A multi-frequency high-precision positioning method and system based on beidou satellite signals

Abstract

The application discloses a multi-frequency high-precision positioning method and system based on Beidou satellite signals and belongs to the technical field of satellite navigation and positioning. The method comprises the following steps: collecting original observation data of Beidou navigation satellites through a monitoring station network; pre-processing the original observation data and solving initial rough coordinates of each monitoring station in the monitoring station network; based on the pre-processed data, using a state estimation algorithm to estimate different frequency signals inside all monitoring station receivers in real time, generating dynamic delay compensation parameters, encapsulating the dynamic delay compensation parameters to form a delay compensation parameter table; based on the pre-processed data and the delay compensation parameter table, detecting cycle slip of observation values of each satellite of each monitoring station, generating differential correction information, encoding based on the differential correction information, and generating an enhanced information data stream; based on the enhanced information data stream and the delay compensation parameter table, a fusion positioning correction model is constructed, and high-precision solving is performed on the fusion positioning correction model.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation and positioning technology, and in particular to a multi-frequency high-precision positioning method and system based on BeiDou satellite signals. Background Technology

[0002] In fields such as surveying and exploration, intelligent transportation, and disaster monitoring, high-precision satellite navigation and positioning have become critical infrastructure. From RTK technology, which relies on real-time differential positioning between reference stations and user stations, to PPP technology, which utilizes precise ephemeris to achieve precise positioning of a single device, the industry continues to pursue solutions with wider coverage, faster convergence, and higher reliability. With the full completion of the BeiDou-3 global system, the open signals from multiple frequencies such as B1C and B2a provided by it have created conditions for obtaining richer and higher-quality observation data. The industry is actively exploring a new generation of high-precision service systems based on dense reference station networks, deep fusion of multi-frequency data, and real-time intelligent processing.

[0003] Existing high-precision positioning methods mainly include: pseudorange smoothing through a fixed window (such as Hatch filtering); approximating the hardware delay of different signals within the receiver as a fixed value; using fixed spatial correlation distance for regional interpolation in ionospheric delay modeling; directly using various differential correction information when fusing positioning at the user end, lacking dynamic evaluation of data timeliness, accuracy, and consistency; and integrity assurance mechanisms that mainly focus on positioning accuracy but rarely make real-time judgments and early warnings on the reliability of the results.

[0004] However, existing technologies have shortcomings. The preprocessing stage lacks adaptive capability for pseudorange smoothing, affecting the accuracy of initial coordinates. Hardware delays of different frequency signals within the receiver are often treated as fixed values, failing to consider their time-varying characteristics, leading to systematic biases in the correction information. The differential correction information generation process exhibits poor adaptability to observation quality and ionospheric disturbances. During user-end fusion positioning, there is a lack of dynamic evaluation and intelligent weighting of the timeliness, accuracy, and consistency of multiple types of correction information. Furthermore, the lack of a multi-layered integrity detection mechanism from the data source to the solution result makes it impossible to provide real-time identification and alerts regarding the reliability of the positioning results. These deficiencies collectively make it difficult for existing methods to simultaneously guarantee high accuracy, robustness, and high reliability in complex environments. Summary of the Invention

[0005] To address the challenge of simultaneously ensuring high accuracy, robustness, and reliability in positioning under complex environments in existing technologies, this invention provides a multi-frequency high-precision positioning method and system based on BeiDou satellite signals. The specific technical solution is as follows: A multi-frequency high-precision positioning method based on BeiDou satellite signals includes the following steps: S1. Collecting raw observation data of BeiDou navigation satellites through a monitoring station network; S2. Preprocessing the raw observation data and calculating the initial coarse coordinates of each monitoring station in the monitoring station network; S3. Based on the preprocessed data, using a state estimation algorithm to estimate the different frequency signals inside the receivers of all monitoring stations in real time, generating dynamic delay compensation parameters, and encapsulating the dynamic delay compensation parameters into a delay compensation parameter table; S4. Based on the preprocessed data and the delay compensation parameter table, performing cycle slip detection on the observation values ​​of each satellite at each monitoring station sequentially, generating differential correction information, encoding the differential correction information, and generating an enhanced information data stream; S5. Based on the enhanced information data stream and the delay compensation parameter table, constructing a fusion positioning correction model, and performing high-precision calculation on the fusion positioning correction model.

[0006] Preferably, step S2 includes the following steps: The original observation data is synchronized to the same master time scale and uniformly converted into a standard format to form standardized observation data; the pseudorange observations and carrier phase observations in the standardized observation data are checked, outgoing errors exceeding the threshold are removed, and the remaining observations are marked as valid observations; for each monitoring station, satellite, and signal frequency, an adaptive weighted recursive smoothing formula based on the satellite elevation angle is used epoch-by-epoch to smooth the valid pseudorange observations, and smoothing is paused if the instantaneous residual exceeds the dynamic threshold during the smoothing process; using the smoothed pseudorange observations of each signal frequency and the broadcast ephemeris, the initial coarse coordinates of the monitoring station in the ground-fixed coordinate system and the receiver clock error are calculated.

[0007] Preferably, step S3 includes the following steps: based on the preprocessed data, extract the effective carrier phase observation values ​​of each monitoring station for each visible BeiDou satellite; construct a geometrically distance-free combined observation value using the frequency corresponding to the effective carrier phase observation values; decompose the differential hardware delay of each monitoring station into constant bias and time-varying components for estimation using state-adaptive Kalman filtering; synthesize the estimated constant bias and time-varying components into a total differential hardware delay estimate, and allocate the hardware delay compensation amount for each frequency according to the frequency ratio; encapsulate the compensation amount and the corresponding uncertainty into a delay compensation parameter table.

