Low-orbit satellite network high-precision positioning enhancement method for complex and remote terrain construction

Abstract

The application discloses a low-orbit satellite network high-precision positioning enhancement method for complex remote terrain construction and belongs to the technical field of satellite navigation. The application comprises the following steps: S1, for the complex working condition of remote terrain construction, satellite observation data, inertial measurement data and environment perception data are collected in real time; S2, inertial measurement unit pre-integration is carried out based on the inertial measurement data, the results of the inertial measurement unit pre-integration are used for dynamic Doppler compensation of a carrier tracking loop of a satellite receiver, satellite original observation values are stabilized and output, quality marks are made according to the metadata of the satellite original observation values, and satellite observation factors with quality weights are generated; S3, a factor graph containing state variables and constraint factors is constructed, the factor graph is subjected to nonlinear optimization, the optimal value of the state variable is solved, and a positioning result is output, the low-orbit satellite network high-precision positioning effectiveness is improved, and the problem of low low-orbit satellite network high-precision positioning effectiveness in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation technology, and in particular to a high-precision positioning enhancement method for low-orbit satellite networks used in construction projects in complex and remote terrain. Background Technology

[0002] Low Earth Orbit (LEO) satellites typically orbit at altitudes between 500 and 2000 kilometers. They are close to the Earth's surface, move at high speeds, and experience minimal signal propagation loss. Current technologies primarily rely on Global Navigation Satellite Systems (GNSS) to achieve high-precision positioning. However, with the deepening of global digital transformation, the demand for location information has expanded from open outdoor environments to complex scenarios such as urban canyons, mountainous areas, and mining areas, particularly in fields like large-scale infrastructure construction, resource exploration, and emergency rescue.

[0003] To improve navigation and positioning performance, various enhancement techniques have been developed in existing technologies. For example, Inertial Navigation Systems (INS), as an autonomous navigation system, can utilize Inertial Measurement Units (IMUs) for dead reckoning, providing continuous navigation information with a high update rate. Therefore, GNSS / INS integrated navigation technology is widely adopted, fusing the two through filters to achieve complementary advantages. At the signal processing level, antenna arrays and adaptive beamforming techniques are used to suppress multipath interference and improve the quality of observations. In the field of precise positioning, linear combinations of multi-frequency, multi-system observations and integer least squares search algorithms (such as the LAMBDA method) are core technologies for achieving rapid resolution of carrier phase ambiguity. Meanwhile, environmental perception technologies, such as Simultaneous Localization and Mapping (SLAM), can maintain relative positioning by perceiving the surrounding environment in the absence of external signals.

[0004] For example, Chinese invention patent CN109001786B discloses a positioning method and system based on navigation satellites and low-Earth orbit augmentation satellites, comprising: S1, acquiring observation data from navigation satellites and low-Earth orbit augmentation satellites at the current epoch; S2, acquiring navigation messages from navigation satellites and low-Earth orbit augmentation satellites to obtain precise orbits and clock biases; S3, correcting positioning errors based on the acquired navigation messages; S4, obtaining a unified linear observation equation by normalization using one of the satellite navigation systems as a reference and calculating the observed values ​​of positioning parameters; S5, calculating estimated values ​​of positioning parameters based on the calculated observed values ​​of positioning parameters and the estimated values ​​of positioning parameters from the previous epoch; S6, generating and saving the positioning and velocity measurement results for the current epoch based on the estimated values ​​of positioning parameters.

[0005] For example, Chinese invention patent CN113687402B discloses a real-time positioning method for low-Earth orbit navigation enhancement that takes into account satellite orbit errors. The method includes: receiving observation data and navigation messages using a ground receiver; calculating the orbits and clock errors of navigation satellites and low-Earth orbit satellites; performing positioning in a conventional manner; and after positioning convergence, estimating orbit error parameters and absorbing the orbit errors of low-Earth orbit satellites and navigation satellites by giving appropriate weights.

[0006] The above-mentioned technology has at least the following technical problems: In existing technologies, the high-speed motion of low-Earth orbit satellites relative to the ground causes severe Doppler shift, and the short coverage time of a single satellite necessitates frequent beam or satellite switching by ground terminals. This results in a high failure rate for switching in signal-blocked areas, easily causing location service interruptions and failing to meet the requirements for continuous operation. Secondly, in complex terrain environments such as canyons and dense forests, satellite signals are easily blocked, and the multipath effect caused by reflections from ground objects is significant, leading to a deterioration in the quality of carrier phase observations. This causes a sharp increase in positioning errors that are difficult to effectively correct using traditional models.

[0007] Furthermore, due to the rapid changes in their geometry, low-Earth orbit (LEO) satellite constellations can theoretically significantly shorten the convergence time for precise point positioning. However, in complex terrain scenarios, the number of available satellites is limited and the observation geometry deteriorates, making the resolution of carrier phase integer ambiguities exceptionally difficult. Consequently, the convergence time is prolonged, resulting in low effectiveness of high-precision positioning in LEO satellite networks. Summary of the Invention

[0008] To address the technical problem of low effectiveness of high-precision positioning in existing low-Earth orbit (LEO) satellite networks, this invention provides a method for enhancing high-precision positioning in LEO satellite networks for construction in complex and remote terrain. The technical solution is as follows: S1. For complex construction conditions in remote terrain, real-time acquisition of satellite observation data, inertial measurement data, and environmental perception data is performed. S2. Based on the inertial measurement data, inertial measurement unit (IMU) pre-integration is performed to generate IMU pre-integration factors. The results of IMU pre-integration are used to perform dynamic Doppler compensation on the carrier tracking loop of the satellite receiver to stabilize the output of the original satellite observations. Quality labeling is performed based on the metadata of the original satellite observations to generate satellite observation factors with quality weights. S3. A factor graph containing state variables and constraint factors is constructed. Nonlinear optimization is performed on the factor graph to solve for the optimal values ​​of all state variables, thereby outputting the positioning results. The state variables of the factor graph include the position, velocity, attitude, and satellite ambiguity parameters of the carrier used to carry the navigation and positioning system.

[0009] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. The high-precision positioning enhancement method for low-orbit satellite networks in complex and remote terrain construction provided by this invention simultaneously collects satellite observation data, inertial measurement data, and environmental perception data. Existing technologies rely solely on single satellite positioning or inertial navigation data, or simply fuse the two types of data. This invention not only fuses satellite observation data and inertial measurement data but also introduces environmental perception data. Environmental perception data provides detailed information about the surrounding environment, helping to solve problems such as satellite signal obstruction and multipath effects. The mutual supplementation and verification of multiple data sources effectively reduces the errors and uncertainties that may exist with a single data source. Furthermore, existing technologies neglect the impact of the Doppler effect caused by carrier motion on signal reception when processing satellite observation data, leading to errors in the original satellite observation values. This invention utilizes the pre-integration results of the inertial measurement unit to perform dynamic Doppler compensation on the carrier tracking loop of the satellite receiver, effectively eliminating interference caused by the Doppler effect and stably outputting accurate original satellite observation values, thus providing a basis for subsequent... High-precision positioning provides a reliable data foundation, avoiding positioning deviations caused by data errors. It uses metadata from raw satellite observations for quality labeling and generates satellite observation factors with quality weights. Existing technologies do not fully consider differences in data quality when processing satellite observation data. Furthermore, existing technologies often employ traditional algorithms such as Kalman filtering for positioning. These algorithms have limitations in handling nonlinear problems under complex conditions and in multi-source data fusion. This invention constructs a factor graph containing state variables and constraint factors, performs nonlinear optimization on the factor graph, solves for the optimal values ​​of all state variables, and outputs the positioning results. Through nonlinear optimization algorithms, under all constraints, it finds the state variable values ​​that best match all observation data and prior information, thereby obtaining the precise position, velocity, and attitude information of the carrier. By setting constraint factors, the constraint capability of the factor graph is further enhanced, making the positioning results more accurate and stable, especially in complex conditions where it can effectively overcome various errors and interferences.

