation value in step S3 includes: monitoring the signal power characteristic values ​​of the in-phase branch and the quadrature branch in the carrier tracking loop; calculating the amplitude ratio between the signal power characteristic values; performing a function mapping between the carrier phase observation value and the amplitude ratio to separate the phase difference between the direct wave signal and the reflected wave signal, so as to generate the phase delay deviation value.

[0023] Preferably, the step of compensating the variance-covariance matrix in step S3 includes: converting the phase delay deviation value into an equivalent variance correction amount of the phase observation value of the corresponding satellite channel; adjusting the diagonal variance elements of the corresponding satellite channel in the variance-covariance matrix based on the equivalent variance correction amount; and reconstructing the ambiguity decorrelation transformation matrix using the adjusted variance-covariance matrix.

[0024] Preferably, step S4, which involves eliminating candidate integer ambiguities, includes: determining the maximum displacement envelope interval of the current epoch relative to the previous epoch based on the preset flight path envelope and dynamic acceleration limit in the three-dimensional motion parameters; converting each candidate integer ambiguity to the spatial position domain to generate candidate displacement increments; and determining that the corresponding candidate integer ambiguity is invalid and performing search branch truncation if the candidate displacement increment is outside the maximum displacement envelope interval.

[0025] Preferably, the maximum physical displacement threshold ΔL satisfies the following determination logic: Where ΔL is the theoretical displacement magnitude of the UAV between adjacent epochs, v is the real-time velocity in the three-dimensional motion parameters, a is the acceleration in the three-dimensional motion parameters, and Δt is the data sampling time interval.

[0026] Preferably, the method further includes the following steps: Step S61, real-time monitoring of the carrier phase jump status in the original satellite observation data; Step S62, when a cycle slip is determined to occur, using inertial navigation data to correct the Doppler frequency shift in the original satellite observation data to compensate for the phase center angular velocity deviation of the UAV antenna; Step S63, using the corrected Doppler frequency shift to predict the carrier phase increment of the current epoch, and performing phase value bridging on the satellite channel where the cycle slip has occurred through the carrier phase increment.

[0027] Preferably, the method further includes the following steps: Step S71, calculating the fluctuation frequency of the phase delay deviation value on the time axis; Step S72, when the fluctuation frequency exceeds the preset jitter threshold, reducing the weight factor of the corresponding satellite channel in the variance-covariance matrix to suppress the intrusion of high-frequency multipath interference on the accuracy of centimeter-level mapping coordinates.

[0028] Preferably, the method further includes the following steps: Step S81, while acquiring centimeter-level mapping coordinates, recording the original three-dimensional point cloud data collected by the UAV payload; Step S82, using the posterior variance factor of the integer ambiguity fixed solution to perform spatial reference alignment on the original three-dimensional point cloud data, so as to correct the local model distortion caused by signal occlusion.

[0029] Preferably, the step of maintaining the full-rank characteristic of the variance-covariance matrix in step S3 includes: eliminating residuals by substituting the phase delay deviation value into the carrier phase double-difference observation equation, so that the multipath damaged signal returns to the equivalent direct wave observation state, thereby ensuring that the number of independent double-difference observation equations involved in the solution is not less than the minimum rank requirement required for integer ambiguity calculation.

[0030] Preferably, step S5, which outputs centimeter-level mapping coordinates, includes: converting the solution coordinates corresponding to the integer ambiguity fixed solution to the projected coordinate system preset by the task; if the calculated standard deviation of the positioning accuracy is less than 30mm, then marking the current centimeter-level mapping coordinates as valid results and storing them in the spatial database.