[0008] Preferably, step S4 includes the following steps: receiving preprocessed data and a delay compensation parameter table; after associating and correcting the delay compensation parameter table with the carrier phase observations of the corresponding monitoring station and frequency, performing cycle slip detection and marking; selecting a reference station; calculating the pseudorange integrated residual based on the quality weighting factor related to adaptive smoothing weight and multipath error; using the pseudorange integrated residual as differential correction information; calculating the vertical delay of the ionospheric puncture point using the carrier phase after cycle slip detection; and generating an ionospheric delay grid model using an adaptive correlation distance interpolation method that dynamically adjusts the correlation length with the regional ionospheric disturbance index; and encoding the differential correction information and the ionospheric delay grid model into an enhanced information data stream.

[0009] Preferably, step S5 includes: receiving BeiDou satellite signals, demodulating to obtain the user's own original multi-frequency pseudorange carrier phase observations, and receiving the enhanced information data stream and delay compensation parameter table; performing real-time quality assessment and spatiotemporal matching on the user's own original multi-frequency pseudorange, carrier phase observations, enhanced information data stream, and delay compensation parameter table, and marking invalid information; for information not marked as invalid, calculating dynamic fusion confidence weights based on information age, nominal accuracy, and multi-source consistency variance; fusing differential correction information and dynamic delay compensation parameters into the observation equation according to the dynamic fusion confidence weights, constructing a fusion positioning correction model and solving for the user's high-precision three-dimensional position coordinates, while recording the weight distribution as a basis for quality assessment.

[0010] Preferably, step S5 further includes: performing multi-layer integrity detection while completing high-precision positioning calculation on the user end, and marking and alarming the positioning results in real time.

[0011] Preferably, the multi-layer integrity detection includes: receiving high-precision three-dimensional position coordinates, dynamically fusing confidence weights, enhanced information data streams, and redundant observation information from the user; performing a first-layer data consistency check to check for information inconsistencies between different correction channels and whether the residuals of observation values ​​exceed limits; performing a second-layer external integrity comparison to compare the error boundary calculated by the user with the protection level in the enhanced information data stream and determine whether it exceeds limits; performing a third-layer internal redundancy check to recalculate and observe whether position changes are abnormal after sequentially removing satellites; and combining the results of the three layers of detection, triggering an audible and visual alarm, recording a fault log, and switching to a low-precision positioning mode if any layer alarm is triggered.

[0012] The preferred adaptive weighted recursive smoothing formula is: in, The adaptive smooth pseudo-range at epoch k-1 This represents the change in the carrier phase observation value at the same frequency between two adjacent epochs. The original pseudorange observation value at epoch k; For epoch k, the dynamic smoothing weight of the i-th monitoring station for satellite s; i is the monitoring station number, f is the frequency number, and s is the satellite number.

[0013] The preferred expression for the sum of observations without geometric distance is: in, At epoch k, the geometrically inverse combination of observations from monitoring station i and satellite s; and These are the original carrier phase observations of monitoring station i and satellite s at frequencies f1 and f2 at epoch k, respectively; f1 and f2 are the two frequencies of the BeiDou satellite signal.

[0014] This invention also provides a multi-frequency high-precision positioning system based on BeiDou satellite signals, which is applied to the aforementioned multi-frequency high-precision positioning method based on BeiDou satellite signals, including: The system comprises the following modules: an acquisition module for collecting raw observation data from BeiDou navigation satellites via a network of monitoring stations; a preprocessing module for preprocessing the raw observation data and calculating the initial coarse coordinates of each monitoring station in the network; a first compensation module for estimating different frequency signals within the receivers of all monitoring stations in real time using a state estimation algorithm based on the preprocessed data, generating dynamic delay compensation parameters, and encapsulating these parameters into a delay compensation parameter table; a second compensation module for performing cycle slip detection on the observation values ​​of each satellite at each monitoring station based on the preprocessed data and the delay compensation parameter table, generating differential correction information, encoding the differential correction information, and generating an enhanced information data stream; and a correction module for constructing a fusion positioning correction model based on the enhanced information data stream and the delay compensation parameter table, and performing high-precision calculations on the fusion positioning correction model.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: Firstly, this method utilizes a monitoring network to collect raw observation data, and through data preprocessing and adaptive smoothing, improves the accuracy of initial coordinate calculation. Secondly, it employs a state estimation algorithm to estimate the time-varying hardware delay within the receiver in real time, generating dynamic compensation parameters to eliminate systematic biases at their source. Then, based on the compensated data, cycle slip detection and gross error removal are performed to generate high-quality differential correction information. At the user end, the differential correction and dynamic delay compensation parameters are received, and a fusion positioning correction model is constructed to achieve precise error correction of the initial coordinates. Finally, while performing high-precision positioning calculations, multi-layer integrity checks are performed in parallel, and the positioning results are marked and alarmed in real time. Thus, this effectively solves the dilemma of balancing accuracy, robustness, and reliability in existing technologies, significantly improving the availability and reliability of BeiDou high-precision services in complex environments. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a flowchart of the multi-frequency high-precision positioning method based on BeiDou satellite signals in Example 1. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0020] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0021] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0022] Please refer to Figure 1 , Figure 1 This is a flowchart of a multi-frequency high-precision positioning method based on BeiDou satellite signals. This embodiment provides a multi-frequency high-precision positioning method based on BeiDou satellite signals, which includes the following steps: S1. Collect raw observation data of BeiDou navigation satellites through the monitoring station network.

[0023] In this example, a network of monitoring stations deployed within the service area synchronously tracks visible BeiDou navigation satellites and collects raw observation data. Specifically, within a pre-defined high-precision positioning service area, three or more BeiDou satellite signal monitoring stations are optimally deployed to form a monitoring network. It should be noted that the pre-defined high-precision positioning service area includes, but is not limited to, provinces, urban clusters, or specific industrial parks. Each monitoring station is equipped with, but is not limited to, a high-performance multi-frequency multi-mode BeiDou receiver, a high-stability atomic clock or temperature-compensated crystal oscillator, a meteorological sensor, and a data communication module, including but not limited to 5G / fiber optic communication. All monitoring station antennas must have their phase centers precisely calibrated and be installed on stable reference piers with a wide field of view.