[0010] 2. This invention, through extended Kalman filtering time updates, can estimate the carrier's motion state in real time, providing accurate foundational information for subsequent Doppler shift prediction. From the state vector obtained after extended Kalman filtering time updates, the real-time velocity estimate of the carrier at the current moment is directly obtained. Using this extracted real-time velocity estimate, the predicted Doppler shift caused by the relative motion between the receiver and the satellite is calculated. This predicted Doppler shift can predict carrier frequency changes in advance, providing accurate frequency compensation information for the carrier tracking loop. This enables the carrier tracking loop to better track the satellite signal carrier and reduce the impact of changes in carrier frequency. The signal loss and tracking errors caused by the Doppler effect can be effectively eliminated or reduced by performing dynamic Doppler compensation on the received satellite signals. This improves the tracking accuracy and stability of the satellite signals, thereby obtaining more accurate raw satellite observations and providing a reliable data foundation for subsequent navigation and positioning calculations. Compared with existing technologies that rely solely on information from the satellite signals themselves for carrier tracking and Doppler compensation, this invention combines data from the inertial measurement unit (IMU) to more accurately predict Doppler frequency shift and achieve more precise dynamic Doppler compensation, thus improving carrier tracking accuracy and navigation positioning. To improve the accuracy and stability of attitude estimation, the system first obtains the angular velocity data measured by the three-axis gyroscope and then processes it using a quaternion integration algorithm to update the attitude information in the state vector. This avoids the singularity problem in the Euler angle representation method. The velocity is obtained by integrating acceleration data, enabling real-time updates of the carrier's velocity and position. Considering the measurement noise characteristics of the inertial measurement unit, the covariance matrix of the state vector is updated to reflect the uncertainty changes in state estimation. This solves the problem of signal loss and increased positioning errors caused by the inability of existing technologies to track satellite signal carriers in a timely and accurate manner in high-dynamic environments. Finally, the real-time velocity estimate of the carrier and the velocity of the satellite are projected onto the line-of-sight direction. The relative radial velocity of the carrier and the satellite in the line-of-sight direction is calculated using vector dot product operations. Based on the Doppler effect, the calculated relative radial velocity and the known satellite signal carrier frequency are used to calculate the Doppler frequency shift prediction value. The Doppler frequency shift prediction value can accurately predict the carrier frequency change caused by the relative motion between the receiver and the satellite, providing accurate frequency compensation information for the carrier tracking loop and improving the accuracy and stability of carrier tracking.

[0011] 3. By quality-labeling the metadata of the raw satellite observations, the satellite observation signals can be quickly and accurately preliminarily screened and classified. Based on the relationship between the carrier-to-noise ratio (CNR) of the raw satellite observations and the preset CNR upper and lower limits, a CNR weight is set, which can preliminarily reflect the quality of the satellite signal strength. Based on the comparison between the satellite elevation angle and the satellite elevation angle limit, an elevation angle weight is set. The satellite elevation angle affects factors such as atmospheric attenuation during signal propagation. By calculating the geometric distribution factor of all currently available satellites and comparing it with the geometric distribution factor limit, the geometric distribution factor weight is determined. The geometric distribution factor reflects the impact of the satellite's distribution in space on positioning accuracy. The quality weight of the satellite observation factor is obtained by multiplying the CNR weight, elevation angle weight, and geometric distribution factor weight. This comprehensively labels the raw satellite observations by considering multiple factors such as signal strength, propagation path, and satellite spatial distribution.

[0012] 4. Satellite observation factor weights are set using carrier-to-noise ratio and geometric distribution factor. Based on quality labeling, satellites with excellent signal quality are given higher weights, allowing for more efficient use of their observation data in factor map construction to improve positioning accuracy. If a satellite observation signal disappears, the satellite observation factor weight is set to the lowest possible weight to avoid adverse effects from poor-quality satellite signals on positioning results. If the satellite signal quality is intermediate, making it difficult to directly determine its future availability, the current real-time trajectory and the future satellite position calculated from ephemeris are input into the satellite quality prediction model. This model outputs the signal availability probability for the future time period, providing a scientific basis for satellite selection and switching. Finally, the model selects the satellite with the best signal quality. Using the satellite with the highest availability as the primary broadcasting satellite ensures the most reliable satellite signal at the current moment, guaranteeing the quality of positioning services. Simultaneously, a backup broadcasting satellite is set up to allow for timely switching in case of problems with the primary broadcasting satellite, improving the reliability of low-Earth orbit satellite network positioning enhancement. By predicting the changing trends of the primary broadcasting satellite signal in advance, a proactive switching strategy is implemented before potential signal problems arise, initiating the acquisition and guidance of the new primary broadcasting satellite signal in advance. This avoids positioning jitter or interruption caused by sudden signal loss, solving the problem of existing technologies that rely solely on current satellite signal quality indicators for satellite selection and switching decisions, resulting in a lack of predictive ability for future signal changes. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0014] Figure 1A flowchart illustrating a method for enhancing high-precision positioning of a low-orbit satellite network for construction in complex and remote terrain, as provided in this application embodiment; Figure 2 A flowchart of inertial measurement unit pre-integration and dynamic Doppler compensation provided for embodiments of this application; Figure 3 A flowchart for obtaining satellite observation factors provided in this application embodiment; Figure 4 A line graph showing the change of positioning error over time before high-precision positioning enhancement of the low-orbit satellite network provided in this application embodiment; Figure 5 A line graph showing the change of positioning error over time after high-precision positioning enhancement of the low-orbit satellite network provided in the embodiments of this application; Figure 6 A flowchart illustrating the satellite observation factor weight setting provided in this application embodiment. Detailed Implementation

[0015] The following provides explanations for some of the terms used in this application. It should be noted that these explanations are for the convenience of those skilled in the art and do not constitute a limitation on the scope of protection claimed in this application.

[0016] The embodiments of this application involve at least one, including one or more; wherein, multiple means two or more. Furthermore, it should be understood that in the description of this specification, terms such as "first," "second," and "third" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order. For example, "first device" and "second device" do not represent the degree of importance of the two or their order, but are merely for descriptive distinction. In the embodiments of this application, "and / or" merely describes an association relationship, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship. In the embodiments of this application, the terms "including," "comprising," "having," and their variations all mean "including but not limited to," unless otherwise specifically emphasized.

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] like Figure 1 The diagram shows a flowchart of a high-precision positioning enhancement method for low-Earth orbit satellite networks used in construction in complex and remote terrain, provided in an embodiment of this application. The method includes the following steps: For complex construction conditions in remote terrain, such as building a hydropower station in a remote mountainous area, there may be challenges such as steep terrain, communication interruptions, and GNSS signals being blocked by mountains.

