High-precision troposphere delay modeling method, device and equipment under sparse GNSS (Global Navigation Satellite System) site

By constructing ZTD residual time series under sparse GNSS stations, generating virtual station ZTD values ​​and performing weighted fusion, and combining spherical cap harmonic analysis, the problem of high precision and high resolution of ZTD maps under sparse GNSS stations was solved, and high-precision ZTD map generation was achieved.

CN121477262AActive Publication Date: 2026-02-06广西壮族自治区自然资源信息中心 +1
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202511731260.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-24
Publication Date
2026-02-06
Estimated Expiration
2045-11-24

AI Technical Summary

Technical Problem

In scenarios with sparse GNSS stations and complex terrain, existing technologies struggle to balance the high spatial resolution and high precision of ZTD maps, making it impossible to generate high-resolution, high-precision ZTD maps and meet the high-precision application requirements of related fields.

Method used

The initial ZTD value is calculated using the first preset empirical model. The ZTD residual time series is constructed by combining the measured ZTD value from GNSS and the periodic signal is determined. The ZTD residual model is constructed to obtain the ZTD residual. The virtual station ZTD value is generated based on the second preset empirical model with the error correction term. The variance component estimation method is used for weighted fusion. The fitting coefficient is solved by spherical cap harmonic analysis to construct a continuous ZTD model and generate a high-resolution, high-precision ZTD map.

Benefits of technology

High-precision tropospheric delay modeling under sparse GNSS stations was achieved, significantly improving the accuracy and spatial continuity of tropospheric delay modeling, generating high-resolution and high-precision ZTD maps, and solving the problem of incomplete data coverage of sparse stations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121477262A_ABST
    Figure CN121477262A_ABST
Patent Text Reader

Abstract

The invention provides a high-precision troposphere delay modeling method, device and equipment under a sparse GNSS (Global Navigation Satellite System) site, and belongs to the technical field of spatial scientificity, and the method comprises the steps: calculating an initial ZTD (Zero Time Division) value of a real GNSS site through a first preset empirical model, constructing a ZTD residual time sequence by combining a GNSS actually measured ZTD value, analyzing a periodic signal, constructing a residual model to obtain a ZTD residual, and calculating the ZTD residual according to the ZTD residual. Further obtaining a ZTD improved value of the GNSS real station; based on a second preset empirical model in which an error correction term is introduced, virtual sites are arranged in the sparse region, and ZTD values of the virtual sites are generated; performing weighted fusion on the real and virtual ZTD values through a variance component estimation method to obtain a unified ZTD observation data set; and adopting spherical crown harmonic analysis to solve the adaptation coefficient, constructing a continuous ZTD model, outputting full-region data, and generating a high-resolution and high-precision ZTD map. According to the method, the defect that sparse GNSS station data coverage is insufficient is effectively overcome, and troposphere delay modeling precision, space continuity and area coverage integrity are remarkably improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of space science and technology, specifically to a high-precision tropospheric delay modeling method for sparse GNSS stations. Background Technology

[0002] Global Navigation Satellite Systems (GNSS), with their high-precision positioning, navigation, and timing capabilities, have been widely applied in key fields such as surveying and mapping, weather forecasting, and intelligent transportation. During GNSS signal propagation, tropospheric delay is one of the core error sources affecting positioning accuracy. Zenith Tropospheric Delay (ZTD), as a standardized benchmark value for tropospheric delay, is crucial for the successful implementation of high-precision GNSS applications. Furthermore, ZTD, as a key physical parameter reflecting atmospheric water vapor distribution and pressure changes, provides vital data support for short-term weather warnings and regional climate monitoring, serving as a core bridge connecting navigation and positioning with atmospheric science.

[0003] Current technologies for ZTD acquisition and modeling face significant limitations in practical applications, especially in scenarios with sparse GNSS stations and complex terrain. On the one hand, ZTD calculation methods based on empirical models (such as the Global Pressure and Temperature Model GPT3) rely on global historical meteorological reanalysis data to build models. While these methods can quickly output ZTD estimates without requiring measured meteorological data, they essentially fit historical atmospheric averages and cannot accurately match real-time local atmospheric changes. They also struggle to adapt to local water vapor distribution differences in complex terrain, resulting in insufficient calculation accuracy in areas with extreme weather or large topographical differences. On the other hand… While methods that infer ZTD from GNSS measured data can provide high-precision results, they are limited by the cost of GNSS station deployment and geographical conditions. In mountainous and remote areas, the distribution of stations is sparse, and traditional spatial interpolation methods (such as inverse distance weighting and ordinary kriging) are difficult to cope with the strong spatial variability of ZTD, and cannot effectively cover blank areas, thus failing to generate high spatial resolution regional ZTD maps. In addition, existing technologies have failed to effectively resolve the contradiction between "sparse station data" and "the need for full regional coverage", resulting in a tradeoff between accuracy and spatial resolution of ZTD maps in scenarios with sparse GNSS stations and complex terrain.

[0004] The most prominent core problem with existing technologies is that it is difficult to maintain both the high spatial resolution and high accuracy of ZTD maps in scenarios with sparse GNSS stations and complex terrain. Summary of the Invention

[0005] In view of this, it is necessary to provide a high-precision tropospheric delay modeling method, apparatus and equipment under sparse GNSS stations, to solve the technical problem that existing models cannot make up for the gaps in sparse GNSS station scenarios due to the sparseness of stations and incomplete data coverage. This results in insufficient accuracy and poor spatial continuity in tropospheric delay modeling, making it impossible to generate high-resolution and high-precision ZTD maps and meet the high-precision application requirements of related fields.

