Rapid convergence satellite clock error resolving method under regional station distribution

Abstract

The invention relates to a fast convergence satellite clock error resolving method under regional station distribution, which belongs to the field of satellite navigation, and comprises the following steps: in a regional monitoring network, constructing a normal equation by using an ionosphere-free combination pseudo range and a carrier phase by using a known precise orbit and a station coordinate; in the Kalman filtering stage, covariance forecasting is carried out on satellite clock correction, receiver clock correction and troposphere delay, cycle slip detection and covariance forecasting are carried out on ambiguity, when the ambiguity has no cycle slip and converges, the forecasting covariance is reset to be a minimum value so as to apply reinforced constraint, and the reinitialization time is remarkably compressed; performing parameter estimation to obtain an optimal value and a pseudo-range post-test residual error; the satellite clock error constant deviation is calculated through multi-station multi-epoch residual errors in a combined mode, and the satellite clock error constant deviation is fused into the original code deviation to form extended code deviation parameter broadcasting; the user side only needs to add the extension correction in the pseudo-range observation value to realize instantaneous centimeter-level positioning. According to the invention, high-precision and high-availability real-time precise single-point positioning service under the regional network condition is realized.

Description

Technical Field

[0001] This invention relates to the fields of satellite navigation, aerospace measurement and control, and in particular to a method for rapidly converging satellite clock bias calculation under regional station deployment. Background Technology

[0002] Satellite navigation systems provide continuous, real-time, and precise location and time information globally, and have been deeply integrated into numerous fields such as autonomous driving, precision agriculture, and mass consumer products. With the rapid growth in demand for real-time, high-precision positioning from emerging applications such as intelligent transportation and unmanned systems, Real-Time Precise Point Positioning (RT-PPP) services, capable of providing dynamic decimeter-level and static centimeter-level positioning accuracy, have become a core direction for the construction and application of global satellite navigation systems.

[0003] Real-time precise satellite clock bias is one of the prerequisite parameters for RT-PPP. Affected by factors such as high-frequency noise from onboard atomic clocks, temperature transients, and orbital dynamics errors, satellite clock bias exhibits significant random walk characteristics. It is typically estimated using recursive estimation algorithms such as Kalman filtering, relying on a globally uniformly deployed monitoring network for continuous tracking and real-time estimation. However, global deployment requires a large number of overseas sites, resulting in extremely high construction and maintenance costs. Most PPP service operators cannot independently control global observation data and can only rely on a posteriori or delayed products provided by a few international institutions, leading to insufficient service autonomy, real-time performance, and security.

[0004] To reduce reliance on global monitoring networks, the regional monitoring network (GMN) model has become a research hotspot in recent years. This model deploys only a limited number of monitoring stations within a country or its surrounding area, achieving real-time estimation of satellite clock bias through localized observation arcs. However, the limited coverage of regional networks means that a single satellite can only be continuously observed for 4–6 hours per day, with the remaining time spent in an intermittent "entry and exit" state, leading to the following prominent problems: Frequent initialization of filters: The clock filter needs to be reinitialized every time a satellite re-enters the regional network's field of view. Traditional random walk models take 3-5 hours to converge to centimeter-level accuracy, while satellites usually leave the area before convergence, resulting in almost zero effective service time.

[0005] Convergence and availability are contradictory: Increasing process noise to shorten convergence time can speed up parameter adjustment, but it will significantly reduce clock error stability; conversely, convergence will be slow and cannot meet the rigid requirement of real-time PPP for "instant availability".

[0006] Indistinguishable constant bias: The weak geometric strength of the regional network observations leads to absorption-coupling between clock bias and hardware delay (DCB), resulting in a systematic constant bias in the estimated clock bias. If this bias cannot be calibrated and broadcast in real time, it will be directly introduced into the user positioning equation, causing a systematic error of several decimeters to several meters.