[0019] S1 collects satellite observation data, inertial measurement data, and environmental perception data in real time. Satellite observation data includes pseudorange, carrier phase, and Doppler shift. Inertial measurement data includes data from the three-axis accelerometer and three-axis gyroscope in the inertial measurement unit. Environmental perception data is point cloud data from the lidar. Pseudorange represents the measured path length of the satellite signal from transmission to reception. Carrier phase represents the phase difference between the carrier signal received by the receiver and the local reference carrier signal during propagation, which can be used for high-precision positioning. When there is relative motion between the satellite and the receiver, the frequency of the satellite signal received by the receiver changes; this frequency change is called Doppler shift, which can be used to measure information such as velocity. The inertial measurement unit consists of a three-axis accelerometer and a three-axis gyroscope, used to measure the acceleration and angular velocity of objects. The fusion of these multiple data sources provides a rich information base for subsequent high-precision positioning. Satellite observation data provides global positioning information, inertial measurement data can provide short-term positioning and attitude information when satellite signals are blocked, and environmental perception data can provide detailed information about the surrounding environment, which helps to improve the accuracy and reliability of positioning, especially in complex working conditions where it can better adapt to different environmental conditions.

[0020] S2. Based on inertial measurement data, inertial measurement unit (IMU) pre-integration is performed to generate IMU pre-integration factors. Through integration processing of the IMU data, discrete acceleration and angular velocity data are converted into changes in position, velocity, and attitude, providing constraint information for subsequent factor map construction. The IMU pre-integration results are used to perform dynamic Doppler compensation on the carrier tracking loop of the satellite receiver to stabilize the output of the original satellite observations. Since the carrier generates the Doppler effect during motion, affecting the reception of satellite signals, dynamic compensation can eliminate the impact on satellite signal reception and ensure the accuracy of the original satellite observations. Quality labeling is performed based on the metadata of the original satellite observations to generate satellite observation factors with quality weights. By evaluating the quality of the observation data, different weights are assigned to different observations. In the subsequent optimization process, high-quality observations will have a greater impact on the positioning results, thereby improving the positioning accuracy.

[0021] like Figure 2 The flowchart shown illustrates the inertial measurement unit (IMU) pre-integration and dynamic Doppler compensation. Specifically, the IMU pre-integration is performed based on the inertial measurement data, and the detailed process is as follows: Within a preset time window, based on the inertial measurement data and its initial values ​​at the start of the preset time window, the attitude change, velocity change, and position change are obtained by integrating the angular velocity and acceleration data of the inertial measurement data. The covariance matrix is ​​then updated using the covariance propagation equation. The preset time window is a pre-defined time interval within which the inertial measurement data is processed and analyzed. The size of the time window can be set according to specific application scenarios and requirements. In this application, the covariance matrix is ​​used to describe the correlation and variance among the attitude change, velocity change, and position change, reflecting the degree of uncertainty of the inertial measurement data changes. Obtaining the covariance matrix provides important uncertainty information for subsequent state estimation, reflecting the measurement accuracy and reliability of the attitude change, velocity change, and position change, enabling more reasonable fusion of inertial measurement data during factor graph optimization and improving the accuracy of state estimation.

[0022] The attitude change is obtained by integrating the three-axis gyroscope data in the inertial measurement data. The three-axis gyroscope data reflects the angular velocity information of the carrier on three orthogonal axes and is used to describe the rotational motion of the object. In this application, the integration process adopts a numerical integration method, such as rectangular integration. Integrating the three-axis gyroscope data in the inertial measurement data can convert instantaneous angular velocity information into accumulated attitude change information, thereby describing the rotational motion of the carrier within a preset time window. The attitude change is a key parameter for describing the attitude change of the carrier in inertial navigation.

[0023] The velocity change is obtained by integrating the triaxial accelerometer data from the inertial measurement system (IMS). The triaxial accelerometer data reflects the acceleration information of an object along three orthogonal axes, including gravitational acceleration and acceleration generated by the object's motion, and can be used to describe the object's translational motion. Since the accelerometer measurements include the influence of gravitational acceleration, the influence of gravitational acceleration needs to be removed before integrating the triaxial accelerometer data in the IMS. For example, the accelerometer measurements can be converted to the navigation coordinate system, and then the gravitational acceleration component can be subtracted. Integrating the triaxial accelerometer data in the IMS converts instantaneous acceleration information into cumulative velocity change information, describing the velocity change of the carrier's translational motion within a preset time window. The velocity change is an important basis for estimating the carrier's velocity in inertial navigation.

[0024] The change in position is obtained by integrating the change in velocity. Through double integration, the acceleration information is converted into position change information, thus describing the position change of the vehicle within a preset time window. The change in position is a key parameter for estimating the position of the vehicle in inertial navigation.

[0025] The inertial measurement variations, including attitude changes, velocity changes, and position changes, along with their covariance matrices, are used as pre-integration factors for the inertial measurement unit. These factors connect the state nodes at the start and end times of a preset time window in the factor graph. The factor graph constructs a global optimization problem by connecting different state nodes and factor nodes, thereby solving for the optimal state estimate. Through factor graph optimization, the complementarity of various sensor data can be fully utilized to improve the accuracy and robustness of the carrier state estimate. At the same time, pre-integration processing reduces the computational load during the integration process.

[0026] Specifically, dynamic Doppler compensation is performed on the carrier tracking loop of the satellite receiver using the pre-integration results of the inertial measurement unit. The specific process is as follows: First, after receiving inertial measurement data, a time update step based on extended Kalman filtering is performed at a preset update time point. This involves using the extended Kalman filter algorithm to update the system's state over time, predicting the current state vector, which includes the carrier's position, velocity, and attitude. Extended Kalman filtering is an extension of standard Kalman filtering. In this application, the extended Kalman filter algorithm predicts the carrier's current state vector based on the state vector from the previous time step and the inertial measurement data. Through time updates using the extended Kalman filter algorithm, the carrier's motion state can be estimated in real time using high-frequency measurement data from the inertial measurement unit, providing an accurate state basis for subsequent dynamic Doppler compensation. This can, to some extent, overcome the uncertainties of the system model and the influence of measurement noise, improving the accuracy of state estimation.

[0027] It should be added that the specific process for performing the time update step based on the extended Kalman filter is as follows: The filtering algorithm function, namely the Extended Kalman Filter, is invoked. The state vector and covariance matrix of the previous update time are used as the initial values ​​for the current time update and as the input values ​​for the Extended Kalman Filter. This provides a benchmark for the current state estimation and ensures the continuity and accuracy of the filtering process.

[0028] The angular velocity data of the carrier around three orthogonal axes is obtained from a three-axis gyroscope. The angular velocity data measured by the three-axis gyroscope is processed by a quaternion integration algorithm to obtain the quaternion at the current moment, so as to update the attitude information in the state vector. The orthogonal axes include roll, pitch and yaw axes. The quaternion integration algorithm includes quaternion differential equations and discrete integrals such as Euler method, midpoint method and Runge-Kutta method. It means that the quaternion is updated by integrating the angular velocity data, thereby reflecting the attitude change of the carrier, and thus updating the attitude information of the carrier in real time and accurately, providing a basis for subsequent coordinate transformation and velocity and position updates.