[0024] Based on real-time satellite ephemeris data, the data processing center generates a unified satellite tracking plan and distributes it to all monitoring stations. Each monitoring station then synchronously and continuously tracks all visible BeiDou satellites in the sky according to this plan, covering multiple frequencies including B1I, B1C, B2I, B2a, and B3I.

[0025] The receiver internally records the following raw observation data synchronously for each frequency of each satellite: pseudorange observations, including but not limited to C / A code and P code; carrier phase observations with millimeter-level accuracy; Doppler shift observations; and decoded BeiDou navigation messages, including but not limited to broadcast ephemeris, satellite clock bias, and ionospheric model parameters. Simultaneously, meteorological sensors collect temperature, pressure, and humidity data for the local station.

[0026] Each monitoring station generates a data packet containing a high-precision timestamp, station identifier, observation data blocks, and meteorological data blocks, using a fixed high sampling rate (which can be set to 1Hz). This data packet is then streamed in real-time to the data processing center via a data communication module using the RTCM protocol. Monitoring stations must periodically perform time synchronization calibration with the data processing center to ensure time consistency.

[0027] S2. Preprocess the raw observation data and calculate the initial coarse coordinates of each monitoring station in the monitoring network. Specifically, step S2 includes the following steps: S21. The raw observation data is synchronized to the same master time scale and uniformly converted into a standard format to form standardized observation data. In practical applications, the data processing center receives the raw observation data stream, initiates the time synchronization process, and, based on the local high-precision timestamp in each data packet and the preset precise network time protocol, synchronizes all observation data to the same master time scale of the data processing center, achieving nanosecond-level precision. Simultaneously, all receiver-defined raw data formats are uniformly converted into an internal standard format using a parsing library, forming standardized observation data with consistent time stamps and identifiers.

[0028] S22. The pseudorange and carrier phase observations in the raw observation data are checked to remove outliers exceeding the threshold. Specifically, the pseudorange and carrier phase observations are quality-marked, and the checks include whether the signal-to-carrier-to-noise ratio is higher than a preset threshold, whether the observations are continuous, and whether they are within the physically permissible range. Data that fails the check is marked as invalid and will not proceed to further processing, but it is recorded to remove extremely obvious outliers caused by instantaneous interference. The remaining observations after removing outliers are marked as valid observations. At the same time, parameters such as satellite position and clock error are parsed from the navigation message and correlated with each observation to form paired data blocks of observation-satellite ephemeris.

[0029] S23. For each monitoring station, satellite, and signal frequency, the pseudorange observation value of the signal frequency is smoothed epoch by epoch using an adaptive weighted recursive smoothing formula based on the satellite elevation angle. If the instantaneous residual exceeds the dynamic threshold during the smoothing process, the smoothing is paused.

[0030] Specifically, for any monitoring station i, satellite s, and frequency f, at epoch k, the adaptive smoothing pseudorange is recursively calculated. The formula for adaptive smoothing pseudorange is: in, The adaptive smoothing pseudorange at epoch k-1; when k=1 or the channel is reinitialized. = That is, it is initialized with the primitive pseudorange of the first epoch; This represents the change in the carrier phase observation value at the same frequency between two adjacent epochs. ; The original pseudorange observation value at epoch k; The carrier phase of satellite s at epoch k. For epoch k, the dynamic smoothing weight of the i-th monitoring station for satellite s is dimensionless. It should be noted that a maximum basic weight is set. In reality The possible value is 300, and the elevation angle of satellite s at epoch k is... The effective weight is dynamically adjusted based on the elevation angle, i.e., the weight. Determined based on satellite altitude segmentation: like They considered the satellite's high-altitude operating conditions to be favorable and assigned it full weight. ; like It is believed that at a medium satellite altitude, signal quality may degrade, and the weight decreases linearly with decreasing elevation angle. ; like It is believed that the satellite's low elevation angle results in poor data reliability, therefore depth smoothing is not performed, only slight smoothing is applied. .

[0031] During adaptive weight smoothing, consistency is achieved simultaneously, i.e., the instantaneous residuals are calculated. The formula for calculating the instantaneous residuals is: if If the preset dynamic threshold is exceeded, it is suspected that the carrier phase L of the current epoch k has experienced a cycle slip or that there is a gross error in the pseudorange P. Once the suspicion is triggered, the current smoothing recursion is immediately paused. The preset dynamic threshold is usually set to 3 to 5 meters.

[0032] Through the above steps, the smoothing weights are strongly correlated with the satellite elevation angle, the geometric parameter that most directly reflects the signal propagation quality, thus achieving adaptability: when the satellite has a high elevation angle, the historical smoothing sequence is given a very high weight, which greatly suppresses pseudorange noise; when the satellite has a low elevation angle, the trust in the current original observation is rapidly increased, preventing low-quality data from contaminating the smoothing sequence.

[0033] S24. Using the smoothed pseudorange observations of each signal frequency and the broadcast ephemeris, calculate the initial coarse coordinates of the monitoring station in the Earth-fixed coordinate system and the receiver clock error.

[0034] After completing the adaptive smoothing of all valid satellite channels at the current epoch, for each monitoring station i, the adaptive smoothed pseudoranges of all satellites s at all frequencies are collected. Using the satellite positions and clock errors provided by the broadcast ephemeris, the initial coordinates of the current monitoring station in the Earth-fixed coordinate system are calculated using the least squares method or fast Kalman filtering. And receiver clock bias. Among them, the satellite position, clock bias and other parameters in the broadcast ephemeris are extracted from the observation-satellite ephemeris paired data block; using the satellite position and clock bias data provided by the observation-satellite ephemeris, the initial coordinates of the monitoring station and the receiver clock bias are calculated by least squares method or fast Kalman filtering.