[0006] To address the aforementioned technical problems, in a first aspect, this invention provides a high-precision tropospheric delay modeling method for sparse GNSS stations, comprising: The initial ZTD value is calculated using the first preset empirical model, and the measured ZTD value is obtained using GNSS measurements. Based on the initial ZTD value and the measured ZTD value, a ZTD residual time series is constructed and its periodic signal is determined. Based on the periodic signal, a ZTD residual model is constructed and the ZTD residual is obtained. The ZTD improvement value of the actual GNSS station is obtained by combining the initial ZTD value and the ZTD residual. Based on a second pre-defined empirical model that introduces an error correction term, virtual station ZTD values ​​are generated in sparse areas of real GNSS stations. The variance component estimation method is used to weight and fuse the ZTD improved value and the virtual site ZTD value to obtain a unified ZTD observation dataset; Using the spherical cap harmonic analysis method, the fitting coefficients that adapt to the ZTD observation dataset and regional spatial distribution are solved by using known basis functions and the ZTD observation dataset. A continuous ZTD model is constructed based on the fitting coefficients, and a high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data of the whole region output by the continuous ZTD model.

[0007] In one possible implementation, calculating the initial ZTD value using a first preset empirical model includes: The ellipsoidal coordinates and time parameters of the actual GNSS station are input into the first preset empirical model, and meteorological parameters are output; the meteorological parameters include air pressure, temperature and water vapor pressure. The dry delay is determined based on the air pressure and temperature, and the wet delay is determined based on the water vapor pressure. The initial ZTD value is obtained by adding the dry delay and the wet delay.

[0008] In one possible implementation, constructing the ZTD residual time series based on the initial ZTD value and the measured ZTD value and determining its periodic signal includes: For each real GNSS station, calculate the difference between the measured ZTD value and the corresponding initial ZTD value at each time point in chronological order, and sort all the differences by time to form a ZTD residual time series; The ZTD residual time series was decomposed and reconstructed using a multi-channel singular spectrum analysis method to extract multi-period signals. The Lomb-Scargle method is used to determine the specific values ​​of the multi-period signal and obtain different periods for different real GNSS stations; the periods include annual periods, semi-annual periods and seasonal periods.

[0009] In one possible implementation, constructing the ZTD residual model based on the periodic signal and obtaining the ZTD residual includes: ZTD residual models are constructed based on annual, semi-annual, and seasonal cycles. The ZTD residual model is shown in the following equation:

[0010] in, Represents the ZTD residual. It is the average value of the ZTD residuals. , , and These represent the amplitudes of the annual cycle, semi-annual cycle, 120-day seasonal cycle, and 90-day seasonal cycle, respectively. , , and These are the phase angles corresponding to each cycle, and doy represents the accumulated days of the year; The ZTD residuals are estimated using the ZTD residual model.

[0011] In one possible implementation, the generation of virtual station ZTD values ​​in sparse areas of real GNSS stations based on a second preset empirical model with an introduced error correction term includes: The ZTD value is calculated based on the second preset empirical model and the difference between the calculated ZTD value and the measured ZTD value to obtain the ZTD deviation value. The error correction term of the virtual station is obtained by modeling the ZTD deviation value using the spherical cap harmonic analysis method. The error correction term is embedded into the second preset empirical model to form an optimized second preset empirical model; Based on the optimized second preset empirical model, virtual stations are deployed in sparse areas of real GNSS stations, and the ZTD value of the virtual stations is calculated.

[0012] In one possible implementation, the variance component estimation method is used to weight and fuse the ZTD improved value and the virtual site ZTD value to obtain a unified ZTD observation dataset, including: Calculate the variances of the ZTD improvement value and the virtual site ZTD value respectively to obtain the variances of the ZTD improvement value and the virtual site ZTD value, and set the weights of the ZTD improvement value and the virtual site ZTD value. The variance of ZTD improvement value unit weights is determined based on the variance of ZTD improvement value and the weight of ZTD improvement value; the variance of virtual site ZTD value unit weights is determined based on the variance of virtual site ZTD value and the weight of virtual site ZTD value. Compare the variance of the ZTD improved value unit weight with the variance of the virtual site ZTD value unit weight. If they are not equal, update the ZTD improved value weight and the virtual site ZTD value weight and iteratively calculate the unit weight variance. Stop iterating when the number of iterations exceeds the first set threshold or when the variance of the ZTD improved value unit weight and the virtual site ZTD value unit weight are equal, and output the final ZTD improved value weight and the virtual site ZTD value weight. Based on the final ZTD improvement value weight and the virtual site ZTD value weight, the ZTD improvement value and the virtual site ZTD value are weighted and fused to obtain a unified ZTD observation dataset.

[0013] In one possible implementation, the ZTD observation dataset includes ZTD improved values, virtual station ZTD values, and their corresponding spatial location parameters; the GNSS real stations and virtual stations together constitute the observation points; the step of employing the spherical cap harmonic analysis method, using known basis functions and the ZTD observation dataset to solve for the fit coefficients that adapt to the ZTD observation dataset and the regional spatial distribution, includes: The spatial location parameters corresponding to each observation point in the ZTD observation dataset are introduced; the spatial location parameters include the radial distance from the observation point to the Earth's center of mass and the spherical cap coordinates of the observation point; A fixed set of spherical cap harmonic basis functions is determined based on the spherical cap coordinates and radial distance of the observation point; The observation equation is constructed based on the spherical cap harmonic basis function; the least squares method is used to solve the fit coefficient between the ZTD observation dataset and the regional spatial distribution in the observation equation.