[0029] By combining the attitude information in the updated state vector, i.e., the quaternion at the current moment, the specific force data measured by the triaxial accelerometer is subjected to coordinate transformation, and the gravitational acceleration component is subtracted to obtain the acceleration of the carrier in the navigation coordinate system. The specific force represents the acceleration measured by the accelerometer, which includes the vector sum of the carrier acceleration and the gravitational acceleration. The coordinate transformation means transforming the specific force data measured by the accelerometer from the carrier coordinate system to the navigation coordinate system, and the coordinate transformation is achieved through a rotation matrix. This eliminates the influence of gravitational acceleration and obtains the true acceleration information of the carrier, providing accurate input for subsequent velocity and position updates. Compared with the Euler angles and other representation methods in the prior art, quaternions have the advantages of high computational efficiency and no singularities, and can more accurately reflect the attitude changes of the carrier.

[0030] The acceleration data is integrated to obtain the vehicle's velocity at the current moment, thus updating the vehicle's velocity information in real time and reflecting the vehicle's motion state. The updated velocity is then integrated to obtain the vehicle's position at the current moment, thus updating the vehicle's position in real time and achieving navigation and positioning for the vehicle.

[0031] The measurement noise of an inertial measurement unit (IMU) is typically modeled as Gaussian white noise and random walk noise. A process noise covariance matrix is ​​constructed based on the IMU's Gaussian white noise and random walk noise. This process noise covariance matrix is ​​then input into a Kalman filter covariance prediction algorithm to update the state vector covariance matrix. This accurately reflects the uncertainty of the state vector, providing accurate prior information for subsequent measurement update steps and improving the accuracy and robustness of the filter. The noise model describes the statistical characteristics of the IMU's measurement error. The covariance matrix update represents adjusting the covariance matrix based on changes in the noise model and the state vector to reflect the uncertainty at the current moment.

[0032] Then, the real-time velocity estimate is directly extracted from the predicted state vector at the current moment using a programming language (such as Python or MATLAB). Accurate extraction of the real-time velocity estimate can provide reliable data for subsequent Doppler frequency shift calculations, ensuring the accuracy of the calculation.

[0033] Next, based on the real-time velocity estimate of the carrier at the current moment, combined with the satellite's position information (usually known or obtained through other means), the carrier frequency change caused by the relative motion between the carrier and the satellite is calculated using the Doppler shift formula; this is the predicted Doppler shift value. The carrier tracking loop is a crucial part of the satellite receiver, used to track the received satellite signal carrier and keep it synchronized with the carrier signal generated by the local oscillator. By accurately tracking the carrier, the information carried in the satellite signal, such as navigation messages, can be demodulated. The Doppler shift affects the accuracy of carrier tracking, so compensation is necessary. When there is relative motion between the receiver and the satellite, the frequency of the received satellite signal changes; this phenomenon is called the Doppler effect, and the change in frequency is the Doppler shift. By calculating the predicted Doppler shift value, the impact of the relative motion between the satellite receiver and the satellite on the carrier frequency can be known in advance, providing accurate frequency change predictions for the carrier tracking loop, enabling it to adjust the frequency of the local oscillator in a timely manner to maintain synchronization with the received carrier.

[0034] It should be added that, based on the real-time velocity estimate, the predicted Doppler frequency shift of the carrier tracking loop in the satellite receiver is calculated. The specific process is as follows: By receiving ephemeris data broadcast by satellites, and using the ephemeris data of the currently tracked satellites and corresponding orbital calculation models, such as the Kepler orbital model, the position coordinates and velocity vector of the satellite in the Earth coordinate system at the current moment can be accurately calculated. Accurately obtaining the satellite's position and velocity is the basis for subsequent calculations. Ephemeris data is a series of data parameters describing the satellite's orbital position and motion state, including information such as the satellite's position, velocity, orbital inclination, and right ascension of the ascending node in a specific coordinate system.

[0035] Vector calculations are performed based on the current position of the carrier and the current position of the tracked satellite. For example, the magnitude of the difference between the current position of the carrier and the current position of the tracked satellite is obtained. The ratio of the difference between the current position of the carrier and the current position of the tracked satellite to the magnitude of the difference between the current position of the carrier and the current position of the tracked satellite is calculated to obtain the line-of-sight vector from the carrier to the satellite. The line-of-sight vector is a vector with direction and magnitude pointing from the carrier position to the satellite position. It describes the relative directional relationship between the carrier and the satellite, thus providing directional information for subsequent calculation of relative radial velocity, enabling accurate extraction of the relative motion components between the carrier and the satellite in the line-of-sight direction.

[0036] By performing vector dot product operations, the real-time velocity estimate of the carrier and the current velocity of the tracked satellite are projected onto the line-of-sight vector. The relative radial velocity in the line-of-sight direction refers to the difference between the projection of the carrier velocity onto the line of sight and the projection of the satellite velocity onto the line of sight. The relative radial velocity is the relative motion velocity between the carrier and the satellite in the line-of-sight direction, that is, the velocity component of the carrier relative to the satellite along the line connecting the two. A positive relative radial velocity indicates that the carrier is moving away from the satellite, and a negative relative radial velocity indicates that the carrier is moving closer to the satellite. Accurate calculation of the relative radial velocity can improve the accuracy of Doppler frequency shift prediction, thereby improving the performance of the satellite receiver carrier tracking loop.

[0037] Based on the Doppler effect principle, the relative radial velocity and carrier frequency are multiplied, and the product of the relative radial velocity and carrier frequency is compared with the speed of light to obtain the predicted Doppler frequency shift value; the carrier frequency is the frequency of the electromagnetic wave carrying information in the satellite signal.

[0038] Finally, the predicted Doppler shift value is fed back to the carrier tracking loop. The carrier tracking loop adjusts the frequency of the local oscillator based on the predicted Doppler shift value, using it as the initial frequency offset of the numerically controlled oscillator (CNC). This ensures a closer match between the local oscillator frequency and the received satellite signal carrier frequency, achieving dynamic Doppler compensation. Simultaneously, after completing signal tracking, the baseband signal processing module inside the GNSS receiver extracts the raw satellite observations from the tracking loop and outputs them via the receiver's configured communication interface (such as UART, USB, or Ethernet) in a specific data protocol format (such as Receiver Independent Exchange Format, RINEX). Dynamic Doppler compensation effectively reduces the impact of Doppler shift on the carrier tracking loop, improving the accuracy and stability of carrier tracking, thereby ensuring the demodulation quality of the satellite signal and ultimately improving the accuracy and reliability of satellite navigation and positioning.

[0039] like Figure 3 The flowchart shown illustrates the acquisition of satellite observation factors. Specifically, quality labeling is performed based on the metadata of the raw satellite observations. The detailed process is as follows: The metadata of the raw satellite observations includes carrier-to-noise ratio, satellite elevation angle, and geometric distribution factor.

[0040] If the carrier-to-noise ratio (CNR) of the satellite signal corresponding to the original satellite observation is greater than the upper limit of CNR, then the CNR weight is set as the first CNR weight. The CNR represents the ratio of carrier power to noise power density, reflecting the quality of the satellite signal. The higher the CNR, the less noise interference the signal is subject to, and the better the signal quality. The output power of the prompt correlator is directly used as the carrier power. The power output of the non-prompt correlator (Early / Late) is used as the noise power density. If the CNR of the original satellite observation is not greater than the upper limit of CNR, but greater than the lower limit of CNR, then the CNR weight is set as the second CNR weight. The upper and lower limits of CNR are manually set thresholds used to define different quality ranges of CNR. The upper limit of CNR is greater than the lower limit of CNR. If the carrier-to-noise ratio (CNR) of the original satellite observations is not greater than the lower limit of the CNR, then the CNR weight is set as the third CNR weight. The first, second, and third CNR weights decrease progressively. The first, second, and third CNR weights are set by pre-defined personnel; for example, the first CNR weight can be set to 0.6, the second to 0.3, and the third to 0.1. By using the upper and lower limits of the CNR, the CNR is divided into three intervals, so that different weights can be assigned to different intervals. A higher CNR indicates better signal quality and is given a higher weight, which will play a greater role in subsequent comprehensive evaluation. A lower CNR indicates poorer signal quality and is given a lower weight to reduce the adverse impact of the satellite signal on the overall results.