[0007] Existing approaches to address these challenges either rely on high-frequency epoch differential speed-up, increase redundancy through global station networks, or depend on real-time streaming multinomial forecasts for precision. However, they generally overlook the fundamental pain points of short regional monitoring network arcs, frequent satellite entry and exit, repeated filter initialization, long convergence time, and difficulty in real-time calibration and broadcasting of constant deviations. As a result, high-precision real-time clock difference services in regional environments still cannot simultaneously achieve accuracy, availability, and autonomy.

[0008] In summary, existing technologies can operate well under global network conditions, but they generally suffer from three major bottlenecks under regional deployment modes: excessively long convergence time, short available time periods, and difficulty in real-time calibration of constant deviations. These limitations prevent them from simultaneously ensuring the accuracy, availability, and autonomy of real-time PPP services. Therefore, there is an urgent need for a real-time satellite clock bias calculation method for regional monitoring networks. This method should be able to achieve convergence within minutes after satellite arrival and simultaneously estimate and broadcast clock bias constant deviations to realize high-precision, highly available real-time precise point positioning services under regional network conditions. Summary of the Invention

[0009] Based on the above analysis, this invention aims to disclose a real-time satellite clock bias calculation method for regional monitoring networks; it can achieve convergence within minutes after satellite arrival and simultaneously estimate the deviation from the broadcast clock bias constant, so as to realize high-precision, high-availability real-time precise single-point positioning service under regional network conditions.

[0010] This invention discloses a real-time satellite clock bias calculation method for regional monitoring networks, comprising: S1. Under the condition of regional monitoring stations, based on the precise orbit of GNSS satellites and the precise coordinates of monitoring stations, the normal equations are constructed using pseudorange and carrier phase observations without ionospheric combination. S2. Perform Kalman filter state prediction, including covariance prediction for satellite clock error, receiver clock error, and tropospheric delay parameters, and cycle slip detection and covariance prediction for ambiguity parameters; when the ambiguity parameter has not experienced cycle slip and has converged to a suboptimal solution, its predicted covariance is reset to apply stronger constraints. S3. Based on the normal equation and state prediction results, perform parameter estimation to obtain the optimal estimated values ​​of each parameter and the pseudorange post-hoc residuals; S4. Calculate the satellite clock error constant deviation using the pseudorange verification residuals of the same satellite at multiple epochs using multiple regional monitoring stations, and integrate the clock error constant deviation into the satellite code deviation parameter to form the extended code deviation parameter and broadcast it. S5. The user terminal applies corrections to the pseudorange and carrier phase observations, wherein the correction of the pseudorange observations includes the correction of the spreading code deviation parameter.

[0011] Furthermore, in S1, the process of constructing the normal equations includes: S101. Using the precise orbit of the GNSS satellite and the precise station coordinates of the ground monitoring station receiver as input, calculate the pseudorange and phase data residuals of the satellite for the regional monitoring receiver; S102. Under the least squares framework, the observation equation is constructed by designing the coefficient matrix and constant vector, and a parameter system to be estimated, including satellite clock error, receiver clock error, ambiguity parameter and tropospheric parameter, is established. Based on the observation equation and the observation value weight matrix, the normal equation matrix and the right matrix of the normal equation are constructed, and they are combined to form the normal equation serving the fast convergence and clock error estimation of the regional network.

[0012] Furthermore, S101 includes: 1) Establish the current epoch Ionospheric combination equations for GNSS satellite pseudorange and carrier phase observation data from regional ground monitoring receivers: ; In the formula, , These are GNSS satellites and ground monitoring receivers, respectively. and These are respectively the pseudorange and carrier phase observations of GNSS satellites by regional ground monitoring receivers; The geometric distance between the GNSS satellite and the ground monitoring receiver is calculated using the precise orbit of the GNSS satellite and the precise station coordinates of the ground monitoring station receiver as input. and These are the clock biases of the ground monitoring receiver and the clock biases of the GNSS satellite. The speed of light in a vacuum. The projection function of tropospheric delay. To calculate the tropospheric delay in the zenith direction using an empirical model, This is a delay in the zenith-direction troposphere; For satellite wavelength, For phase ambiguity, and These are pseudorange measurement error and phase measurement error, respectively. 2) Calculate the pseudorange and phase data residuals of the regional monitoring receiver for the satellite based on the ionospheric-free combination equation; ; in, and These represent the pseudo-range non-ionospheric combination residual and the phase non-ionospheric combination residual, respectively.