[0035] Finally, a preprocessed data packet is generated for each monitoring station. The preprocessed data packet contains a time stamp, monitoring station ID, smoothed multi-frequency pseudorange, original (but time-aligned) multi-frequency carrier phase, initial coordinates, and related satellite elevation and azimuth information.

[0036] Step S3. Based on the preprocessed data, a state estimation algorithm is used to estimate the different frequency signals inside the receivers of all monitoring stations in real time, generating dynamic delay compensation parameters, and encapsulating the dynamic delay compensation parameters into a delay compensation parameter table. Specifically, step S3 includes the following steps: S31. Based on the preprocessed data, extract the effective carrier phase observation values ​​of each monitoring station for each visible BeiDou satellite. Receive the preprocessed data packets, and based on the data in the preprocessed data packets, for each monitoring station i, at each epoch k, extract the effective carrier phase observation values ​​of the current monitoring station for all visible satellites s.

[0037] S32. Construct geometrically distance-free combined observations using the frequencies f1 and f2 corresponding to the effective carrier phase observations. . The expression is: in, At epoch k, the geometrically inverse combination of observations from monitoring station i and satellite s; and These are the original carrier phase observations of monitoring station i and satellite s at frequencies f1 and f2 at epoch k, respectively; f1 and f2 are two frequencies of the BeiDou satellite signal, which are known physical constants of the BeiDou system, fixed values ​​(e.g., ...). =1561.098MHz, =1207.140MHz).

[0038] S33. The differential hardware delay of each monitoring station is decomposed into constant bias and time-varying components and estimated by using state-adaptive Kalman filtering.

[0039] Because the hardware delay (differential hardware delay) of different frequency signals inside the receiver drifts slowly over time (affected by temperature, circuit aging, etc.), traditional methods treat it as a constant, leading to estimation errors. This step decomposes it into a constant bias (long-term stable) and a time-varying component (slowly changing), and uses a state-dependent adaptive Kalman filter for real-time estimation.

[0040] Specifically, the state vector for each monitoring station i is defined as: ,in, It is a constant bias, which reflects the inherent asymmetry of the hardware. Theoretically, it is a constant, but it is allowed to change very slowly. It is a time-varying component, which reflects the slow drift caused by factors such as temperature and circuit aging.

[0041] The state prediction equation is generated based on the state vector. The state prediction equation is as follows: Where F is the state transition matrix, It is the correlation constant of the time-varying component. Δt is the epoch interval, representing the time difference between two adjacent epochs, and its unit is seconds. τ is the correlation time, which can take a value of 3600 seconds. It is close to 1 but less than 1.

[0042] The observation equation of the filter is a transformation of the linear combination of observations. The observations are defined as follows: in, It is a fixed integer ambiguity combination value that has been determined and maintained through the initial time-span wide-lane ambiguity fixing technique, and is a constant. At epoch k, the combined observations of monitoring station i and satellite s without geometric distance are given. It should be noted that the wide-lane ambiguity fixing technique is a well-known technique in the field. Its principle is as follows: a Melbourne-Wübbena combination (MW combination) is constructed using dual-frequency pseudorange and carrier phase observations. This combination eliminates error terms such as geometric distance, ionospheric delay, satellite clock bias, and receiver clock bias, retaining only the wide-lane ambiguity information. By smoothing or rounding the MW combination observations, the wide-lane integer ambiguity can be fixed. In this embodiment, after fixing the wide-lane ambiguity using this technique during the initial observation period, it is used as a constant. And it remains constant in subsequent epochs.

[0043] The observation equation is: in, It is observation noise. For state vectors, This is the observation matrix. This indicates that the mapping relationship between two state variables is formed by directly linearly superimposing them to form the observation value, that is, the observation value is equal to the sum of the constant bias and the time-varying component (plus noise).

[0044] Observation noise variance in traditional Kalman filtering Since it is a constant, this step combines it with the adaptive smoothing weights from the preprocessing stage. Correlation enables adaptive adjustment and observation of noise variance. The expression is: in, The basic observation noise variance is pre-calibrated based on the carrier phase measurement accuracy. It should be noted that for high-quality geographic mapping GNSS receivers, the carrier phase observation noise is typically in the millimeter range. If measured in "weeks" (1 week ≈ 0.19 meters), the basic observation noise variance... The typical order of magnitude is approximately 0.001 to 0.01 cycles². Specific values ​​need to be obtained through instrument calibration or zero-baseline testing, depending on the type of receiver used and the observation environment. This is the adjustment coefficient, a fixed value, which can be 2.

[0045] For each epoch, Kalman gain calculation, state update, and covariance update are performed sequentially, ultimately outputting the optimal state estimation vector for the current epoch. ,in, This is a constant bias estimate. This is the estimated value of the time-varying component.

[0046] Specifically, after calculating the variance of the basic observation noise, state prediction and covariance prediction are performed: The formula for state prediction is: ; The formula for covariance prediction is: ,in, Let be the transition matrix. Let be the posterior covariance matrix of the previous epoch. The process noise covariance matrix is... ,in, constant bias The corresponding process noise, For time-varying components The corresponding process noise.

[0047] Then calculate the Kalman gain, the formula for which is: ,in, Here is the Kalman gain matrix. To predict the covariance matrix, For the transpose of the observation matrix, To observe the noise variance.

[0048] After calculating the Kalman gain, state and covariance updates are performed. The state update formula is as follows: ,in, Let the prior state vector be... For Kalman gain, For the observed values, This is the observation matrix. After the state update, it becomes the optimal state estimate vector for the current epoch.

[0049] The covariance update formula is ,in, It is the identity matrix. For Kalman gain. Covariance update is used to provide prior information for the next epoch.

[0050] S34. Combine the estimated constant bias with the time-varying components to obtain the total differential hardware delay estimate, and allocate the hardware delay compensation amount for each frequency according to the frequency ratio.