[0041] If the satellite elevation angle corresponding to the original satellite observation is less than the satellite elevation angle limit, the elevation angle weight is set to the first elevation angle weight; otherwise, the elevation angle weight is set to the second elevation angle weight, with the first elevation angle weight being less than the second elevation angle weight. The satellite elevation angle represents the elevation angle between the line connecting the satellite to the observation point and the ground plane. The elevation angle affects the propagation path of the satellite signal. If the elevation angle is too small, the signal has a longer path through the atmosphere, is more susceptible to atmospheric interference, and has poorer signal quality. The satellite elevation angle is obtained through RTKLIB. The satellite elevation angle limit is a threshold set manually to define the quality of the satellite elevation angle. The first and second elevation angle weights are set by preset personnel; for example, the first elevation angle weight can be set to 0.3, and the second elevation angle weight can be set to 0.7. By setting different elevation angle weights, the quality of the satellite observation can be further evaluated, making the quality assessment more comprehensive and accurate.

[0042] The geometric distribution factor of all currently available satellites is obtained based on their geometric distribution, reflecting the degree of influence of satellite geometry on positioning accuracy. The geometric distribution factor is obtained based on the direction cosine matrix, specifically: the positions of the satellite receiver and all available satellites are obtained; in the same coordinate system, the unit direction vector pointing from each satellite to the receiver is calculated; the direction cosines and clock bias terms of all available satellites are combined into a geometric matrix; the weight inverse matrix is ​​obtained from the geometric matrix; and the geometric accuracy attenuation factor is extracted from the diagonal elements of the weight inverse matrix and denoted as the geometric distribution factor. The geometric accuracy attenuation factor comprehensively reflects the geometric influence of three-dimensional position and time errors. If the geometric distribution factor is greater than the geometric distribution factor limit, then the geometric distribution factor is... The distribution factor is set as the weight of the first geometric distribution factor; otherwise, the weight of the geometric distribution factor is set as the weight of the second geometric distribution factor, with the weight of the first geometric distribution factor being less than that of the second geometric distribution factor. The geometric distribution factor limit is a threshold set manually to define the quality requirements of the geometric distribution factor. Based on this limit, the geometric distribution factor is divided into two intervals and assigned different weights. The weights of the first and second geometric distribution factors are set by pre-selected personnel. For example, the weight of the first geometric distribution factor can be set to 0.3, and the weight of the second geometric distribution factor can be set to 0.7. By assigning different weights to the geometric distribution factors, the impact of the satellite's geometric distribution on the quality of the observations can be reflected, making the quality assessment more consistent with the actual situation.

[0043] The product of the carrier-to-noise ratio weight, elevation angle weight, and geometric distribution factor weight is denoted as the quality weight of the satellite observation factor. The quality weight is a comprehensive weight value obtained by taking into account the carrier-to-noise ratio weight, elevation angle weight, and geometric distribution factor weight, and is used to measure the quality status of the original satellite observations. The satellite observation factor includes the original satellite observations and the quality weight.

[0044] like Figure 4 The line graph shown illustrates the change in positioning error over time before high-precision positioning enhancement for the low-Earth orbit satellite network. Figure 5 The graph shown is a line graph illustrating the change in positioning error over time after high-precision positioning enhancement of the low-Earth orbit satellite network. Figure 4 and Figure 5 The graphs show the changes in positioning error over time before and after augmentation for low-Earth orbit satellite positioning. Different colors are used to distinguish errors in different directions: gray line represents x-direction error, red line represents y-direction error, and blue line represents z-direction error. Furthermore, the graphs also show... Figure 4 and Figure 5 The comparison shows that the positioning error decreases significantly faster after enhancement. At the same time point, the error values ​​in each direction before positioning enhancement are generally higher than those after positioning enhancement. This intuitively reflects that without LEO satellite positioning enhancement technology, high-precision positioning of LEO satellite networks takes longer to reach a relatively stable state, and the positioning accuracy is relatively low.

[0045] In addition, due to Figure 4 and Figure 5 The basic principles of localization are the same, and the interference factors are similar, but the optimization of localization results by this invention is limited. Figure 4 and Figure 5 The overall trends are similar. From Figure 4 and Figure 5 The data shows that the positioning error decreases over time. When time is 0, the satellite receiver needs to initialize parameters, such as its initial position and clock offset. The positioning algorithm performs preliminary calculations based on inaccurate initial parameters, leading to a larger positioning error in the initial stage. As time progresses, by continuously receiving and processing satellite signals, these parameters are gradually optimized, causing the positioning result to gradually approach the true value. The positioning error then decreases and stabilizes. Specifically, the positioning error in all directions decreases rapidly between 0-10 minutes, and gradually stabilizes between 15-40 minutes, with the positioning error becoming stable and relatively small. Between 10-15 minutes, a satellite handover operation may occur. During the handover, the receiver needs to relock to the new satellite's signal and update relevant parameters. During this brief period, the satellite information used for the positioning calculation is incomplete or inaccurate, leading to an increase in error. In satellite positioning, to ensure continuous positioning service, when the signal quality of a satellite deteriorates or is about to fall outside the reception range, the system switches to another satellite.

[0046] The system uses a base station as the test point. The coordinates of the base station are calculated based on the received satellite signals. The Euclidean distance formula is used to compare the calculated coordinates of the base station with the coordinates of the standard base station to obtain the positioning error.

[0047] S3. Construct a factor graph containing state variables and constraint factors. Perform nonlinear optimization on the factor graph to find the optimal values ​​of all state variables, thereby outputting the positioning results. That is, use nonlinear optimization algorithms (such as the Gauss-Newton method) to iteratively calculate the factor graph, continuously adjusting the values ​​of state variables to minimize the objective function (such as the sum of squared residuals), thus obtaining the optimal estimates of the state variables. These optimal values ​​are the positioning results, such as the precise position and clock error of the receiver. In other words, the factor graph is used to represent the relationship between state variables and constraint factors. Through nonlinear optimization algorithms, under the condition of satisfying all constraints, find the state variable values ​​that best match all observation data and prior information, thereby obtaining the precise position, velocity, and attitude information of the carrier, i.e., quaternions.

[0048] The state variables in the factor diagram include the position, velocity, attitude, and satellite ambiguity parameters of the carrier used to carry the navigation and positioning system. The state variables describe the motion state of the carrier and the propagation of the satellite signal. The satellite ambiguity parameters are denoted as integer ambiguity. The constraint factors include the inertial measurement unit pre-integration factor connecting the state variables at adjacent time points, the satellite observation factor connecting the state variables and the satellite ambiguity parameters, and the relative pose constraint factor providing the relative pose relationship of the carrier at different time points, as well as the a priori constraint factors, including the position prior factor, the atmospheric delay prior factor, and the ambiguity prior factor. These constraint factors establish the relationship between the state variables, forming a complete constraint network. When using carrier phase for positioning, due to the periodicity of the carrier signal, the receiver cannot directly determine the integer number of signal propagation cycles. The integer number of cycles is the satellite ambiguity parameter.