[0014] In one possible implementation, the step of constructing a continuous ZTD model based on the adaptation coefficients, and generating a high-resolution, high-precision ZTD map based on the full-area continuous and accurate ZTD data output by the continuous ZTD model, includes: Substituting the obtained fitting coefficients into the fixed set of spherical cap harmonic basis functions, a continuous ZTD mathematical model covering the target region is formed; Input spatial location parameters of any location within the target area into the continuous ZTD mathematical model to obtain the ZTD value corresponding to that location, and output continuous and accurate ZTD data for the entire area. Based on the continuous and accurate ZTD data of the entire region, a high-resolution and high-precision ZTD map is generated.

[0015] Secondly, the present invention also provides a high-precision tropospheric delay modeling device for sparse GNSS stations, comprising: The ZTD improvement value acquisition unit is used to calculate the initial ZTD value through a first preset empirical model, obtain the measured ZTD value using GNSS measurements, construct a ZTD residual time series based on the initial ZTD value and the measured ZTD value and determine its periodic signal, construct a ZTD residual model based on the periodic signal and obtain the ZTD residual, and combine the initial ZTD value and the ZTD residual to obtain the ZTD improvement value of the GNSS real station. The virtual station ZTD value acquisition unit is used to generate virtual station ZTD values ​​in sparse areas of real GNSS stations based on a second preset empirical model that introduces an error correction term. The ZTD observation dataset acquisition unit is used to perform weighted fusion of the ZTD improved value and the virtual site ZTD value using the variance component estimation method to obtain a unified ZTD observation dataset. The ZTD map generation unit is used to solve for the fit coefficients that fit the ZTD observation dataset and the regional spatial distribution using the spherical cap harmonic analysis method and the known basis functions and the ZTD observation dataset. Based on the fit coefficients, a continuous ZTD model is constructed, and a high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data of the whole region output by the continuous ZTD model.

[0016] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the high-precision tropospheric delay modeling method for sparse GNSS stations described in any of the above implementations.

[0017] The beneficial effects of this invention are as follows: The high-precision tropospheric delay modeling method provided by this invention calculates the initial ZTD value through a first preset empirical model and combines it with the measured ZTD value of GNSS, constructs the ZTD residual time series, determines the periodic signal, and constructs a residual model to obtain the ZTD residual, effectively correcting the initial value error to obtain a high-precision improved ZTD value of the real GNSS station; based on a second preset empirical model that introduces an error correction term, virtual station ZTD values ​​are generated in sparse areas, making up for the data coverage gap in sparse areas; the variance component estimation method is used to weight and fuse the two types of ZTD values, ensuring the consistency and reliability of the unified ZTD observation dataset; the fitting coefficient is solved by spherical cap harmonic analysis and a continuous ZTD model is constructed, outputting continuous and accurate ZTD data for the entire region and generating a high-resolution and high-precision ZTD map, realizing the transformation from discrete data to continuous regional modeling, and significantly improving the accuracy, spatial continuity, and regional coverage integrity of tropospheric delay modeling. Attached Figure Description

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

[0019] Figure 1 This is a schematic flowchart of an embodiment of the high-precision tropospheric delay modeling method under sparse GNSS stations provided by the present invention; Figure 2 This is a schematic flowchart of another embodiment of the high-precision tropospheric delay modeling method under sparse GNSS stations provided by the present invention; Figure 3 This is a schematic diagram of an embodiment of the present invention for calculating the initial ZTD value using a first preset empirical model; Figure 4 This is a schematic diagram of an embodiment of the present invention for constructing ZTD residual time series and determining its periodic signal; Figure 5 This is a schematic diagram of an embodiment of the present invention for generating virtual station ZTD values ​​in sparse areas of real GNSS stations; Figure 6 This is a schematic diagram of another embodiment of the present invention for generating virtual station ZTD values ​​in sparse areas of real GNSS stations; Figure 7 This is a schematic diagram of an embodiment of the present invention for obtaining a unified ZTD observation dataset; Figure 8 This is a schematic diagram of an embodiment of the present invention for solving the adaptation coefficient that is compatible with the ZTD observation dataset and the spatial distribution of the region; Figure 9 This is a schematic diagram of an embodiment of the process for generating high-resolution, high-precision ZTD maps provided by the present invention; Figure 10 This is a schematic diagram of an embodiment of the high-precision tropospheric delay modeling device under sparse GNSS stations provided by the present invention; Figure 11 This is a schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0020] 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 a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0022] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0023] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0024] Before demonstrating the embodiments, the following terms will be explained.

[0025] ZTD: Total Tropospheric Delay, refers to the total vertical delay of an electromagnetic signal as it passes through the troposphere. It consists of dry delay and wet delay and is a key parameter in GNSS positioning.

[0026] GNSS: Global Navigation Satellite System, is a collective term for satellite navigation systems that provide all-weather positioning, navigation and timing services to users worldwide, including BeiDou, GPS, GLONASS, etc.

[0027] Tropospheric delay: The propagation time delay caused by atmospheric refraction when satellite signals pass through the Earth's troposphere is one of the main sources of error affecting GNSS positioning accuracy.

[0028] Spherical Cap Harmonic Analysis (SCHA) is a mathematical method specifically designed for modeling regional geophysical fields. It approximates the physical field distribution of a local region using specific basis functions and is suitable for modeling local regions such as gravity fields and atmospheric parameters.

[0029] Variance component estimation: An adjustment method for processing multi-source observation data. It iteratively calculates the variance of each group of observations, reasonably determines the data weights, and optimizes the accuracy of the adjustment results.

[0030] Multichannel singular spectrum analysis (MSSA): An extended method for time series analysis that can process multidimensional synchronous data and extract periodic signals and trend features from the data through decomposition and reconstruction.

[0031] The Lomb-Scargle method is a period detection method designed specifically for non-uniformly sampled time series. It identifies periods by fitting a sine function and has strong resistance to noise and data sparsity.