[0051] The estimated state variables are combined to form a total differential hardware delay estimate: .according to Based on the known frequency relationships, allocate the hardware delay compensation amounts for individual frequencies f1 and f2 according to empirical proportions. and .

[0052] Specifically, the differential hardware delay (DCB, differential code offset) inside the receiver is typically frequency-dependent. When using a dual-frequency, geometrically indistinguishable combination ( Estimate the total differential hardware delay. Then, the commonly used empirical formula for assigning it to individual frequencies f1 and f2 is: Δ1 = [f2² / (f1² - f2²)] Δ2 = [f1² / (f1² - f2²)] This allocation ratio is derived from the coefficient relationship of the ionosphere-free combination and is a standard practice in GNSS data processing.

[0053] S35. Encapsulate the compensation amount and the corresponding uncertainty into a delay compensation parameter table.

[0054] Hardware delay compensation for individual frequencies f1 and f2 and The corresponding uncertainties and valid timestamps are encapsulated into a delay compensation parameter table.

[0055] It should be noted that the "corresponding uncertainty" refers to the standard deviation (or variance) of the estimation error of the hardware delay compensation, which is derived from the covariance result of the Kalman filter and is used to characterize the reliability of the compensation.

[0056] S4. Based on the preprocessed data and the delay compensation parameter table, cycle slip detection is performed on the observations of each satellite at each monitoring station to generate differential correction information. This differential correction information is then encoded to generate an enhanced information data stream. The specific steps in S4 include the following: S41. Receive the preprocessed data and delay compensation parameter table, and after associating and correcting the delay compensation parameter table with the carrier phase observation values ​​of the corresponding monitoring station and frequency, perform cycle slip detection and marking.

[0057] The system synchronously receives the main input and key correction input. The main input is preprocessed data, containing information such as smoothed pseudorange, original carrier phase observations, initial coordinates, elevation angle, and azimuth angle for each epoch of each monitoring station. The key correction input is a delay compensation parameter table, containing the hardware delay compensation amount for different signal frequencies at each epoch of each monitoring station. And its uncertainties.

[0058] Based on the monitoring station ID, satellite PRN number, signal frequency (e.g., B1I, B2a), and epoch timestamp, the compensation values ​​in the compensation parameter table are matched one-to-one with the effective carrier phase observations of the corresponding monitoring station. If a channel lacks compensation parameters at a certain epoch (e.g., just initialized), cycle slip detection for that channel is temporarily skipped.

[0059] For each successfully matched carrier phase observation Application compensation amount: in, This is the amount of hardware latency compensation. It is the wavelength of that frequency. For carrier phase observations, This is the updated wave phase observation.

[0060] Using the preprocessed carrier phase, fine-grained cycle slip detection is performed using distance-free and ionosphere-free combinations (such as the Melbourne-Wübbena combination). For observations that detect cycle slips, cycle slip markers are added to the data stream.

[0061] S42. Select a reference station, calculate the pseudorange integrated residual based on the quality weighting factor related to the adaptive smoothing weight and multipath error, and use the pseudorange integrated residual as differential correction information.

[0062] During the generation of differential correction information, it is necessary to calculate the comprehensive residual between the pseudorange observations and theoretical calculations from the reference station. This residual includes the combined effects of satellite orbital errors, satellite clock errors, ionospheric delay, tropospheric delay, and multipath errors. To obtain more reliable differential correction information, different quality weighting factors need to be assigned to the observations from different satellites and monitoring stations, so that high-precision, low-noise observations have a greater weight in subsequent fusion.

[0063] First, select a monitoring station with known coordinates and long-term stability from the monitoring station network as the reference station (e.g., the station site has been precisely measured, the antenna phase center has been calibrated, and there is little environmental interference).

[0064] For each base station i and each satellite s, calculate the combined residual of its pseudorange observations. : in, At epoch k, the adaptive smooth pseudorange of reference station i and satellite s at the reference frequency; c is the speed of light, a physical constant; The geometric distance is calculated using the known coordinates of the base station and the precise orbital products of the satellite; For satellite clock bias (from precision clock products or broadcast ephemeris). To use a model to initially estimate the ionospheric delay (which could be the Klobuchar model), These are corrections for tropospheric delay estimated using a model (which could be the Saastamoinen model), and are all conventional corrections. As a quality-weighted factor, it should be noted that... ,in, The variance of the multipath error is estimated by analyzing the sequence of pseudorange and carrier phase residuals or by using a satellite elevation angle model. For an effective satellite set.

[0065] In the estimation based on the residual sequence, the high precision of carrier phase observations is utilized to construct a combination quantity with no or weak multipath (such as MP combination, i.e., multipath combination). The time-varying sequence of this quantity reflects the pseudorange multipath error. By calculating the variance of this sequence within a certain time window, the pseudorange multipath error can be obtained. The estimated value.

[0066] When estimating using an elevation angle model, the multipath effect typically increases as the satellite elevation angle decreases. A sinusoidal function model related to the elevation angle can be used for estimation, for example: = a + b exp(-El / El0), where El is the satellite elevation angle, and a, b, and El0 are empirical model parameters.

[0067] Finally, the pseudorange integrated residual is used as differential correction information.

[0068] S43. The vertical delay of the ionospheric puncture point is calculated using the carrier phase after cycle slip detection, and an adaptive correlation distance interpolation method with the correlation length dynamically adjusted according to the regional ionospheric disturbance index is used to generate an ionospheric delay grid model.