[0013] Furthermore, S102 includes: 1) Construct the observation equation; ; in, for Time residual vector , To design the coefficient matrix, Let be the state vector to be estimated. It is a constant vector; 2) Construct the normal equation matrix and the right normal equation matrix based on the observation equation and the observation weight matrix; Wherein, the normal equation matrix for: ; Right matrix of normal equations for: ; in, It is the observation weight matrix; 3) The normal equations are constructed by combining the normal equation matrix and the right-hand matrix of the normal equations. Solving for the normal equations yields the optimal parameter estimate. ; The normal equation is: .

[0014] Further, S2 includes: S201. Establish a random walk prediction model; use the identity matrix as the state transition matrix, and directly inherit the optimal estimate from the previous epoch as the prediction value for the current epoch; simultaneously transfer the covariance matrix from the previous epoch to the current epoch, and superimpose the process noise matrix to quantify the error accumulation caused by model uncertainty. The diagonal elements of the forecast covariance matrix represent the forecast variances of receiver clock error, satellite clock error, zenith tropospheric delay, and phase ambiguity, while the off-diagonal elements represent the cross-covariances between the diagonal elements, thus comprehensively characterizing the uncertainty structure of multi-parameter coupling.

[0015] S202. Implement fixed incremental covariance expansion for three types of parameters: satellite clock bias, receiver clock bias, and zenith tropospheric delay. S203. Perform cycle slip detection and convergence judgment on the phase ambiguity parameter. If either the condition of "cycle slip has occurred" or "converged" is met, immediately reset the current epoch ambiguity prediction variance to the minimum value and apply a pseudo-observation strong constraint. If no cycle slip has occurred and convergence has not occurred, maintain the original variance level and continue filtering. S204. The covariance elements of each sub-block processed by S202 and S203 are integrated into a unified matrix according to the parameter order to generate a complete forecast covariance matrix, which serves as the input for parameter estimation in the next step, realizing rapid convergence and high-precision maintenance of clock error calculation in the regional monitoring network environment.

[0016] Furthermore, in S202, a fixed incremental covariance expansion is applied to three types of parameters: satellite clock bias, receiver clock bias, and zenith tropospheric delay; including: ; in, The current epoch receiver clock bias prediction variance, This represents the receiver clock bias of the previous epoch; The current epoch satellite clock error prediction variance, The satellite clock bias for the previous epoch; The variance of the zenith tropospheric delay forecast for the current epoch. For the zenith troposphere delay of the previous epoch, The time difference between epochs, The noise intensity coefficient for the clock bias random walk process. This represents the noise intensity coefficient during the random walk process in the troposphere.

[0017] Furthermore, in S203, cycle slip detection and convergence judgment are performed on the phase ambiguity parameter; From the era Towards Cycle slip detection; if a cycle slip occurs, variance is reset. If cycle slips occur, the covariance prediction of the ambiguity parameters is as follows: ; The variance of phase ambiguity; For the phase ambiguity prediction variance, As the first empirical threshold, select one that is close to the upper limit of the variance of the entire search area. The second empirical threshold is selected to be slightly larger than the lower limit of the variance of the measurement noise.

[0018] Furthermore, in S3, parameter estimation is performed to obtain the optimal solution for parameter estimation: ; in, The normal equation matrix, The right matrix of the normal equation, Let covariance matrix be the variance matrix. This yields the optimal estimate of the parameters; pseudo-distance post-test residual The calculation method is as follows: ; in, These are the original pseudorange observations without an ionosphere. To calculate the geometric distance using prior precise orbits and station coordinates, This is the projection function of the tropospheric delay; To calculate the tropospheric delay in the zenith direction using an empirical model, The zenith retardation is the residual tropospheric delay, and c is the speed of light in a vacuum. For satellite clock bias, This refers to the receiver clock bias.