[0032] Ellipsoidal coordinates: A spatial coordinate system based on a reference ellipsoid, which accurately describes the position of a point on the Earth's surface using geodetic longitude, geodetic latitude, and geodetic height.

[0033] Doy: The number of days accumulated starting from January 1st of each year, used to quantify time parameters and facilitate periodic signal analysis and model calculation.

[0034] Dry Delay: A component of ZTD, caused by dry components in the atmosphere such as nitrogen and oxygen. Its changes are stable and can be accurately predicted by meteorological models.

[0035] Wet delay: A component of ZTD, caused by changes in the water vapor content in the air. It has strong spatiotemporal variability and is the main factor affecting the accuracy of ZTD.

[0036] Virtual stations: Virtual observation points deployed in sparse areas of real GNSS stations. Their ZTD values ​​are calculated through a model and are used to supplement data coverage gaps.

[0037] Fit coefficient: A parameter obtained in spherical cap harmonic analysis, used to adapt the observation dataset to the spatial distribution characteristics of the region, and is the core of constructing a continuous ZTD model.

[0038] This invention provides a method, apparatus, and equipment for high-precision tropospheric delay modeling under sparse GNSS stations, which will be described below.

[0039] Figure 1 This is a schematic flowchart of an embodiment of the high-precision tropospheric delay modeling method for sparse GNSS stations provided by the present invention, as shown below. Figure 1 As shown, the high-precision tropospheric delay modeling method under sparse GNSS stations includes: S101. Calculate the initial ZTD value through the first preset empirical model, obtain the measured ZTD value using GNSS measurements, construct the ZTD residual time series based on the initial ZTD value and the measured ZTD value and determine its periodic signal, construct the ZTD residual model based on the periodic signal and obtain the ZTD residual, and combine the initial ZTD value and the ZTD residual to obtain the ZTD improvement value of the GNSS real station. S102. Based on the second preset empirical model with introduced error correction term, virtual station ZTD values ​​are generated in the sparse area of ​​real GNSS stations. S103. The variance component estimation method is used to weight and fuse the ZTD improved value and the virtual site ZTD value to obtain a unified ZTD observation dataset. S104. Using the spherical cap harmonic analysis method, the fitting coefficients that adapt to the ZTD observation dataset and regional spatial distribution are solved by using known basis functions and ZTD observation dataset. A continuous ZTD model is constructed based on the fitting coefficients. A high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data of the whole region output by the continuous ZTD model.

[0040] In some embodiments of the present invention, the first preset empirical model is the GPT2w model, which is a mature global tropospheric delay empirical model that can quickly provide initial ZTD values ​​and meet the basic requirements of "initial value calculation"; the second preset empirical model is the Gtrop model, which performs better in terms of accuracy and regional adaptability in high-altitude areas and is suitable for generating virtual station ZTD values ​​in sparse areas.

[0041] It should be noted that these two types of preset experience models are not the only options, and can be replaced with other similar models based on the following dimensions.

[0042] In summary, the high-precision tropospheric delay modeling method provided by this invention calculates initial ZTD values ​​using a first preset empirical model and combines them with measured GNSS ZTD values ​​to construct a ZTD residual time series, determine periodic signals, and construct a residual model to obtain ZTD residuals. This effectively corrects initial value errors to obtain high-precision improved ZTD values ​​for real GNSS stations. Based on a second preset empirical model that introduces error correction terms, virtual station ZTD values ​​are generated in sparse regions, filling data coverage gaps in sparse areas. A variance component estimation method is used to weight and fuse the two types of ZTD values, ensuring the consistency and reliability of the unified ZTD observation dataset. By using spherical cap harmonic analysis to solve for the fitting coefficients and construct a continuous ZTD model, continuous and accurate ZTD data for the entire region is output, and a high-resolution, high-precision ZTD map is generated. This realizes the transformation from discrete data to continuous regional modeling, significantly improving the accuracy, spatial continuity, and regional coverage integrity of tropospheric delay modeling.

[0043] In some embodiments of the present invention Figure 2 This is a schematic flowchart of an embodiment of the high-precision tropospheric delay modeling method for sparse GNSS stations provided by the present invention.

[0044] In some embodiments of the present invention, such as Figure 3 As shown, the initial ZTD value is calculated using the first preset empirical model, including: S301. Input the ellipsoidal coordinates and time parameters of the actual GNSS station into the first preset empirical model, and output meteorological parameters; meteorological parameters include air pressure, temperature and water vapor pressure; S302. Dry delay is determined based on air pressure and temperature, and wet delay is determined based on water vapor pressure. S303. Add the dry delay and wet delay to obtain the initial ZTD value.

[0045] In some embodiments of the present invention, the first preset empirical model is GPT2w (global pressure and temperature 2 wet), which is one of the more widely used tropospheric delay empirical models and can provide meteorological parameters such as air pressure, temperature, and water vapor pressure.

[0046] It should be noted that the Zenith Tropospheric Delay (ZTD) is composed of both the dry delay (ZHD) and the wet delay (ZWD), with the dry delay playing a dominant role.

[0047] In some embodiments of the present invention, the initial ZTD value is calculated using the GPT2w model, with the ellipsoidal coordinates (latitude, longitude, altitude) of the GNSS station and the corresponding time parameters (modified Julian day data) as input. The GPT2w model is based on ECMWFERA-Interim reanalysis data and outputs meteorological parameters (such as pressure, temperature, vapor pressure, etc.).

[0048] It should be noted that modifying the Julian Day data is an operation that adjusts the Julian Day data according to the actual situation; the two are related as "operation action" and "operation object".

[0049] ZHD and ZWD are calculated separately. ZHD is calculated using parameters such as surface pressure, as shown in Equation 1 below: (1) Where P is the surface pressure, in hPa, and h is the geodetic height of the station, in km. The latitude of the station is expressed in radians.