[0069] Using the dual-frequency f1 and f2 compensated carrier phases of all monitoring stations (including the reference station and ordinary monitoring stations), the ionospheric vertical delay at the puncture point of each station-satellite connection is calculated. In the discrete When interpolating to an ionospheric delay grid covering the service area, an adaptive correlation distance interpolation method is proposed: in, For the vertical ionospheric delay of the grid points to be determined, This is the calculated vertical delay value for the nth neighboring puncture point. Let N be the projected distance between a grid point and its nth neighboring puncture point on the ground, where N is the total number of neighboring puncture points, and n is the variable index. This is the adaptive correlation length. The expression for the adaptive correlation length is: The reference length under a calm ionosphere is a preset value based on expert experience, such as 200 kilometers. The regional ionospheric disturbance index is calculated based on the carrier phase change rate of all monitoring stations in the region over the past few minutes. This is the threshold for ionospheric disturbance, typically set at 0.1–0.5 TECU / min. Reference: Studies have shown that during periods of ionospheric calm, The values ​​are typically small (e.g., about 0.02 TECU / min); however, when more pronounced ionospheric scintillation or disturbances occur, The value will increase significantly (e.g., ≥2 TECU / minute). Therefore, setting the threshold within this range can effectively distinguish between calm and disturbed states. This is an adjustment coefficient, which is determined based on expert experience, and can be set to 1.0.

[0070] For each grid point, a vertical delay estimate is calculated, along with uncertainties. This ultimately generates an ionospheric delay grid model covering the entire service area.

[0071] S44. Encode the differential correction information and ionospheric delay grid model into an enhancement information data stream. In this example, the calculated weighted average differential correction information, the generated adaptive ionospheric delay grid model, and the corresponding satellite integrity information, carrier phase ambiguity information, etc., are encoded according to a standard differential protocol (such as RTCM SSR) to generate an enhancement information data stream. This data stream is broadcast in real time to users within the service area via satellite link or mobile network.

[0072] S5. Based on the enhanced information data flow and delay compensation parameter table, a fusion positioning correction model is constructed, and the fusion positioning correction model is solved with high precision.

[0073] Furthermore, step S5 includes the following sub-steps: S51. Receives BeiDou satellite signals, demodulates to obtain the user's own original multi-frequency pseudorange carrier phase observation values, and receives the enhanced information data stream and delay compensation parameter table.

[0074] The receiver antenna receives radio frequency signals emitted by BeiDou satellites, and after demodulation, obtains the user's original multi-frequency pseudorange. And carrier phase observations. Simultaneously, it continuously receives augmented information data streams from the data processing center, and delay compensation parameter tables.

[0075] S52. Perform real-time quality assessment and spatiotemporal matching on the user's own original multi-frequency pseudorange, carrier phase observations, enhanced information data stream, and delay compensation parameter table, and mark invalid information.

[0076] Before fusion positioning is performed on the user end, all input data needs to be filtered for quality and aligned for consistency. Unreliable or inapplicable information should be removed to ensure that the data used by the subsequent fusion positioning model has sufficient timeliness, spatial matching degree and physical rationality, thereby improving positioning accuracy and integrity.

[0077] The real-time quality assessment includes timeliness verification. Timeliness verification is defined as calculating the age Δt of each piece of information by subtracting the data generation time from the current time. If the age Δt is greater than a preset value, it is marked as expired and unusable.

[0078] Spatiotemporal matching employs spatial correlation judgment, which involves interpolating the ionospheric delay correction value at the user's puncture point from the ionospheric grid model based on the user's current location. If, during interpolation, the puncture point is found to be outside the grid coverage area (e.g., beyond the outermost grid ring), the ionospheric correction information is marked as spatial mismatch (invalid) and will not be used for subsequent positioning.

[0079] Expired, incomplete, or unusable information is marked as invalid and will not participate in subsequent fusion. Expired data refers to data whose age Δt is greater than a preset value; incomplete data refers to data marked as invalid due to discontinuity or being outside the physical allowable range of pseudorange and carrier phase observations during the check in step S22, as well as data packets lacking necessary fields (such as monitoring station ID, satellite PRN, frequency, etc.); unusable data refers to data that has been explicitly marked as "invalid" during the preprocessing or enhancement information generation stage, such as observations rejected due to gross errors in step S22, and carrier phases with cycle slips detected in step S41. These marks will be transmitted with the data stream, and the user terminal will directly reject them accordingly.

[0080] S53. For information not marked as invalid, calculate the dynamic fusion confidence weight based on information age, nominal precision, and multi-source consistency variance. The formula for calculating the dynamic fusion confidence weight is: in, To integrate confidence weights, j represents the information type, and s represents the satellite. The current age of the j-th type of information is Δt; the larger Δt is, the older the information. For the j-th type of information, the nominal accuracy index for satellite s is derived from the augmented information data stream for differential correction and from the uncertainty in the compensation parameter table for hardware delay compensation. The consistency variance of multi-source information is obtained by performing short-time statistical comparisons of estimates of the same error term from different sources. and This is an adjustable empirical coefficient used to balance the effects of the accuracy and consistency terms, as well as to control the rate of aging decay. It can take the value 0.7. The possible value is 0.05.

[0081] S54. The differential correction information and dynamic delay compensation parameters are fused into the observation equation according to their weights to construct a fused positioning correction model and solve for the user's high-precision three-dimensional position coordinates. At the same time, the weight distribution is recorded as a basis for quality assessment.

[0082] Specifically, the expression for the fusion localization correction model is: in, The raw pseudorange observations of the user receiver itself. The specific correction amount for the j-th type of correction information applied to this satellite and frequency includes hardware delay compensation and differential correction; The geometric distance from the user's location to satellite s; For the user receiver clock bias; For satellite clock bias; is the ionospheric delay factor at frequency f, which is dimensionless; For tropospheric delay correction; To observe noise.

[0083] The weighted least squares method or extended Kalman filter is used to solve the fusion positioning correction model to obtain high-precision three-dimensional position coordinates and receiver clock error. Simultaneously, the distribution of various dynamic fusion confidence weights during the fusion process is recorded as a dimension for evaluating the quality of the positioning results.