[0031] Example 1: In a high-frequency dynamic aerial survey scenario for smart land acquisition and resettlement, when there are partially demolished reinforced concrete building structures and exposed metal skeletons on the ground, the onboard BeiDou receiver signal generates multipath effects and non-line-of-sight reception when the UAV flies close to this irregular, highly reflective surface. Multiple visible satellites suffer from multipath contamination. If the weighting is reduced or abnormal nodes are removed according to conventional logic, the number of effective independent double-difference observation equations participating in the solution is sharply reduced. The deterioration of the geometric spatial structure causes observation rank starvation, the variance-covariance matrix is ​​distorted and the search ellipsoid is distorted, resulting in the inability to fix the integer ambiguity and output a floating-point solution. The original satellite observation data containing BeiDou satellite carrier phase observations, pseudorange observations and Doppler frequency shifts are obtained, and the three-dimensional motion parameters containing the UAV's real-time velocity and acceleration are obtained. The signal-to-noise ratio within the set time window is extracted from the original satellite observation data. The rate of change is used to determine whether multipath fading exists in the corresponding satellite channel when the absolute value of the rate of change is greater than the threshold of 0.5 dB-Hz / s. For multipath fading satellite channels, the weighting is stopped and carrier phase compensation is triggered. The line-of-sight geometric vector of a single satellite is calculated by combining ephemeris data. The three-dimensional motion parameters are integrated along the line-of-sight geometric vector to calculate the line-of-sight geometric displacement increment of the UAV between adjacent epochs. The Doppler frequency shift of the current epoch is extracted to calculate the corresponding velocity integral displacement. If the deviation between the line-of-sight geometric displacement increment and the velocity integral displacement is less than the preset tolerance range, it is confirmed as the reference line-of-sight constraint increment. The carrier phase observation value is extracted to calculate the measured phase increment between adjacent epochs. The difference between the measured phase increment and the reference line-of-sight constraint increment is used to generate the phase delay deviation value. In step S3, the signal power characteristic value of the in-phase branch of the carrier tracking loop is monitored in real time. Power characteristics of orthogonal branch signals The amplitude ratio R between the two is calculated. The fifth-order polynomial fitting coefficient matrix, which is pre-stored in non-volatile memory, is called to convert the amplitude ratio R into a phase delay deviation value. The polynomial fitting coefficient matrix is ​​obtained by pre-calibrating multipath signals with different incident angles and signal-to-noise ratio gradients in an anechoic chamber environment. The input feature vector includes the carrier signal-to-noise ratio and the amplitude ratio R. The corresponding phase deviation compensation amount is output to remove the deterministic phase offset caused by irregular reflections, so that the contaminated satellite channel returns to the equivalent direct wave state under the full observation weight condition, maintains the full-rank characteristic of the double-difference observation equation system, and provides a stable geometric space constraint matrix for subsequent ambiguity resolution. The phase delay deviation value is substituted as a reverse compensation term into the variance-covariance matrix of the carrier phase double-difference observation equation.

[0032] To maintain the full-rank property of the variance-covariance matrix without removing interfered satellite observations, the maximum physical displacement threshold ΔL between adjacent epochs is calculated based on the three-dimensional motion parameters. This maximum physical displacement threshold satisfies the logical relationship... Where ΔL is the theoretical displacement magnitude, v is the real-time velocity, a is the acceleration, and Δt is the data sampling time interval, a position constraint model of the carrier phase double-difference observation equation is generated based on the maximum physical displacement threshold. In step S4, the attitude update quaternion is constructed using the three-axis angular velocity increments output in real time by the airborne inertial measurement unit, the three-dimensional acceleration vector in the body coordinate system is mapped to the local navigation coordinate system, the gravitational acceleration component is stripped and mapped twice to the geocentric-ground-fixed coordinate system, and the precise line-of-sight geometric displacement increment between adjacent epochs is calculated; according to the formula Determine the dynamic physical displacement tolerance threshold T, where T is the dynamic calculated displacement tolerance limit and k is the dimensionless confidence ratio coefficient. The standard deviation of the noise in the acceleration measurement of the inertial measurement unit. λ is the data sampling time interval, and λ is the carrier wavelength. The standard deviation of the receiver phase-locked loop Doppler frequency tracking error is defined as follows: if the absolute value of the difference between the candidate displacement increment generated by converting the candidate integer ambiguity to the spatial position domain and the precise line-of-sight geometric displacement increment is greater than the threshold T, then the candidate solution is deemed invalid and the search branch is truncated.

[0033] By utilizing inherent hardware noise parameters to define physical kinematic constraint boundaries, the system achieves objective data convergence for residual reconstruction in high-frequency dynamic environments. Using a position constraint model as the search boundary, within the ambiguity space defined by the compensated variance-covariance matrix, each candidate integer ambiguity is transformed into the spatial position domain to generate candidate displacement increments. If a candidate displacement increment is outside the maximum displacement envelope interval, the corresponding candidate integer ambiguity is deemed invalid, and combinations of candidate integer ambiguities exceeding the displacement constraints are eliminated. The remaining candidate solutions are then subjected to dimensionality reduction search to obtain fixed solutions for integer ambiguities. Based on the fixed solutions for integer ambiguities, centimeter-level mapping coordinates are calculated and output, and then transformed into a preset projection coordinate system. Random phase deviations exhibited by multipath interference are mapped to delay residuals and canceled out in the reverse direction. Contaminated satellite channels revert to the equivalent direct wave state while maintaining full observation weights, avoiding the risk of geometric matrix distortion caused by observation rank starvation. Carrier phase residual convergence is achieved without sacrificing the underlying spatial geometric strength. Mapping coordinates with a positioning accuracy standard deviation of less than 30 mm are stored in the spatial database to form the benchmark for mapping data.