[0019] Furthermore, in S4, the spreading code deviation parameter for: ; in, The actual inter-symmetric deviation of satellites due to hardware delay; The clock constant deviation calculated for the server; ; in, To participate in satellite The total number of ground monitoring stations observed For each station, satellite In continuous The number of samples of effective pseudorange residuals within each epoch.

[0020] Furthermore, in S5, precise satellite orbit, group delay parameters, and precise clock error parameters are corrected for pseudorange observations; ; in, This is the corrected pseudorange; These are the original pseudorange observations; For track corrections; For precision satellite clock bias; Group delay parameter; Precise satellite orbit and precise clock error parameters are applied to the phase observations; ; in, For the corrected phase; These are the original phase observations.

[0021] This invention can achieve one of the following beneficial effects: 1. A model, method, and process for real-time filtering and calculation of satellite clock bias under the condition of regional observation stations are proposed to achieve high-precision clock bias estimation within the coverage area of ​​observation data; 2. A method for accelerating the convergence of real-time filtering estimation of navigation satellite clock bias under the case of regional monitoring stations with strong constraints on ambiguity parameters is proposed, which accelerates the convergence process of real-time filtering clock bias under the case of regional monitoring stations and improves the accuracy and availability of real-time clock bias. 3. A method for calculating satellite clock bias in the case of regional monitoring stations is proposed, and the idea of ​​absorbing the bias using code bias parameters is suggested. This establishes an application mode where the pseudorange and carrier phase differ, thereby reducing satellite clock bias errors. The proposed design can balance the accuracy, convergence performance, and service availability of precise point positioning services. Attached Figure Description

[0022] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Figure 1 This is a flowchart of a real-time satellite clock bias calculation method for regional monitoring networks according to an embodiment of the present invention. Detailed Implementation

[0023] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and, together with the embodiments of the present invention, serve to illustrate the principles of the present invention.

[0024] One embodiment of the present invention discloses a real-time satellite clock bias calculation method for regional monitoring networks, such as... Figure 1 As shown, it includes: S1. Under the condition of regional monitoring stations, based on the precise orbit of GNSS satellites and the precise coordinates of monitoring stations, the normal equations are constructed using pseudorange and carrier phase observations without ionospheric combination. S2. Perform Kalman filter state prediction, including covariance prediction for satellite clock error, receiver clock error, and tropospheric delay parameters, and cycle slip detection and covariance prediction for ambiguity parameters; when the ambiguity parameter has not experienced cycle slip and has converged to a suboptimal solution, its predicted covariance is reset to apply stronger constraints. S3. Based on the normal equation and state prediction results, perform parameter estimation to obtain the optimal estimated values ​​of each parameter and the pseudorange post-hoc residuals; S4. Calculate the satellite clock error constant deviation using the pseudorange verification residuals of the same satellite at multiple epochs using multiple regional monitoring stations, and integrate the clock error constant deviation into the satellite code deviation parameter to form the extended code deviation parameter and broadcast it. S5. The user terminal applies corrections to the pseudorange and carrier phase observations, wherein the correction of the pseudorange observations includes the correction of the spreading code deviation parameter, while the correction of the spreading code deviation parameter is not applied to the carrier phase observations.

[0025] Specifically, in S1, the normal equation construction process includes: S101. Using the precise orbit of the GNSS satellite and the precise station coordinates of the ground monitoring station receiver as input, calculate the pseudorange and phase data residuals of the satellite for the regional monitoring receiver; S102. Under the least squares framework, the observation equation is constructed by designing the coefficient matrix and constant vector, and a parameter system to be estimated, including satellite clock error, receiver clock error, ambiguity parameter and tropospheric parameter, is established. Based on the observation equation and the observation value weight matrix, the normal equation matrix and the right matrix of the normal equation are constructed, and they are combined to form the normal equation serving the fast convergence and clock error estimation of the regional network.