[0049] By solving the problem through nonlinear optimization, the accuracy and reliability of positioning are improved. The introduction of relative pose constraint factors and prior constraint factors further enhances the constraint capability of the factor graph, making the positioning results more accurate and stable, and providing an effective solution for navigation and positioning in complex scenarios such as construction in remote terrain.

[0050] like Figure 6 The flowchart shown illustrates the satellite observation factor weight setting process. Specifically, after quality labeling based on the metadata of the original satellite observations and before constructing the factor graph containing state variables and constraint factors, the process also includes: If the carrier-to-noise ratio (CNR) is greater than the upper limit and the geometric distribution factor (GDP) is not greater than the GDP limit, it indicates that the satellite observation signal is good. The satellite observation factor weight is then set to the highest weight, making the original satellite observation values ​​more important during the nonlinear optimization of the factor graph. The highest satellite observation factor weight is set by pre-selected personnel based on experience, for example, it can be set to 1. This accurately identifies satellite observation data with good signals and assigns them the highest weight, enabling the use of the original satellite observation values ​​during the nonlinear optimization of the factor graph, thus improving positioning accuracy and reliability. This is because the better the satellite observation signal, the more accurate and reliable the information contained in the satellite signal, and the greater its contribution to the positioning results.

[0051] If the carrier-to-noise ratio (CNR) is not greater than the lower limit of the CNR and the geometric distribution factor (GDP) is greater than the GDP limit, it indicates that the satellite observation signal has disappeared. The satellite observation factor weight is set to the lowest possible weight, and the observation data of this satellite is ignored as much as possible in subsequent data processing to avoid adverse effects on the positioning results. The lowest satellite observation factor weight is set by the pre-set personnel based on experience, for example, it can be set to 0.1. This allows for the timely identification of satellite observation data with missing signals or extremely poor quality, and setting its weight to the lowest possible value. This can effectively avoid interference and misleading of the positioning results by low-quality raw satellite observations. If poor-quality raw satellite observations are used during the positioning process, it may lead to increased positioning errors or even positioning failure.

[0052] If the above conditions are not met—namely, if the carrier-to-noise ratio (CNR) is greater than the lower limit but not greater than the upper limit, if the CNR is greater than the upper limit and the geometric distribution factor (GDP) is greater than the GDP limit, or if the CNR is not greater than the lower limit and the GDP is not greater than the GDP limit—then the satellite observation signal quality is in an intermediate state, and its future availability cannot be directly determined. The current real-time trajectory (reflecting the receiver's motion state and position change trend) and the future satellite position calculated from the ephemeris are input into the satellite quality prediction model, which outputs the signal availability probability for a future time period. The satellite quality prediction model is a Long Short-Term Memory (LSTM) network model. The ephemeris describes the satellite's orbit and includes parameters such as the satellite's position and velocity at various times. Through the ephemeris, the satellite's spatial position at any given time can be accurately calculated, providing basic data support for satellite positioning. Predicting the signal availability probability for a future time period through the satellite quality prediction model provides forward-looking information for subsequent data processing and decision-making. In the positioning system, satellites participating in the calculation can be selected based on the prediction results to optimize the positioning scheme and improve the continuity and accuracy of positioning.

[0053] It should be added that the current real-time trajectory and the future satellite positions calculated from ephemeris are input into the satellite quality prediction model, which outputs the signal availability probability for a future time period, and then includes: Based on the signal availability probability of each satellite in the future time period output by the satellite quality prediction model, all satellites are ranked, and the satellite with the highest signal availability probability is selected as the primary broadcasting satellite. This maximizes the accuracy and reliability of positioning and other operations because the signal of this satellite is most likely to be stably available during the predicted time period. The satellite with the second highest signal availability probability is selected as the backup broadcasting satellite. The signal availability probability represents an indicator of the likelihood that a satellite signal can be received normally and used for positioning and other applications within a specific time period. Setting up backup broadcasting satellites provides redundancy protection. When the primary broadcasting satellite encounters a problem, it can be quickly switched to the backup satellite, reducing the positioning interruption time caused by satellite failure and improving the fault tolerance and stability of low-Earth orbit satellite positioning.

[0054] Simultaneously, the system continuously monitors and predicts the signal availability probability of the broadcast satellite for future time periods. If the predicted signal availability probability is lower than the signal availability probability limit after a preset time, an active handover strategy is implemented to avoid positioning jitter or interruption caused by sudden signal interruption. Otherwise, the system continues to acquire the satellite signal from the broadcast satellite, avoiding unnecessary satellite handover operations and reducing system resource consumption. The signal availability probability limit is a pre-set threshold used to determine whether the satellite signal meets the conditions for normal use. When the satellite signal availability probability is lower than this limit, the signal is considered unreliable. By predicting and judging the signal availability probability of the broadcast satellite, potential signal interruption risks can be detected in advance, avoiding positioning jitter or interruption caused by handover only when the signal is suddenly interrupted, thus improving the continuity and stability of positioning.

[0055] The proactive switching strategy manifests itself in several ways: First, it initiates the acquisition and tracking of new satellite signals in advance, avoiding positioning jitter or interruption caused by sudden signal outages. In this application, "acquisition" refers to the receiving device initially locking onto a satellite signal and acquiring satellite observation data through a process of searching and identifying the signal. "Tracking" refers to further adjusting the receiving device's parameters after acquiring the satellite signal, enabling it to more accurately track changes in the satellite signal, maintain stable signal reception, and improve the quality and reliability of signal reception. Initiating the acquisition and tracking of new satellite signals in advance ensures that the receiving device completes signal preparation before the new signal becomes available. When the original satellite signal is interrupted, it can immediately use the new satellite signal for positioning, achieving seamless switching and effectively avoiding positioning jitter or interruption.

[0056] Specifically, the output includes the location results, followed by: After obtaining the positioning results, the reference station coordinates, raw satellite observations, and atmospheric delay information are extracted from the factor map. This information forms the basis for differential positioning. The reference station coordinates provide a reference benchmark, the raw satellite observations contain the measurement information required for positioning, and the atmospheric delay information is used to correct for atmospheric effects on satellite signals during propagation. This provides an accurate data source for subsequent differential data generation. For example, variable nodes can be extracted by searching for markers of the reference station coordinates, and the values ​​can be read directly. In differential positioning, the coordinates of stations with known precise locations are used as reference station coordinates. Satellite signals are delayed by the ionosphere and troposphere when passing through the atmosphere. The atmospheric delay information includes ionospheric delay and tropospheric delay.

[0057] The extracted information is encoded according to international standard formats to generate differential data streams. In the broadcasting of differential data, specific international standard formats (such as Radio Technical Commission for Maritime Services, RTCM format) are followed. The differential data stream represents a data stream containing base station coordinates, raw satellite observations, and atmospheric delay information after encoding. It is broadcast to the rover station in real time through a communication link for differential positioning calculations. The use of international standard formats for encoding ensures that the generated differential data stream has universality and compatibility.