[0034] Example 2: In an outdoor test field simulating high-frequency dynamic aerial surveying for intelligent land acquisition and resettlement, semi-demolished concrete components with exposed reinforcing steel are deployed on the ground to form an irregular interference array with controllable reflectivity. The test platform is equipped with an inertial measurement unit with a sampling frequency of 100Hz and a multi-frequency BeiDou satellite navigation system receiver. Gaussian white noise with a signal-to-noise ratio of 20dB is actively superimposed on the original carrier phase signal, along with a dynamic multipath reflection signal with Doppler frequency shift characteristics, to construct a degraded test condition. The setting of the signal-to-noise ratio change rate judgment threshold in the system is balanced. To assess the sensitivity of multipath fading feature acquisition and the false positive rate caused by high-frequency UAV maneuvers, a smaller judgment threshold is set to ensure that multipath signals are not missed when the normal Doppler frequency shift broadening index generated by UAV maneuvers is small. Based on the receiver phase-locked loop thermal noise bandwidth and the upper limit of typical aerial survey speed, 0.5dB-Hz / s is determined as the normal median value of this threshold. To assess the boundary effectiveness of this threshold setting, an absolute lower limit of 0.1dB-Hz / s and an absolute upper limit of 1.0dB-Hz / s are selected to form a three-point support boundary comparison model.

[0035] Multipath interference intensity was set to correspond to low, medium, and high levels of building obstruction density gradient. A control group using a conventional signal-to-noise ratio (SNR) weighting strategy, a control group using phase compensation but partially missing location constraint models, and an experimental group using the complete scheme of this invention were constructed. Noisy raw observation data was input, and the rate of change of SNR within a set time window was extracted. Under low-density interference conditions, the integer ambiguity fixation rate of the control group remained at 85.2%, while that of the experimental group was 91.5%. When the interference escalated to high-density obstruction conditions, the rate of change of SNR for multiple satellite channels exceeded 0.5 dB-Hz / s. The control group, due to the large number of observations removed, experienced a significant decrease in variance-covariance moments. The integer ambiguity fixation rate decreased to 21.4% and a floating-point solution was output due to array distortion. In the partially missing control group, the fixation rate rebounded to 62.7% after compensation of the variance-covariance matrix, but candidate solution divergence still existed. The experimental group triggered carrier phase compensation for the multipath fading satellite channel, maintained the full rank of the variance-covariance matrix, and used the maximum physical displacement threshold ΔL to truncate the divergent search branches. The integer ambiguity fixation rate output by the experimental group reached 88.6%, and the root mean square error of its centimeter-level mapping coordinates stabilized at 26.4 mm. The phase delay deviation value decoupled from the measured phase increment and the reference line-of-sight constraint increment suppressed the random polarization noise caused by multipath.

[0036] The impact of exceeding the threshold of the signal-to-noise ratio change rate judgment on the verification data shows a nonlinear performance inflection point. When the judgment threshold is below the lower limit of 0.1 dB-Hz / s, the system misjudges the normal high-frequency maneuvers of the UAV as multipath interference, frequently triggering invalid compensation and increasing the solution delay to 42.3 ms. When the judgment threshold is above the upper limit of 1.0 dB-Hz / s, the system produces a saturation effect, and a large number of real multipath contaminations are missed, causing the positioning error to suddenly increase to 1.25 m. Near the operating point of 0.5 dB-Hz / s, the integer ambiguity fixing time is shortened to 1.4 s, and the coordinate residual convergence is stable. The experimental data confirms that the carrier phase reverse compensation term and the physical motion envelope constraint model produce synergistic effects, eliminating the ill-conditioned risk of the spatial geometric matrix while retaining the full observation weight, and meeting the technical requirements of robustness of the positioning reference and topological consistency of the mapping model under complex occlusion conditions.