[0050] ZWD is calculated based on the following formula 2: (2) in, (This is the molar gas constant, with a value of 8.314 J / (mol·K)). and The molar masses of water and dry air are 18.0152 g / mol and 28.9644 g / mol, respectively. and The constants for the refractive index formula are 77.604±0.014 K / mbar, 64.79±0.08 K / mbar, and (3.776±0.004)×10⁵ K / mbar, respectively. Let be the specific gas constant of dry air. Let gravitational acceleration be the center of mass of the vertical column of atmosphere. For weighted average temperature, It is a water vapor reduction factor. This represents the corrected coefficient, which takes into account the effect of moisture content on soil compressibility, derived from the original coefficient. The result is obtained by subtracting a term related to moisture content. It refers to the void ratio when the soil is completely saturated, that is, when the pores are completely filled with water.

[0051] The initial ZTD value is obtained by adding ZHD and ZWD, as shown in Equation 3 below: (3) The initial ZTD value is obtained by adding the dry delay and the wet delay, which accurately separates the components of the tropospheric delay and provides a scientific basis for the calculation of the initial ZTD value.

[0052] In some embodiments of the present invention, such as Figure 4 As shown, a ZTD residual time series is constructed based on the initial ZTD value and the measured ZTD value, and its periodic signal is determined, including: S401. For each GNSS real station, calculate the difference between the measured ZTD value of GNSS at each time point and the corresponding initial ZTD value in chronological order, and sort all the differences by time to form a ZTD residual time series. S402. Multi-channel singular spectrum analysis is used to decompose and reconstruct the ZTD residual time series and extract multi-period signals from it. S403. Use the Lomb-Scargle method to determine the specific values ​​of multi-period signals and obtain different periods for different GNSS real stations; the periods include annual periods, semi-annual periods and seasonal periods.

[0053] In some embodiments of the present invention, multi-channel singular spectrum analysis is performed to verify whether periodic signals exist in the ZTD residual time series. First, the ZTD residual time series ΔZTD(t) is converted into a Hankel matrix form. Assuming the time series length is N, a window length L is selected (usually L≈N / 2). L is preferably longer than one year, as empirically, the annual periodic component has a relatively significant impact. Singular value decomposition is performed on the Hankel matrix to obtain eigenvalues ​​and eigenvectors. The reconstructed components (RCs) corresponding to the k largest eigenvalues ​​are selected; preferably, the present invention selects the first ten RCs. The RC sequences are observed to see if they exhibit patterns such as annual, semi-annual, or seasonal cycles, and the periodic signals are extracted.

[0054] In some embodiments of the present invention, to accurately determine the periodic values ​​in the ZTD residual time series, the Lomb-Scargle periodogram method is used to discriminate the specific values ​​of the residual periodic signal, calculate the Lomb-Scargle periodogram, and calculate the power spectral density at different frequencies f. Correspondingly, the period T = 1 / f. Calculate the power spectral density at different frequencies f. As shown in Equation 4 below: (4) In the formula, Indicates at a specific frequency The power spectral density calculated at a given frequency reflects the significance of the periodic component in the time series. Represents a point in time The observed value is the ZTD residual at time t; and This is the average of all observations. Subtracting it is to eliminate DC offset in the data, allowing the analysis to focus on the fluctuation signal itself. In the formula... It is the observation time point. It is a crucial time offset that varies with frequency. Dynamic adjustment ensures that the sine and cosine basis functions maintain mathematical orthogonality at non-uniform time points, thus guaranteeing the fairness and accuracy of periodic detection. It also identifies peaks in the power spectrum that are significantly higher than the noise level.

[0055] The Lomb-Scargle method is used to determine the specific values ​​of multi-period signals, effectively separating the periodic features in the ZTD residual time series, and providing accurate periodic parameter support for the construction of residual models.

[0056] In some embodiments of the present invention, constructing a ZTD residual model based on a periodic signal and obtaining the ZTD residual includes: ZTD residual models are constructed based on annual, semi-annual, and seasonal cycles. The ZTD residual model is shown in Equation 5 below: (5) in, Represents the ZTD residual. It is the average value of the ZTD residuals. , , and These represent the amplitudes of the annual cycle, semi-annual cycle, 120-day seasonal cycle, and 90-day seasonal cycle, respectively. , , and These are the phase angles corresponding to each cycle, and doy represents the accumulated days of the year; ZTD residuals are estimated using the ZTD residual model.

[0057] In some embodiments of the present invention, the unknown parameters in the residual model are estimated by the least squares method.

[0058] In some embodiments of the present invention, the initial ZTD value derived from the GPT2w model is combined with the ZTD residual estimated by the ZTD residual model to obtain the improved ZTD value, as shown in Equation 6 below: (6) in, This represents the initial ZTD value derived from the GPT2w model. This represents the ZTD residuals estimated using the ZTD residual model. This represents the ZTD improvement value of a real GNSS site.

[0059] ZTD residual models are constructed and ZTD residuals are estimated based on annual, semi-annual and seasonal cycles. The time variation of residuals is accurately fitted, providing a precise quantitative basis for error correction of the initial ZTD value.

[0060] In some embodiments of the present invention, such as Figure 5 As shown, based on a second preset empirical model that introduces an error correction term, virtual station ZTD values ​​are generated in sparse areas of real GNSS stations, including: S501. Calculate the ZTD value based on the second preset empirical model and subtract it from the measured ZTD value to obtain the ZTD deviation value. Model the ZTD deviation value using the spherical cap harmonic analysis method to obtain the error correction term of the virtual station. S502. Embed the error correction term into the second preset empirical model to form an optimized second preset empirical model; S503. Based on the optimized second preset empirical model, virtual stations are deployed in sparse areas of real GNSS stations, and the ZTD value of the virtual stations is calculated.