[0026] Specifically, S101 includes: 1) Establish the current epoch Ionospheric combination equations for GNSS satellite pseudorange and carrier phase observation data from regional ground monitoring receivers: ; In the formula, , These are GNSS satellites and ground monitoring receivers, respectively. and These are respectively the pseudorange and carrier phase observations of GNSS satellites by regional ground monitoring receivers; The geometric distance between the GNSS satellite and the ground monitoring receiver is calculated using the precise orbit of the GNSS satellite and the precise station coordinates of the ground monitoring station receiver as input. and These are the clock biases of the ground monitoring receiver and the clock biases of the GNSS satellite. The speed of light in a vacuum. The projection function of tropospheric delay. To calculate the tropospheric delay in the zenith direction using an empirical model, This is a delay in the zenith-direction troposphere; For satellite wavelength, For phase ambiguity, and These are pseudorange measurement error and phase measurement error, respectively. 2) Calculate the pseudorange and phase data residuals of the regional monitoring receiver for the satellite based on the ionospheric-free combination equation; ; in, and These represent the pseudo-range non-ionospheric combination residual and the phase non-ionospheric combination residual, respectively.

[0027] Specifically, S102 includes: 1) Construct the observation equation; ; in, for Time residual vector , To design the coefficient matrix, Let be the state vector to be estimated. It is a constant vector; ; 2) Construct the normal equation matrix and the right normal equation matrix based on the observation equation and the observation weight matrix; Wherein, the normal equation matrix for: ; Function: Its inverse matrix is ​​the covariance matrix of the parameter estimation, which reflects the reliability and correlation of the parameter solution; Right matrix of normal equations for: ; in, The observation weight matrix is ​​obtained by first assigning weights to the pseudorange and phase according to the satellite elevation angle model, and then iteratively correcting the residuals after inspection using a robust scheme, finally obtaining the diagonal weight matrix.

[0028] 3) The normal equations are constructed by combining the normal equation matrix and the right-hand matrix of the normal equations. Solving for the normal equations yields the optimal parameter estimate. ; The normal equation is: .

[0029] Specifically, S2 includes: S201. Establish a random walk prediction model; use the identity matrix as the state transition matrix, and directly inherit the optimal estimate from the previous epoch as the prediction value for the current epoch; simultaneously transfer the covariance matrix from the previous epoch to the current epoch, and superimpose the process noise matrix to quantify the error accumulation caused by model uncertainty. When the state transition matrix is ​​an identity matrix, the prediction equation for the parameter to be estimated is: ; ; in, For the current epoch The forecast covariance matrix, For the previous era The estimated covariance matrix, This is process noise; ; ; Among them, the forecast covariance matrix The diagonal elements are the prediction variances of receiver clock bias, satellite clock bias, zenith tropospheric delay, and phase ambiguity, respectively, while the off-diagonal elements are the cross-variances between the diagonal elements; this fully characterizes the uncertainty structure of multi-parameter coupling.

[0030] Diagonal elements (variance) Receiver clock bias prediction variance, reflecting the uncertainty of receiver clock drift (unit: m²). Satellite clock bias prediction variance reflects the uncertainty of satellite atomic clock frequency drift (unit: m²). : Variance of zenith tropospheric delay forecast, reflecting the time-varying uncertainty of the atmosphere (unit: m²) Phase ambiguity prediction variance reflects the time-varying uncertainty of ambiguity (unit: weeks² or m²). Off-diagonal elements (covariance) : Predicted covariance between receiver clock bias and satellite clock bias (unit: m²) Forecast covariance of receiver clock bias and tropospheric delay (unit: m²) Forecast covariance of satellite clock bias and tropospheric delay (unit: m²) Receiver clock bias and ambiguity prediction covariance (unit: m·week) Predictive covariance of satellite clock bias and ambiguity (Note: should be in m·week) Forecast covariance of ambiguity and tropospheric delay (unit: weeks·m) S202. Implement fixed incremental covariance expansion for three types of parameters: satellite clock bias, receiver clock bias, and zenith tropospheric delay. include: ; ; in, The current epoch receiver clock bias prediction variance, This represents the receiver clock bias of the previous epoch; The current epoch satellite clock error prediction variance, The satellite clock bias for the previous epoch; The variance of the zenith tropospheric delay forecast for the current epoch. For the zenith troposphere delay of the previous epoch, The time difference between epochs, The noise intensity coefficient for the clock bias random walk process. This represents the noise intensity coefficient during the random walk process in the troposphere.