[0037] Example 3: When near-ground mapping tasks in smart land acquisition and resettlement areas cause UAVs to frequently perform large yaw and pitch maneuvers, the three-axis acceleration and three-axis angular velocity output by the onboard inertial measurement unit are natively attached to the dynamically changing body coordinate system. However, the spatial coordinates and carrier phase observations of the BeiDou satellite are calculated in the geocentric coordinate system. To eliminate the risk of topological misalignment between the two heterogeneous spatial references and avoid phase compensation divergence caused by projection distortion, the system extracts the three-axis angular velocity increment of the inertial measurement unit in the current sampling period to construct an attitude update quaternion. This attitude update quaternion is used to rotate and map the three-axis acceleration in the body coordinate system to the local navigation coordinate system, thereby separating the gravitational acceleration component and the Coriolis force component of Earth's rotation, and extracting the local navigation coordinate system and the geocentric coordinate system. The latitude and longitude transformation matrix between the Earth-fixed coordinate system and the Geocentric Earth-fixed coordinate system is used to map the three-dimensional motion acceleration after removing interference components to the Geocentric Earth-fixed coordinate system to generate an absolute spatial acceleration vector. At the same time, the three-dimensional spatial coordinates of the multipath fading satellite channel corresponding to the current epoch are obtained by analyzing the broadcast ephemeris of the BeiDou satellite navigation system. The difference between the three-dimensional spatial coordinates of the UAV phase center and the three-dimensional spatial coordinates of the multipath fading satellite channel is calculated and normalized to generate the line-of-sight geometric unit vector in the Geocentric Earth-fixed coordinate system. The dot product of the absolute spatial acceleration vector and the line-of-sight geometric unit vector is calculated to extract the scalar acceleration of the UAV along the line of sight of satellite signal propagation. The scalar acceleration is then integrated twice in the time domain according to the sampling time interval to output the accurate line-of-sight geometric displacement increment that eliminates the coordinate system rotation error.

[0038] To determine the consistency between the extracted line-of-sight geometric displacement increment and the velocity integral displacement calculated based on Doppler frequency shift, the system constructs a dynamic calibration model that integrates multi-source fundamental noise to replace the fixed empirical constants. It reads the standard deviation of acceleration measurement noise and the standard deviation of Doppler frequency tracking error output from the current epoch receiver's phase-locked loop from the inertial measurement unit's calibration parameters. Based on the error propagation law, it generates a dynamic physical displacement tolerance threshold T between adjacent epochs. This dynamic physical displacement tolerance threshold satisfies the logical constraint formula... Where T is the dynamically calculated displacement tolerance limit, k is the dimensionless confidence scaling factor, and 0.5 is the integration constant. λ represents the standard deviation of the acceleration measurement noise of the inertial measurement unit, Δt is the data sampling time interval, and λ is the signal wavelength of the current carrier channel. To determine the standard deviation of the Doppler frequency tracking error of the receiver phase-locked loop, the system inputs the calculated dynamic physical displacement tolerance threshold into the preset tolerance range judgment logic. When the absolute value of the difference between the line-of-sight geometric displacement increment and the velocity integral displacement is less than the dynamic physical displacement tolerance threshold, it is determined that the spatial geometric integral path is within the linear constraint range. Then, the line-of-sight geometric displacement increment is confirmed as the reference line-of-sight constraint increment and decoupled from the measured phase increment to obtain the deterministic multipath phase delay. This model uses sensor hardware noise parameters to define the physical kinematic constraint boundary, maintains the real-time alignment between the multipath phase extraction trigger condition and the external signal evolution state, and enables the residual reconstruction process of the surveying system in a high-frequency dynamic environment to have an objective data convergence basis.

[0039] Example 4: Under the initial assembly condition of the intelligent land acquisition and relocation UAV mapping platform, the underlying hardware baseline calibration procedure is initiated to determine the set time window length for the signal-to-noise ratio. The fundamental vibration frequency of the UAV rotor in hovering state and the thermal noise bandwidth parameter of the onboard BeiDou satellite navigation system receiver's RF front-end are read. The fundamental vibration frequency and thermal noise bandwidth parameter are input into the Nyquist aliasing determination matrix to calculate the lower boundary frequency of the overlapping frequency band. Based on the lower boundary frequency, the corresponding minimum fluctuation period is extracted. The set time window length is assigned as a preset constant multiple of this minimum fluctuation period to construct an offline calibration parameter set. In this set, the confidence ratio coefficient is fixedly set to a number. A value of 3 is used to define a 99.7% statistical confidence interval for the error, ensuring that the physical displacement judgment envelope has sufficient engineering margin. At the same time, the time window length is set to 200 milliseconds based on the 50Hz rotor vibration frequency characteristics of the UAV and the fluctuation period measured by the oscilloscope at the initial power-on stage. This length fully covers 10 vibration energy cycles, thereby effectively filtering out false signal-to-noise ratio changes caused by mechanical vibration of the fuselage. This calibration process establishes a data mapping between the observation period of the multipath signal evolution characteristics and the inherent resonance properties of the hardware, eliminating the signal truncation bias caused by short time slices and the calculation lag bias caused by long time slices in the multipath fading judgment.