[0058] Differential data streams are broadcast via a communication link, which represents the channel used to transmit differential data streams. Broadcasting differential data streams via a communication link enables data sharing between the base station and the rover, allowing the rover to use the high-precision information from the base station to correct observation errors, thereby improving the positioning accuracy of the rover and achieving high-precision real-time or near-real-time positioning services.

[0059] Example 2: Based on Example 1, it also includes: By fitting historical observation data into a physical model describing satellite motion, the parameters in the model are determined, thus establishing a satellite state dynamics model. This physical model considers various perturbations, such as Earth's gravity, gravity caused by Earth's non-spherical shape and non-uniform mass, solar radiation pressure, and the gravitational forces of the Sun, Moon, and other planets. Through mathematical modeling and numerical calculations, the influence of these perturbations on the satellite orbit is quantified. High-precision numerical integration methods, such as the Runge-Kutta method and the Adams method, can be used to solve the satellite state dynamics model, more accurately simulating the satellite's orbital state, including orbital changes under various perturbations. To accurately calculate the gravitational effects of the Sun, Moon, and other planets on the satellite, high-precision astronomical ephemeris (such as the JPL planetary ephemeris) can be used to obtain celestial position information. Simultaneously, a higher-order gravity field model (such as EGM2008) is used to describe the complex changes in Earth's gravitational field. The satellite state dynamics model provides a basic framework for satellite orbit prediction and state estimation, enabling a preliminary description of the satellite's motion trend in space.

[0060] It is important to understand that the inputs to the satellite state dynamics model include perturbation model parameters and historical observation data; the perturbation model parameters include Earth's gravity, the gravitational force caused by Earth's non-spherical shape and non-uniform mass, solar radiation pressure, and the gravitational forces of the Sun, Moon, and other planets; the outputs of the satellite state dynamics model include the satellite state sequence and orbital elements; the satellite state sequence includes the satellite position, satellite velocity, and satellite acceleration; the orbital elements include the semi-major axis, eccentricity, orbital inclination, right ascension of the ascending node, argument of perigee, and true anomaly.

[0061] Using the satellite state dynamics model as a prior model for Kalman filtering, the Kalman filtering algorithm is used to recursively process historical observation data. In each recursive step, the algorithm combines the model predictions of the satellite state dynamics model with new observations, and adjusts the model parameters of the satellite state dynamics model by calculating the Kalman gain. This makes the model predictions of the satellite state dynamics model more consistent with the observation data, thereby continuously optimizing the model parameters of the satellite state dynamics model. By processing historical observation data through the Kalman filtering algorithm, model information and observation information are effectively integrated, overcoming the limitations of a single data source, significantly improving the accuracy of satellite orbit parameters, and providing a more accurate foundation for subsequent real-time positioning.

[0062] During satellite operation, real-time observation data is acquired through remote sensing processing software such as ERDAS Imagine. This data is then input into an optimized Kalman filter framework. Based on the new real-time observation data and the current model state, the Kalman filter calculates the optimal estimate of the satellite's state at the current moment and simultaneously updates the model parameters of the satellite state dynamics model. The updated model is then used to predict the satellite's state over a future period. Real-time observation data refers to data acquired in real time during satellite operation through ground observation stations, onboard sensors, and other equipment related to the satellite's position, velocity, and signal propagation time. By introducing observation data in real time and dynamically correcting the model and satellite state, the system can promptly reflect changes in the satellite's state during actual operation, such as minor disturbances. This improves the tracking capability and prediction accuracy of low-Earth orbit (LEO) satellite states, making LEO satellite orbit prediction more accurate and reliable, and enhancing the real-time performance and adaptability of LEO satellite positioning.

[0063] Point cloud data from different regions, such as digital elevation models, are collected. Historical satellite positioning data is compared with standard single-point positioning results at the same time and location. Positioning error vectors for different regions at different time periods are statistically analyzed. Correlation analysis is performed between terrain features and positioning error vectors. Through data mining and modeling techniques (such as geographic weighted regression), a regional error mapping model is constructed that can predict satellite positioning errors based on geographical location. Terrain features refer to the topographical characteristics of different regions on the Earth's surface. Terrain features affect the propagation of satellite signals. For example, in mountainous areas, satellite signals may be blocked or reflected by mountains, causing changes in the signal propagation path and resulting in errors. Historical error distribution represents the distribution of errors in different regions and time periods obtained through analysis of a large amount of past satellite positioning data. The regional error mapping model considers the impact of terrain factors on satellite signal propagation and can predict potential positioning errors based on the characteristics of different regions, providing a geospatial basis for subsequent error correction and helping to improve the accuracy and uniformity of satellite positioning in different regions.

[0064] The specific construction and training method of the regional error mapping model is as follows: A machine learning model, such as a neural network, is used to model multipath errors, and historical positioning residuals are used as training samples for training. These historical positioning residuals contain information about multipath errors. Historical observation data (such as satellite signal strength and arrival time) are used as input features of the regional error mapping model, and historical positioning residuals are used as labels to train the machine learning model, enabling it to learn the mapping relationship between input features and multipath errors. During satellite signal propagation, when it encounters reflective objects such as buildings and the ground, the reflected signal and the direct signal arrive at the receiver simultaneously, causing interference between the received signals and resulting in measurement errors. This error caused by multipath propagation is called multipath error. The historical positioning residual represents the difference between the actual satellite positioning result and the true location. The machine learning model can automatically learn the complex characteristics and patterns of multipath errors from large amounts of data. Compared with traditional physical model-based methods, it can more accurately capture changes in multipath errors, improve the modeling accuracy of multipath errors, and provide a more effective basis for subsequent error correction.

[0065] Based on real-time acquired satellite position information and ground receiver geographic location information, a regional error mapping model is used to determine the possible distribution of ionospheric and tropospheric errors in the current region. Simultaneously, a trained machine learning model is used to predict the multipath error at the current moment based on real-time observation data, generating ionospheric, tropospheric, and multipath error correction parameters for a preset time window in real time. These ionospheric, tropospheric, and multipath error correction parameters represent parameters used to correct the effects of satellite signals passing through the ionosphere and troposphere, as well as errors caused by multipath effects. The real-time generated error correction parameters can be promptly applied to the satellite positioning process to correct various errors and further improve the accuracy of satellite positioning.

[0066] The ionospheric delay correction and its uncertainty, as well as the tropospheric delay correction and its uncertainty, can be used as the prior constraint factors of the factor map in Example 1. The multipath error prediction values ​​are subtracted from the original pseudorange and carrier phase obtained in Example 1, and the satellite observation factor weights of the satellite observation data and the square of the uncertainty of the multipath error are calculated to obtain the new satellite observation factor weights.

[0067] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)). Where there is no conflict, the solutions in the above embodiments can be combined.

[0068] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0069] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0070] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0071] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0072] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope and intent of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and variations.

Claims

1. A high-precision positioning enhancement method for low-orbit satellite networks used in construction projects in complex and remote terrain, characterized in that: Includes the following steps: S1 is designed for complex construction conditions in remote terrain, and collects satellite observation data, inertial measurement data, and environmental perception data in real time. S2, based on the inertial measurement data, perform inertial measurement unit pre-integration to generate inertial measurement unit pre-integration factor, use the result of inertial measurement unit pre-integration to perform dynamic Doppler compensation on the carrier tracking loop of the satellite receiver to stabilize the output of the original satellite observation value, and perform quality labeling based on the metadata of the original satellite observation value to generate a satellite observation factor with quality weight; S3. Construct a factor graph containing state variables and constraint factors, perform nonlinear optimization on the factor graph, solve for the optimal values ​​of all state variables, and output the positioning result. The state variables of the factor graph include the position, velocity, attitude and satellite ambiguity parameters of the carrier used to carry the navigation and positioning system.

2. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 1, characterized in that: The specific process for pre-integrating the inertial measurement unit based on the inertial measurement data is as follows: Within a preset time window, based on the inertial measurement data and the initial values ​​of the inertial measurement data at the beginning of the preset time window, the covariance matrices of attitude change, velocity change, and position change are obtained. The attitude change is obtained by integrating the three-axis gyroscope data in the inertial measurement data; The change in velocity is obtained by integrating the triaxial accelerometer data from the inertial measurement data. The change in position is obtained by integrating the change in velocity; The inertial measurement variations, including attitude changes, velocity changes, and position changes, and their covariance matrix are used as pre-integration factors for the inertial measurement unit to connect state nodes in the factor graph within a preset time window.

3. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 1, characterized in that: The specific process for performing dynamic Doppler compensation on the carrier tracking loop of the satellite receiver using the pre-integration result of the inertial measurement unit is as follows: After receiving inertial measurement data, at a preset update time point, a time update step based on extended Kalman filtering is performed to predict the state vector at the current moment, which includes the position, velocity and attitude of the vehicle. Extract the real-time velocity estimate for the current moment from the predicted state vector at the current moment; Based on the real-time velocity estimate, the predicted Doppler frequency shift of the carrier tracking loop in the satellite receiver is calculated to predict the frequency change caused by the relative motion between the receiver and the satellite. The predicted Doppler frequency shift is fed back to the carrier tracking loop, and the original satellite observations are output for dynamic Doppler compensation.

4. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 3, characterized in that: The specific process for performing the time update step based on extended Kalman filtering is as follows: Read the state vector and covariance matrix from the previous update time; The angular velocity data measured by the three-axis gyroscope is processed by a quaternion integration algorithm to update the attitude information in the state vector; By combining the attitude information in the updated state vector, coordinate transformation is performed on the specific force data measured by the triaxial accelerometer to obtain the acceleration of the vehicle in the navigation coordinate system; The acceleration data is integrated to update the velocity of the carrier; The updated speed is integrated to update the carrier's position; The covariance matrix of the state vector is updated based on the noise model of the inertial measurement unit.

5. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 3, characterized in that: The specific process for calculating the predicted Doppler frequency shift of the carrier tracking loop in the satellite receiver based on the real-time velocity estimate is as follows: Obtain the ephemeris data of the currently tracked satellite to calculate its position and velocity; Calculate the line-of-sight vector from the vehicle to the satellite based on the current position of the vehicle and the current position of the tracked satellite; Calculate the relative radial velocity in the line-of-sight direction based on the real-time velocity estimate and the current velocity of the tracked satellite; The Doppler shift prediction is calculated based on the relative radial velocity and the carrier frequency of the currently tracked satellite.

6. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 1, characterized in that: The process of quality labeling based on the metadata of the original satellite observations is as follows: If the carrier-to-noise ratio of the satellite signal corresponding to the original satellite observation is greater than the upper limit of the carrier-to-noise ratio, then the carrier-to-noise ratio weight is set as the first carrier-to-noise ratio weight. If the carrier-to-noise ratio of the original satellite observation is not greater than the upper limit of the carrier-to-noise ratio, but is greater than the lower limit of the carrier-to-noise ratio, then the carrier-to-noise ratio weight is set as the second carrier-to-noise ratio weight. If the carrier-to-noise ratio of the original satellite observations is not greater than the lower limit of the carrier-to-noise ratio, then the carrier-to-noise ratio weight is set as the third carrier-to-noise ratio weight. The values ​​of the first carrier-to-noise ratio weight, the second carrier-to-noise ratio weight, and the third carrier-to-noise ratio weight decrease sequentially. If the satellite elevation angle in the original satellite observation is less than the satellite elevation angle limit, then the elevation angle weight is set to the first elevation angle weight; otherwise, the elevation angle weight is set to the second elevation angle weight, and the first elevation angle weight is less than the second elevation angle weight. Based on the geometric distribution of the currently available satellites, the geometric distribution factor of all currently available satellites is obtained. If the geometric distribution factor is greater than the geometric distribution factor limit, the geometric distribution factor weight is set to the first geometric distribution factor weight; otherwise, the geometric distribution factor weight is set to the second geometric distribution factor weight. The first geometric distribution factor weight is less than the second geometric distribution factor weight. The product of the carrier-to-noise ratio weight, elevation angle weight, and geometric distribution factor weight is denoted as the quality weight of the satellite observation factor. The satellite observation factors include raw satellite observations and quality weights.

7. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 6, characterized in that: The process of quality labeling based on the metadata of the raw satellite observations also includes: If the carrier-to-noise ratio is greater than the upper limit of the carrier-to-noise ratio and the geometric distribution factor is not greater than the limit of the geometric distribution factor, it indicates that the satellite observation signal is good. The satellite observation factor weight is set to the highest satellite observation factor weight. The satellite observation factor weight is used to quantify the degree of influence of the satellite observation factor on the nonlinear optimization process of the factor map. If the carrier-to-noise ratio is not greater than the lower limit of the carrier-to-noise ratio and the geometric distribution factor is greater than the limit of the geometric distribution factor, then the satellite observation factor weight is set to the lowest satellite observation factor weight. If the conditions for setting the highest satellite observation factor weight are not met, and the conditions for setting the lowest satellite observation factor weight are not met, then the current real-time trajectory and the future satellite position calculated from the ephemeris are input into the satellite quality prediction model, and the signal availability probability for the future time period is output.

8. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 7, characterized in that: The process of inputting the current real-time trajectory and the future satellite position calculated from ephemeris into the satellite quality prediction model, and outputting the signal availability probability for a future time period, further includes: The satellite with the highest signal availability probability will be used as the main broadcasting satellite, and the satellite with the second highest signal availability probability will be used as the backup broadcasting satellite. Meanwhile, if it is predicted that the signal availability probability of the satellite broadcaster is less than the signal availability probability limit after a preset time, an active switching strategy is implemented to avoid positioning jitter or interruption caused by sudden signal interruption; otherwise, the satellite signal broadcaster continues to be acquired.

9. The high-precision positioning enhancement method for low-orbit satellite networks in complex and remote terrain construction as described in any one of claims 1-6, characterized in that: Also includes: A satellite state dynamics model is established based on historical observation data and physical models describing satellite motion. The historical observation data is processed using the Kalman filter algorithm to optimize the parameters of the satellite state dynamics model; Real-time observation data is introduced, and the dynamic model and satellite state are dynamically corrected and predicted using a Kalman filter framework. A regional error mapping model is constructed by combining terrain features with historical error distribution; A machine learning model is used to model multipath error, and historical positioning residuals are used as labels for training. Based on the aforementioned regional error mapping model, ionospheric, tropospheric, and multipath error correction parameters are generated in real time within a preset time window.

10. The high-precision positioning enhancement method for low-orbit satellite networks for construction in complex and remote terrain as described in claim 1, characterized in that: The output positioning result then includes: Extract the base station coordinates, raw satellite observations, and atmospheric delay information from the factor plot; The extracted information is encoded according to a preset encoding format to generate a differential data stream; The differential data stream is broadcast via the communication link.

Images