[0031] Preferred, , .

[0032] S203. Perform cycle slip detection and convergence judgment on the phase ambiguity parameter. If either the condition of "cycle slip has occurred" or "converged" is met, immediately reset the current epoch ambiguity prediction variance to the minimum value and apply a pseudo-observation strong constraint. If no cycle slip has occurred and convergence has not occurred, maintain the original variance level and continue filtering. In S203, cycle slip detection and convergence judgment are performed on the phase ambiguity parameter; From the era Towards Cycle slip detection; if a cycle slip occurs, variance is reset. If cycle slips occur, the covariance prediction of the ambiguity parameters is as follows: ; The variance of phase ambiguity; For the phase ambiguity prediction variance, As the first empirical threshold, select one that is close to the upper limit of the variance of the entire search area. The second empirical threshold is selected to be slightly larger than the lower limit of the variance of the measurement noise.

[0033] Preferred, , .

[0034] S204. The covariance elements of each sub-block processed by S202 and S203 are integrated into a unified matrix according to the parameter order to generate a complete forecast covariance matrix, which serves as the input for parameter estimation in the next step, realizing rapid convergence and high-precision maintenance of clock error calculation in the regional monitoring network environment.

[0035] Specifically, in S3, parameter estimation is performed to obtain the optimal solution for parameter estimation: ; in, The normal equation matrix, The right matrix of the normal equation, Let covariance matrix be the variance matrix. This yields the optimal estimate of the parameters; pseudo-distance post-test residual The calculation method is as follows: ; in, These are the original pseudorange observations without an ionosphere. To calculate the geometric distance using prior precise orbits and station coordinates, This is the projection function of the tropospheric delay; For the residual zenith-direction tropospheric delay, For satellite clock bias, This refers to the receiver clock bias.

[0036] Among them, the residual zenith-direction tropospheric delay Satellite clock bias and receiver clock difference The result is estimated for step S2.

[0037] Specifically, S4, using a polynomial method, the satellite clock error constant deviation is calculated by using the pseudorange verification residuals of the same satellite at multiple epochs from multiple regional monitoring stations. The clock error constant deviation is then fused into the satellite code deviation parameter to form the extended code deviation parameter, which is then broadcast to the user.

[0038] Spread code deviation parameter (Unit: meters, resolution 0.01 meters): ; in, The actual inter-symmetric deviation of satellites due to hardware delay; The clock constant deviation calculated for the server; ; in, To participate in satellite The total number of ground monitoring stations observed For each station, satellite In continuous The number of valid pseudorange residuals within each epoch; all available epochs within a single satellite transit arc.

[0039] The precise satellite clock bias parameters estimated in S2 The expansion code deviation parameter estimated in this step is sent to the user.

[0040] In S5, after receiving the real-time precise point positioning message, which includes precise satellite clock bias parameters and spreading code deviation parameters, the user performs positioning processing; among which, Precise satellite orbit, group delay parameters, and precise clock error parameters are corrected for pseudorange observations; ; in, This is the corrected pseudorange; These are the original pseudorange observations; For track corrections; For precision satellite clock bias; Group delay parameter; Precise satellite orbit and precise clock error parameters are applied to the phase observations; ; in, For the corrected phase; These are the original phase observations.

[0041] In summary, this embodiment can achieve the following effects: 1. A model, method, and process for real-time filtering and calculation of satellite clock bias under the condition of regional observation stations are proposed to achieve high-precision clock bias estimation within the coverage area of ​​observation data; 2. A method for accelerating the convergence of real-time filtering estimation of navigation satellite clock bias under the case of regional monitoring stations with strong constraints on ambiguity parameters is proposed, which accelerates the convergence process of real-time filtering clock bias under the case of regional monitoring stations and improves the accuracy and availability of real-time clock bias. 3. A method for calculating satellite clock bias in the case of regional monitoring stations is proposed, and the idea of ​​absorbing the bias using code bias parameters is suggested. This establishes an application mode where the pseudorange and carrier phase differ, thereby reducing satellite clock bias errors. The proposed design can balance the accuracy, convergence performance, and service availability of precise point positioning services.