[0061] In some embodiments of the present invention, the second preset empirical model is the GTrop model. It should be noted that the GTrop model is a model that can provide high-precision tropospheric delay and temperature parameters globally. It is mainly used to correct GNSS positioning errors and support meteorological research, and is particularly suitable for application in areas lacking measured meteorological data and in complex terrain environments such as high altitudes.

[0062] In some embodiments of the present invention, such as Figure 6 As shown, Figure 6 This is a schematic diagram of another embodiment of the process for generating virtual station ZTD values ​​in sparse areas of real GNSS stations, as provided in this application. Combined with... Figure 6 In areas with sparse GNSS stations, virtual stations are generated to supplement spatial information and improve the spatial resolution of ZTD maps. Introducing a correction term into the GTrop model allows for a more accurate reflection of actual ZTD changes. The GTrop model with the correction term embedded is the optimized second preset empirical model, which can be called the GTrop-Cor model, and its implementation is shown in Equation 7 below: (7) in, These represent the latitude, longitude, and altitude of the observation point, respectively. This represents the ZTD (i.e., the ZTD of the virtual site) of the GTProp-Cor model. It is the ZTD value of the real site derived from the GTPop model, and It is the difference between real-world sites derived from the GTrop model and GNSS; Depend on After modeling and interpolation using Spherical Harmonic Analysis (SCHA), SCHA can fit a continuous spatial function based on the dZTD values ​​at sparse sites, thus predicting the dZTD value at any point within the study area.

[0063] In some embodiments of the present invention, the location of the virtual sample site is determined according to the following principles: Virtual sample stations should be evenly distributed in areas with missing data, and grid points from empirical models should be used preferentially. The number of virtual sample stations should be comparable to that of sparse real GNSS stations, and the longitude and latitude intervals should be set according to actual needs. Finally, a certain number of virtual sample stations are generated, each station having three-dimensional coordinates (longitude, latitude, and altitude).

[0064] Based on the optimized second preset empirical model, virtual stations are deployed in sparse areas of real GNSS stations and ZTD values ​​are calculated to reduce the calculation deviation of virtual station ZTD values ​​and improve the reliability and adaptability of supplementary data in sparse areas.

[0065] like Figure 6 As shown, Figure 6 This is a schematic diagram of another embodiment of generating virtual station ZTD values ​​in sparse areas of real GNSS stations provided in this application. In some embodiments of the present invention, such as Figure 7 As shown, the variance component estimation method is used to weight and fuse the improved ZTD values ​​and the virtual site ZTD values ​​to obtain a unified ZTD observation dataset, including: S701. Calculate the variances of the ZTD improvement value and the virtual site ZTD value respectively to obtain the variances of the ZTD improvement value and the virtual site ZTD value, and set the weights of the ZTD improvement value and the virtual site ZTD value. S702. Determine the weight variance of ZTD improved value unit based on the variance of ZTD improved value and the weight of ZTD improved value; determine the weight variance of virtual site ZTD value unit based on the variance of virtual site ZTD value and the weight of virtual site ZTD value. S703. Compare the variance of the ZTD improved value unit weight with the variance of the virtual site ZTD value unit weight. If they are not equal, update the ZTD improved value weight and the virtual site ZTD value weight and iteratively calculate the unit weight variance. Stop iterating when the number of iterations exceeds the first set threshold or when the variance of the ZTD improved value unit weight and the variance of the virtual site ZTD value unit weight are equal, and output the final ZTD improved value weight and the virtual site ZTD value weight. S704. Based on the final ZTD improvement value weight and the virtual site ZTD value weight, the ZTD improvement value and the virtual site ZTD value are weighted and fused to obtain a unified ZTD observation dataset.

[0066] In some embodiments of the present invention, the fusion strategy for weighted fusion of ZTD improvement values ​​and virtual site ZTD values ​​is to use the Helmert Variance Component Estimation (HVCE) method for adaptive weighting. The ZTD improvement values ​​are balanced through HVCE. ) and virtual sample site data ( To reduce the uncertainty of ZTD maps and improve their accuracy.

[0067] It should be noted that the HVCE method updates the weights iteratively to ensure that the unit weight variances of the two types of data are equal.

[0068] set up and These are the ZTD improvement values ​​( ) and virtual site ZTD value ( The variance of these two sets of data. The goal of the Helmert variance component estimation method is to find a weight... This ensures that the variance per unit weight is equal. First, initialize the weights. Let the initial weights for the ZTD improvement values ​​be... The initial weight of the virtual site's ZTD value is . Then calculate the unit weight variance for the first... In the next iteration, the unit weight variance is calculated as shown in Equation 8-9 below: (8) (9) in, Indicates the first Weights of the ZTD improvement value in the next iteration This represents the weight of the virtual site's ZTD value in the i-th iteration. express The unit weight variance express The unit weighted variance; The weights are updated by comparing the unit weight variances, and the weight update form is shown in Equation 10 below: (10) in Indicates the first The weight of the improved ZTD value in the next iteration.

[0069] Repeat the calculation of unit weight variance and update the weights until the unit weight variances are equal, as shown in Equation 11 below: (11) If the number of iterations exceeds a first set threshold (e.g., 20 times), the iteration stops and the final weights are output.

[0070] The ZTD improvement value and the virtual site ZTD value are weighted and fused based on the final weight, and the weight allocation of the two types of data is dynamically optimized to maximize the synergistic value of real high-precision data and virtual supplementary data.