[0084] In this embodiment, step S5 further includes step S55. While the user terminal completes the high-precision positioning calculation, multi-layer integrity detection is performed, and the positioning results are marked and alarmed in real time.

[0085] Specifically, it receives high-precision three-dimensional position coordinates and their fused confidence weights, enhanced information data streams, and its own redundant observation information, and performs multi-layer integrity checks on the above information. This includes: The first-level data consistency check is performed to examine information inconsistencies between different correction channels and whether the observation residuals exceed limits. Specifically, within the fused correction model, it checks whether there are inconsistencies in the same type of information provided by different correction channels. For the same satellite, the same type of correction information may be obtained from different sources. The difference between the two is calculated. If the difference exceeds a preset threshold, an information inconsistency is determined, triggering the first-level alarm. The user observation residuals are then checked to see if they exceed the physically reasonable range predicted based on the current satellite geometry and model accuracy. Using the currently calculated user and satellite positions, the carrier phase / pseudorange observation residuals for each satellite are calculated. Based on the current satellite geometry and model nominal accuracy, a reasonable range for the residuals is predicted. If the absolute value of a satellite's residual exceeds this range, the satellite is marked as potentially faulty, triggering the first-level alarm.

[0086] A second-layer external integrity comparison is performed, comparing the user-calculated error boundary with the protection level in the enhanced information data stream. If the boundary exceeds the limit, a second-layer alarm is triggered. The estimated position error calculated by the user is compared with the integrity information in the enhanced information data stream. If the user-calculated error boundary exceeds the externally provided protection level, a second-layer alarm is triggered.

[0087] A third-layer internal redundancy check is performed by sequentially removing satellites and recalculating the position to observe for any abnormalities in position changes. In this step, the user's position is recalculated after removing each satellite, resulting in a set of position solutions with the removed satellites. The difference between the position calculation results before and after removal (e.g., horizontal deviation, 3D deviation) is calculated. If the difference exceeds a preset threshold related to the current positioning accuracy and protection level, it is considered "abnormal," triggering a third-layer warning. This third-layer check can improve the protection level in the self-reinforcing information data stream, the system's nominal accuracy, or the statistically based fault detection threshold.

[0088] Based on the combined results of the three layers of detection, an alarm at any layer triggers an audible and visual alarm, records a fault log, and switches to a low-precision positioning mode. If any of the first, second, or third layers generates an alarm, the system determines the positioning result to be unreliable. This three-layer progressive integrity check allows for multi-dimensional evaluation of the positioning result's reliability, considering data consistency, external comparison, and internal redundancy. This prevents erroneous high-precision locations from being provided to safety-critical applications, enhancing the system's robustness and security. It's important to note that the low-precision positioning mode refers to Standard Single Point Positioning (SPP), with accuracy ranging from meters to tens of meters. SPP is a mature and fundamental technology in satellite navigation. This mode serves as a safety degradation solution when the system detects unreliable high-precision positioning results. Its purpose is to ensure that the system can still provide basic positioning services after a failure, rather than becoming completely ineffective, thus guaranteeing system availability and security.

[0089] In summary, this method improves the accuracy of initial coordinate calculation by utilizing raw observation data collected from a monitoring network and performing data preprocessing and adaptive smoothing. Secondly, it employs a state estimation algorithm to estimate the time-varying hardware delay within the receiver in real time, generating dynamic compensation parameters to eliminate systematic biases at their source. Then, based on the compensated data, cycle slip detection and gross error removal are performed to generate high-quality differential correction information. At the user end, the differential correction and dynamic delay compensation parameters are received, and a fusion positioning correction model is constructed to achieve precise error correction of the initial coordinates. Finally, while performing high-precision positioning calculations, multi-layer integrity checks are performed in parallel, and the positioning results are marked and alarmed in real time. This effectively solves the dilemma of balancing accuracy, robustness, and reliability in existing technologies, significantly improving the availability and reliability of BeiDou high-precision services in complex environments.

[0090] Example 2 This embodiment provides a multi-frequency high-precision positioning system based on BeiDou satellite signals, which is applied to the multi-frequency high-precision positioning method based on BeiDou satellite signals in Embodiment 1, and includes: The data acquisition module is used to collect raw observation data from BeiDou navigation satellites through a network of monitoring stations. The preprocessing module is used to preprocess the raw observation data and calculate the initial coarse coordinates of each monitoring station in the monitoring station network; The first compensation module is used to estimate the different frequency signals inside the receivers of all monitoring stations in real time based on the preprocessed data using a state estimation algorithm, generate dynamic delay compensation parameters, and encapsulate the dynamic delay compensation parameters into a delay compensation parameter table. The second compensation module is used to perform cycle slip detection on the observation values ​​of each satellite at each monitoring station based on the preprocessed data and the delay compensation parameter table, generate differential correction information, encode the differential correction information, and generate an enhanced information data stream. The correction module is used to construct a fusion positioning correction model based on the enhanced information data stream and the delay compensation parameter table, and to perform high-precision calculation on the fusion positioning correction model.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A multi-frequency high-precision positioning method based on BeiDou satellite signals, characterized in that, Includes the following steps: S1. Collect raw observation data of BeiDou navigation satellites through a network of monitoring stations; S2. Preprocess the raw observation data and calculate the initial coarse coordinates of each monitoring station in the monitoring station network; S3. Based on the preprocessed data, a state estimation algorithm is used to estimate the different frequency signals inside the receivers of all monitoring stations in real time, generate dynamic delay compensation parameters, and encapsulate the dynamic delay compensation parameters to form a delay compensation parameter table. S4. Based on the preprocessed data and the delay compensation parameter table, cycle slip detection is performed on the observation values ​​of each satellite at each monitoring station to generate differential correction information. Based on the differential correction information, encoding is performed to generate an enhanced information data stream. S5. Based on the enhanced information data flow and delay compensation parameter table, a fusion positioning correction model is constructed, and the fusion positioning correction model is solved with high precision.

2. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 1, characterized in that, Step S2 includes the following steps: The raw observation data are synchronized to the same main time scale and uniformly converted into a standard format to form standardized observation data; The pseudorange and carrier phase observations in the standardized observation data are checked, gross errors exceeding the threshold are removed, and the remaining observations are marked as valid observations. For each monitoring station, satellite, and signal frequency, an adaptive weighted recursive smoothing formula based on the satellite elevation angle is used to smooth the effective pseudorange observations every epoch. If the instantaneous residual exceeds the dynamic threshold during the smoothing process, the smoothing is paused. Using the smoothed pseudorange observations of each signal frequency and the broadcast ephemeris, the initial coarse coordinates of the monitoring station in the Earth-fixed coordinate system and the receiver clock error are calculated.

3. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 2, characterized in that, Step S3 includes the following steps: Based on the preprocessed data, the effective carrier phase observation values ​​of each monitoring station for each visible BeiDou satellite are extracted. Construct geometrically indistinguishable combined observations using the frequencies corresponding to the effective carrier phase observations; The differential hardware delay of each monitoring station is decomposed into constant bias and time-varying components and estimated by using state-adaptive Kalman filtering. The estimated constant bias and time-varying components are combined to form the total differential hardware delay estimate, and the hardware delay compensation amount for each frequency is allocated according to the frequency ratio. The compensation amount and the corresponding uncertainty are encapsulated into a delay compensation parameter table.

4. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 1, characterized in that, Step S4 includes the following steps: The system receives preprocessed data and a delay compensation parameter table, and after associating and correcting the delay compensation parameter table with the carrier phase observation values ​​of the corresponding monitoring station and frequency, it performs cycle slip detection and marking. A reference station is selected, and the pseudorange integrated residual is calculated based on the quality weighting factor related to the adaptive smoothing weight and multipath error. The pseudorange integrated residual is used as differential correction information. The vertical delay of the ionospheric puncture point is calculated using the carrier phase after cycle slip detection, and an adaptive correlation distance interpolation method with the correlation length dynamically adjusted according to the regional ionospheric disturbance index is used to generate an ionospheric delay grid model. The differential correction information and the ionospheric delay grid model are encoded into an enhanced information data stream.

5. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 1, characterized in that, Step S5 includes: It receives BeiDou satellite signals, demodulates to obtain the user's own original multi-frequency pseudorange carrier phase observation values, and receives the enhanced information data stream and delay compensation parameter table; Real-time quality assessment and spatiotemporal matching are performed on the user's original multi-frequency pseudorange, carrier phase observations, enhanced information data stream, and delay compensation parameter table, and invalid information is marked. For information that is not marked as invalid, a dynamic fusion confidence weight is calculated based on the information age, nominal precision, and multi-source consistency variance. The differential correction information and dynamic delay compensation parameters are fused into the observation equation according to the dynamic fusion confidence weights to construct a fusion positioning correction model and solve for the user's high-precision three-dimensional position coordinates. At the same time, the weight distribution is recorded as a basis for quality assessment.

6. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 5, characterized in that, Step S5 also includes: while completing the high-precision positioning calculation at the user end, performing multi-layer integrity detection, and marking and alarming the positioning results in real time.

7. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 6, characterized in that, The multi-layer integrity detection includes: It receives high-precision three-dimensional position coordinates, dynamically fuses confidence weights, enhances information data streams, and includes redundant observation information from the user's own observations. Perform the first-level data consistency check to check for information inconsistencies in different correction channels and whether the residuals of observed values ​​exceed the limits; Perform a second-layer external integrity comparison, compare the error boundary calculated by the user with the protection level in the enhanced information data stream, and determine whether the limit is exceeded; Perform a third-level internal redundancy check by sequentially removing satellites and recalculating to observe whether the position changes are abnormal. Based on the combined detection results from the three layers, an alarm at any layer will trigger an audible and visual alarm, record a fault log, and switch to a low-precision positioning mode.

8. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 2, characterized in that, The adaptive weighted recursive smoothing formula is: in, The adaptive smooth pseudo-range at epoch k-1 This represents the change in the carrier phase observation value at the same frequency between two adjacent epochs. The original pseudorange observation value at epoch k; For epoch k, the dynamic smoothing weight of the i-th monitoring station for satellite s, At epoch k, the carrier phase of satellite s; i is the monitoring station number, f is the frequency number, and s is the satellite number.

9. The multi-frequency high-precision positioning method based on BeiDou satellite signals according to claim 3, characterized in that, The expression for the observations without geometric distance combination is: in, At epoch k, the geometrically inverse combination of observations from monitoring station i and satellite s; and These are the original carrier phase observations of monitoring station i and satellite s at frequencies f1 and f2 at epoch k, respectively; f1 and f2 are the two frequencies of the BeiDou satellite signal.

10. A multi-frequency high-precision positioning system based on BeiDou satellite signals, characterized in that, The multi-frequency high-precision positioning method based on BeiDou satellite signals, applied to any one of claims 1 to 9, includes: The data acquisition module is used to collect raw observation data of BeiDou navigation satellites through the monitoring station network; The preprocessing module is used to preprocess the raw observation data and calculate the initial coarse coordinates of each monitoring station in the monitoring station network; The first compensation module is used to estimate the different frequency signals inside the receivers of all monitoring stations in real time based on the preprocessed data using a state estimation algorithm, generate dynamic delay compensation parameters, and encapsulate the dynamic delay compensation parameters into a delay compensation parameter table. The second compensation module is used to perform cycle slip detection on the observation values ​​of each satellite at each monitoring station based on the preprocessed data and the delay compensation parameter table, generate differential correction information, encode the differential correction information, and generate an enhanced information data stream. The correction module is used to construct a fusion positioning correction model based on the enhanced information data stream and the delay compensation parameter table, and to perform high-precision calculation on the fusion positioning correction model.

Images