[0042] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A satellite clock error solving method with fast convergence under regional station distribution, characterized in that, Comprise: S1, under the condition of regional monitoring station, based on GNSS satellite precise orbit and monitoring station accurate coordinate, using ionosphere-free combination of pseudorange and carrier phase observation value to construct normal equation; S2, Kalman filter state prediction, wherein the satellite clock error, receiver clock error, troposphere delay parameter covariance prediction, ambiguity parameter for cycle slip detection and covariance prediction; wherein, when the ambiguity parameter has not occurred cycle slip and has converged to the suboptimal solution, its prediction covariance is reset to impose strong constraints; S3, based on the normal equation and state prediction results for parameter estimation, obtain the optimal estimation value and pseudorange posteriori residual of each parameter; S4, using multiple regional monitoring stations to calculate satellite clock error constant bias for the same satellite at multiple epochs, and fusing the clock error constant bias into satellite code bias parameter to form extended code bias parameter and broadcast; S5, user side to pseudorange and carrier phase observation value to impose correction, wherein the pseudorange observation value to impose correction contains the extended code bias parameter correction.

2. The regional station layout rapid convergence satellite clock error solving method according to claim 1, characterized in that, S1, the normal equation construction process comprises: S101, taking GNSS satellite precise orbit and ground monitoring station receiver accurate station coordinates as input, calculating the pseudorange and phase data residual of the satellite for the regional monitoring receiver; S102, under the least square framework, the observation equation is constructed by designing the coefficient matrix and constant vector, and the parameter system containing satellite clock error, receiver clock error, ambiguity parameter and troposphere parameter is established; the normal equation matrix and the right matrix of the normal equation are constructed according to the observation equation and the observation value weight matrix, and the normal equation serving for regional network rapid convergence and clock bias estimation is formed.

3. The regional station layout rapid convergence satellite clock error solving method according to claim 2, characterized in that, S101 comprises: 1) Establish current epoch The ionosphere-free combination equation of GNSS satellite pseudorange and carrier phase observation data for regional ground monitoring receivers: ； wherein , are the GNSS satellite and ground monitoring receiver respectively, and are the pseudorange and carrier phase observations of the regional ground monitoring receivers to the GNSS satellites respectively; is the geometric distance between the GNSS satellite and the ground monitoring receiver, calculated using as input the GNSS satellite precise orbit and the ground monitoring station receiver precise station coordinates, and are the ground monitoring receiver and GNSS satellite clock errors respectively, is the speed of light in vacuum, is the projection function of the tropospheric delay, is the zenith tropospheric delay calculated using an empirical model, is the residual zenith tropospheric delay; is the satellite wavelength, is the phase ambiguity, and are the pseudorange and phase measurement errors respectively; 2) calculating the pseudorange and phase data residual of the satellite for the regional monitoring receiver according to the ionosphere-free combination equation; ； wherein, and respectively represent the pseudorange ionosphere-free combined residual and the phase ionosphere-free combined residual.