[0071] In some embodiments of the present invention, such as Figure 8As shown, the ZTD observation dataset includes ZTD improved values, virtual station ZTD values, and their corresponding spatial location parameters; GNSS real stations and virtual stations together constitute the observation points; using the spherical cap harmonic analysis method, the fit coefficients that adapt to the ZTD observation dataset and regional spatial distribution are solved using known basis functions and the ZTD observation dataset, including: S801. Introduce the spatial location parameters corresponding to each observation point in the ZTD observation dataset; the spatial location parameters include the radial distance from the observation point to the Earth's center of mass and the spherical cap coordinates of the observation point; S802. Determine a fixed set of spherical cap harmonic basis functions based on the spherical cap coordinates and radial distance of the observation point; S803. Construct observation equations based on spherical cap harmonic basis functions; use the least squares method to solve the fit coefficients between the ZTD observation dataset and the regional spatial distribution in the observation equations.

[0072] In some embodiments of the present invention, the SCHA method is used to fit the weighted ZTD data to generate a high-precision ZTD map. The SCHA method is a regional modeling technique that can effectively reduce the impact of complex terrain. The expression of the SCHA model is shown in Equation 12 below: (12) in, The basis functions of the SCHA model are expressed in Equations 13-14 below: (13) (14) in, These are coordinates of the spherical crown; and These are the radial distance and the average radius of the Earth, respectively. and These are the degree and order of the spherical cap harmonic model ( (This is the maximum degree). It is a related Legendre function; and It is the harmonic coefficient of the spherical cap.

[0073] It should be noted that the SCHA method includes two scenarios: fitting and prediction. In the fitting phase, after processing the data into a spherical coordinate format, it can be used based on the matrix [ and corresponding To solve for the fit coefficient and In the prediction phase, having known the location information that needs to be estimated allows us to prepare in advance. and Once the coefficients in the fitting step are obtained and Then the value of the SCHA model can be calculated immediately.

[0074] The least squares method is used to solve the fitting coefficients in the observation equations, so that the spherical cap harmonic basis functions are accurately matched with the spatial features of the target region, providing core fitting parameters for the construction of continuous ZTD models.

[0075] In some embodiments of the present invention, such as Figure 9 As shown, a continuous ZTD model is constructed based on the adaptation coefficient. A high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data output from the continuous ZTD model covering the entire region, including: S901. Substitute the obtained fitting coefficients into a set of fixed spherical cap harmonic basis functions to form a continuous ZTD mathematical model covering the target area; S902. Input the spatial location parameters of any location within the target area into the continuous ZTD mathematical model, obtain the ZTD value corresponding to that location, and output continuous and accurate ZTD data for the entire area. S903 generates a high-resolution, high-precision ZTD map based on continuous and accurate ZTD data across the entire region.

[0076] High-resolution and high-precision ZTD maps are generated based on continuous and accurate ZTD data across the entire region, achieving seamless and continuous coverage of ZTD data in the target area and ensuring the high resolution and high precision of the output map.

[0077] To better implement the high-precision tropospheric delay modeling method under sparse GNSS stations in this invention embodiment, based on the high-precision tropospheric delay modeling method under sparse GNSS stations, the corresponding method is as follows: Figure 10 As shown, this embodiment of the invention also provides a high-precision tropospheric delay modeling device under sparse GNSS stations. The high-precision tropospheric delay modeling device 1000 under sparse GNSS stations includes: ZTD improvement value acquisition unit 1001 is used to calculate the initial ZTD value through the first preset empirical model, obtain the measured ZTD value using GNSS measurements, construct the ZTD residual time series based on the initial ZTD value and the measured ZTD value and determine its periodic signal, construct the ZTD residual model based on the periodic signal and obtain the ZTD residual, and combine the initial ZTD value and the ZTD residual to obtain the ZTD improvement value of the GNSS real station. The virtual station ZTD value acquisition unit 1002 is used to generate virtual station ZTD values ​​in sparse areas of real GNSS stations based on a second preset empirical model that introduces an error correction term. ZTD observation dataset acquisition unit 1003 is used to perform weighted fusion of ZTD improved values ​​and virtual site ZTD values ​​using the variance component estimation method to obtain a unified ZTD observation dataset; ZTD map generation unit 1004 is used to solve for the fit coefficients that fit the ZTD observation dataset and regional spatial distribution by using the spherical cap harmonic analysis method and the known basis functions and ZTD observation dataset. Based on the fit coefficients, a continuous ZTD model is constructed, and a high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data of the whole region output by the continuous ZTD model.

[0078] like Figure 11 As shown, the present invention also provides an electronic device 1100, including a memory and a processor, wherein the electronic device 1100 includes a processor 1101, a memory 1102 and a display 1103. Figure 11 Only some components of the electronic device 1100 are shown, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead.

[0079] In some embodiments, processor 1101 may be a central processing unit (CPU), microprocessor or other data processing chip, used to run program code stored in memory 1102 or process data, such as the high-precision tropospheric delay modeling method under sparse GNSS stations in this invention.

[0080] In some embodiments, processor 1101 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 1101 may be local or remote. In some embodiments, processor 1101 may be implemented on a cloud platform. In one embodiment, the cloud platform may include a private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-cloud, etc., or any combination thereof.

[0081] In some embodiments, memory 1102 may be an internal storage unit of electronic device 1100, such as a hard disk or memory of electronic device 1100. In other embodiments, memory 1102 may also be an external storage device of electronic device 1100, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on electronic device 1100.

[0082] Furthermore, the memory 1102 may include both internal storage units of the electronic device 1100 and external storage devices. The memory 1102 is used to store application software and various types of data installed on the electronic device 1100.

[0083] In some embodiments, display 1103 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 1103 is used to display information from electronic device 1100 and to display a visual user interface. Components 1101-1103 of electronic device 1100 communicate with each other via a system bus.