[0040] In the pre-deployment phase of the surveying and mapping system in the smart land acquisition and resettlement area, environmental constraint data of the survey area is loaded to fine-tune the dynamic acceleration limit. Real-time extreme values ​​of gust wind shear at the takeoff position are obtained through airborne meteorological probes, and the total takeoff mass after the surveying and mapping payload is loaded is read. The upper limit of usable maneuvering acceleration under the survey environment is calculated and generated. The available upper limit of maneuver acceleration satisfies the physical constraint equation. ,in This represents the upper limit of usable maneuver acceleration. For the maximum available thrust of the UAV's power system, To cope with the aerodynamic drag generated by extreme gust wind shear. The upper limit of available maneuvering acceleration is set for the total takeoff mass. The dynamic acceleration limit register written to the three-dimensional motion parameters updates the calculation boundary of the maximum physical displacement threshold of the carrier phase double difference observation equation. The pre-debugging procedure enables the position constraint model to generate an adaptive shrinking calculation boundary when dealing with different weather changes and load changes. The truncation boundary of the integer ambiguity space search is set within the range of the physical flight limit envelope data.

[0041] Example 5: To address the hardware alignment and noise separation requirements of the receiver and inertial measurement unit (IMU) of the UAV-borne BeiDou satellite navigation system during the spatial data fusion process, the system initiates a procedure for sensor arm error measurement and low-level noise characteristic extraction during the pre-calibration stage. It measures the three-dimensional physical offset distance from the navigation center of the IMU to the phase center of the receiver antenna, constructs the arm error matrix, and collects at least 30 minutes of raw observation sequences from the inertial sensor and satellite under static open conditions. The system inputs the static output sequence of the inertial sensor into an autoregressive moving average model, extracts the Gaussian white noise variance and random walk coefficients, synchronously statistically analyzes the reference frequency tracking residual sequence of the phase-locked loop under unobstructed conditions, and establishes the initial boundary condition set for noise variables in the dynamic physical displacement tolerance threshold calculation model.

[0042] In the multipath error modeling stage, a nonlinear mapping between signal power characteristics and carrier phase delay is constructed to establish a function fitting optimization path based on gradient descent. Signal power characteristic value sequences of in-phase and orthogonal branches of the same satellite channel under different multipath interference intensities are extracted. The discretized ratio of the power characteristic value amplitudes is calculated to construct a characteristic ratio vector. The corresponding known measured phase delay is used as the supervisory signal input to a polynomial fitting network. Iterative calculations are performed until the sum of squared residuals output by the network converges to a target interval of no more than 0.05 cycles. The system locks the polynomial coefficient matrix at this convergence state as a fixed parameter for the characteristic function mapping relationship, and the parameter matrix is ​​called in real-time during flight mapping. The new input amplitude ratio feature vector, the 5th order polynomial fitting coefficients are set in descending order of power as 0.00000018, 0.0000042, 0.00055, 0.018, 0.35 and a constant term of 1.25; the input quantity is the real-time amplitude ratio of the in-phase branch signal power to the quadrature branch signal power in the carrier tracking loop, the effective physical mapping range is defined as 0.05 to 0.95, the mapping calculation period is locked at 20 milliseconds to match the signal fading frequency of the UAV in complex obstruction environment, and the output is an equivalent variance correction quantity that matches the current environment, which meets the underlying calculation requirements of ambiguity decorrelation transformation matrix reconstruction under complex electromagnetic interference.