4. The regional station layout rapid convergence satellite clock error solving method according to claim 2, characterized in that, S102 comprises: 1) constructing the observation equation; ； wherein is the time instant residual vector , is the design coefficient matrix, is the state vector to be estimated, is the constant vector; 2) constructing the normal equation matrix and the right matrix of the normal equation according to the observation equation and the observation value weight matrix; wherein the normal equation matrix is: ； right matrix of normal equations is: ； wherein is the observation weight matrix; 3) The simultaneous equation matrix and the right matrix of the equation of law constitute the equation of law, and the optimal estimation of the parameter is obtained by solving ; The normal equation is: 。 5. The regional station layout rapid convergence satellite clock error solving method according to claim 2, characterized in that, S2 comprises: S201, a random walk prediction model is established; a unit matrix is used as a state transition matrix, and the optimal estimation value of the last epoch is directly inherited as the prediction value of the current epoch; the covariance matrix of the last epoch is transmitted to the current epoch at the same time, and a process noise matrix is superimposed to quantify the error accumulation caused by the uncertainty of the model; Wherein, the diagonal elements of the prediction covariance matrix are respectively the prediction variances of the receiver clock error, the satellite clock error, the zenith troposphere delay and the phase ambiguity, and the non-diagonal elements are the mutual covariances between the diagonal elements; to fully characterize the uncertainty structure of the coupling of multiple parameters; S202, fixed increment covariance inflation is implemented on the three types of parameters of satellite clock error, receiver clock error and zenith tropospheric delay; S203, cycle slip detection and convergence judgment are performed on the phase ambiguity parameters, and if any of the conditions of "cycle slip occurs" or "convergence is achieved" is met, the current epoch ambiguity prediction variance is immediately reset to a minimum value, and a pseudo-observation strong constraint is applied; if neither cycle slip nor convergence occurs, the original variance level is maintained to continue filtering; S204, the covariance elements of each sub-block processed in S202 and S203 are integrated according to the parameter sequence to form a unified matrix, and a complete prediction covariance matrix is generated, which is used as the input of parameter estimation in the next step to realize fast convergence and high-precision maintenance of clock error solution in the regional monitoring network environment.

6. The satellite clock error solution method with fast convergence under regional station arrangement according to claim 5, characterized in that, In S202, fixed increment covariance inflation is implemented on the three types of parameters of satellite clock error, receiver clock error and zenith tropospheric delay; including: ； wherein, is the predicted variance of the receiver clock error at the current epoch, is the receiver clock error at the previous epoch; is the predicted variance of the satellite clock error at the current epoch, is the satellite clock error at the previous epoch; is the predicted variance of the zenith tropospheric delay at the current epoch, is the zenith tropospheric delay at the previous epoch, is the time difference between epochs, is the clock error random walk process noise intensity coefficient, is the tropospheric random walk process noise intensity coefficient.

7. The satellite clock error solution method with fast convergence under regional station arrangement according to claim 6, characterized in that, In S203, cycle slip detection and convergence judgment are performed on the phase ambiguity parameters; Epoch To Cycle slip detection, variance reset if cycle slip occurs; If cycle slip occurs, the covariance prediction of ambiguity parameters is: ； is the variance of the phase ambiguities; is the predicted variance of the phase ambiguities, is a first empirical threshold, chosen to be close to the upper variance limit of the search region of an integer cycle, is a second empirical threshold, chosen to be slightly larger than the lower variance limit of the measurement noise.

8. The method of claim 5, wherein the method further comprises: In S3, parameter estimation is performed to obtain the optimal solution of parameter estimation: ； wherein, is the normal equation matrix, is the normal equation right matrix, is the covariance matrix, is the optimal estimate of the parameters. Pseudorange post-fit residuals The calculation is: ； wherein, is the original ionosphere-free combined observation value of pseudorange, is the geometric distance calculated using the prior precise orbit and the precise station coordinates, is the projection function of tropospheric delay; is the zenith tropospheric delay calculated using the empirical model, is the residual zenith tropospheric delay, c is the speed of light in vacuum, is the satellite clock error, is the receiver clock error.

9. The satellite clock error solution method with fast convergence under regional station arrangement according to claim 8, characterized in that, In S4, the spreading code offset parameter is: ； wherein, a hardware-delayed satellite true inter- code bias; a server-computed clock-constant bias; ； wherein, to participate in the satellite total number of ground monitoring stations observing, to each station for the satellite valid pseudorange residuals over a number of epochs.

10. The satellite clock error solution method with fast convergence under regional station arrangement according to claim 9, characterized in that, In S5, the pseudo-range observations are corrected by precise satellite orbits, group delay parameters and precise clock error parameters; ； wherein, is the corrected pseudorange; is the original pseudorange observation; is the orbit correction number; is the precise satellite clock error; is the ionospheric delay parameter; The phase observations are corrected by precise satellite orbits and precise clock error parameters; ； wherein, is the corrected phase; is the original phase observation.

Images