[0043] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-precision mapping and positioning method for intelligent land acquisition and relocation based on the fusion of BeiDou and UAVs, characterized in that, Includes the following steps: Step S1: Obtain raw satellite observation data including BeiDou satellite carrier phase observations, pseudorange observations, and Doppler frequency shift, and obtain three-dimensional motion parameters including the real-time velocity and acceleration of the UAV. Step S2: Construct carrier phase double-difference observation equations based on the original satellite observation data, and calculate the maximum physical displacement threshold between adjacent epochs based on the three-dimensional motion parameters to generate the position constraint model of the carrier phase double-difference observation equations. Step S3: The multipath reflection signal is modeled using signal power characteristics to generate phase delay deviation value, and the phase delay deviation value is substituted as a reverse compensation term into the variance-covariance matrix of the carrier phase double difference observation equation to maintain the full rank characteristic of the variance-covariance matrix without removing the observation values ​​of the interfered satellite. Step S4: Using the position constraint model as the search boundary, within the ambiguity space determined by the compensated variance-covariance matrix, candidate integer ambiguities exceeding the maximum physical displacement threshold are eliminated, and dimensionality reduction search is performed on the remaining candidate solutions to obtain fixed solutions for integer ambiguities. Step S5: Calculate and output centimeter-level mapping coordinates based on the integer ambiguity fixed solution.

2. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, The step of generating the phase delay deviation value in step S3 includes: monitoring the signal power characteristic values ​​of the in-phase branch and the quadrature branch in the carrier tracking loop; calculating the amplitude ratio between the signal power characteristic values; performing a function mapping between the carrier phase observation value and the amplitude ratio to separate the phase difference between the direct wave signal and the reflected wave signal, so as to generate the phase delay deviation value.

3. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, Step S3, which involves compensating for the variance-covariance matrix, includes: converting the phase delay deviation value into an equivalent variance correction amount for the phase observation value of the corresponding satellite channel; adjusting the diagonal variance elements of the corresponding satellite channel in the variance-covariance matrix based on the equivalent variance correction amount; and reconstructing the ambiguity decorrelation transformation matrix using the adjusted variance-covariance matrix.

4. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, Step S4, which involves eliminating candidate integer ambiguities, includes: determining the maximum displacement envelope interval of the current epoch relative to the previous epoch based on the preset flight path envelope and dynamic acceleration limit in the three-dimensional motion parameters; converting each candidate integer ambiguity to the spatial position domain to generate candidate displacement increments; and determining that the corresponding candidate integer ambiguity is invalid and performing search branch truncation if the candidate displacement increment is outside the maximum displacement envelope interval.

5. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 4, characterized in that, The maximum physical displacement threshold ΔL satisfies the following judgment logic: Where ΔL is the theoretical displacement magnitude of the UAV between adjacent epochs, v is the real-time velocity in the three-dimensional motion parameters, a is the acceleration in the three-dimensional motion parameters, and Δt is the data sampling time interval.

6. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, It also includes the following steps: Step S61: Monitor the carrier phase jump status in the original satellite observation data in real time; Step S62: When a cycle slip is determined to occur, use inertial navigation data to correct the Doppler frequency shift in the original satellite observation data to compensate for the phase center angular velocity deviation of the UAV antenna. Step S63: Predict the carrier phase increment of the current epoch using the corrected Doppler frequency shift, and perform phase value bridging on the satellite channel where cycle slip occurs using the carrier phase increment.

7. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, It also includes the following steps: Step S71, calculate the fluctuation frequency of the phase delay deviation value on the time axis; Step S72, when the fluctuation frequency exceeds the preset jitter threshold, reduce the weight factor of the corresponding satellite channel in the variance-covariance matrix to suppress the intrusion of high-frequency multipath interference on the accuracy of centimeter-level mapping coordinates.

8. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, It also includes the following steps: Step S81: While acquiring centimeter-level mapping coordinates, record the original 3D point cloud data collected by the UAV payload; Step S82: Use the posterior variance factor of the integer ambiguity fixed solution to perform spatial reference alignment on the original 3D point cloud data in order to correct the local model distortion caused by signal occlusion.

9. The intelligent land acquisition and resettlement high-precision mapping and positioning method based on the fusion of BeiDou and UAVs as described in claim 1, characterized in that, The steps in step S3 to maintain the full-rank characteristic of the variance-covariance matrix include: eliminating residuals by substituting the phase delay deviation value into the carrier phase double-difference observation equation, so that the multipath damaged signal returns to the equivalent direct wave observation state.

10. A high-precision mapping and positioning method for intelligent land acquisition and relocation based on the fusion of BeiDou and UAVs, as described in claim 1, is characterized in that... The steps in step S5 to output centimeter-level mapping coordinates include: converting the solution coordinates corresponding to the fixed integer ambiguity solution to the projected coordinate system preset by the task; if the standard deviation of the calculated positioning accuracy is less than 30mm, then the current centimeter-level mapping coordinates are marked as valid results and stored in the spatial database.

Images