[0084] In one embodiment, when processor 1101 executes the high-precision tropospheric delay modeling program for sparse GNSS stations stored in memory 1102, the following steps can be implemented: The initial ZTD value is calculated using the first preset empirical model. The measured ZTD value is obtained using GNSS measurements. The ZTD residual time series is constructed based on the initial ZTD value and the measured ZTD value, and its periodic signal is determined. The ZTD residual model is constructed based on the periodic signal, and the ZTD residual is obtained. The ZTD improvement value of the actual GNSS station is obtained by combining the initial ZTD value and the ZTD residual. Based on a second pre-defined empirical model that introduces an error correction term, virtual station ZTD values ​​are generated in sparse areas of real GNSS stations. The variance component estimation method is used to weight and fuse the ZTD improved value and the virtual site ZTD value to obtain a unified ZTD observation dataset; The spherical cap harmonic analysis method is adopted to solve the fitting coefficients that fit the ZTD observation dataset and regional spatial distribution using known basis functions and ZTD observation dataset. A continuous ZTD model is constructed based on the fitting coefficients, and a high-resolution, high-precision ZTD map is generated based on the continuous and accurate ZTD data of the whole region output by the continuous ZTD model.

[0085] It should be understood that when the processor 1101 executes the high-precision tropospheric delay modeling program for sparse GNSS stations in the memory 1102, in addition to the functions mentioned above, it can also perform other functions, as detailed in the description of the corresponding method embodiments above.

[0086] Furthermore, the embodiments of the present invention do not specifically limit the type of electronic device 1100 mentioned. Electronic device 1100 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, electronic device 1100 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0087] Accordingly, this application also provides a computer-readable storage medium for storing computer-readable programs or instructions. When the programs or instructions are executed by a processor, they can implement the steps or functions in the high-precision tropospheric delay modeling method for sparse GNSS stations provided in the above-described method embodiments.

[0088] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0089] The above provides a detailed description of the high-precision tropospheric delay modeling method, apparatus, and equipment for sparse GNSS stations provided by this invention. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for high-precision modeling of tropospheric delay under sparse GNSS sites, characterized in that, The method comprises the following steps: calculating an initial ZTD value by a first preset empirical model, obtaining a measured ZTD value by GNSS measurement, constructing a ZTD residual time series based on the initial ZTD value and the measured ZTD value and determining a periodic signal thereof, constructing a ZTD residual model based on the periodic signal and obtaining a ZTD residual, and obtaining an improved ZTD value of a GNSS real site by combining the initial ZTD value and the ZTD residual; generating a virtual site ZTD value in a sparse area of the GNSS real site based on a second preset empirical model with an error correction term introduced; performing weighted fusion on the improved ZTD value and the virtual site ZTD value by a variance component estimation method to obtain a unified ZTD observation data set; using a spherical cap harmonic analysis method, solving an adaptive coefficient which is adapted to the ZTD observation data set and regional spatial distribution by using a known base function and the ZTD observation data set, constructing a continuous ZTD model based on the adaptive coefficient, and generating a high-resolution and high-precision ZTD map based on full-area continuous and accurate ZTD data output by the continuous ZTD model.

2. The method of claim 1, wherein, The method of calculating the initial ZTD value by the first preset empirical model comprises the following steps: inputting an ellipsoidal coordinate and a time parameter of the GNSS real site into the first preset empirical model to output a meteorological parameter; the meteorological parameter comprises air pressure, temperature and water vapor pressure; determining a dry delay based on the air pressure and the temperature, and determining a wet delay based on the water vapor pressure; adding the dry delay and the wet delay to obtain the initial ZTD value.

3. The method of claim 1, wherein, The method of constructing the ZTD residual time series based on the initial ZTD value and the measured ZTD value and determining the periodic signal thereof comprises the following steps: for each GNSS real site, calculating a difference value between a GNSS measured ZTD value and a corresponding initial ZTD value at each time in chronological order, and forming a ZTD residual time series by sorting all the difference values in chronological order; decomposing and reconstructing the ZTD residual time series by a multi-channel singular spectrum analysis method to extract a multi-periodic signal therefrom; determining a specific value of the multi-periodic signal by a Lomb-Scargle method to obtain different periods of different GNSS real sites; the periods comprise an annual period, a semi-annual period and a seasonal period.

4. The method of claim 3, wherein, The method of constructing the ZTD residual model based on the periodic signal and obtaining the ZTD residual comprises the following steps: constructing a ZTD residual model based on the annual period, the semi-annual period and the seasonal period; the ZTD residual model is shown in the following formula: wherein, denotes the ZTD residual, is the mean value of the ZTD residual, , , and denote the amplitudes of the annual cycle, semi-annual cycle, 120-day seasonal cycle and 90-day seasonal cycle, respectively, , , and are the phase angles corresponding to the respective cycles, and doy denotes the day of year; estimating the ZTD residual by the ZTD residual model.

5. The method of claim 1, wherein, The method of generating the virtual site ZTD value in the sparse area of the GNSS real site based on the second preset empirical model with the error correction term introduced comprises the following steps: calculating a ZTD value by the second preset empirical model and obtaining a ZTD deviation value by subtracting the measured ZTD value therefrom, modeling the ZTD deviation value by a spherical cap harmonic analysis method to obtain an error correction term of the virtual site; embedding the error correction term into the second preset empirical model to form an optimized second preset empirical model; based on the optimized second preset empirical model, arranging a virtual site in the sparse area of the GNSS real site to calculate a virtual site ZTD value.

6. The method of claim 5, wherein, The ZTD improved value and the virtual station ZTD value are weighted and fused based on the final ZTD improved value weight and the virtual station ZTD value weight to obtain a unified ZTD observation data set. The ZTD improved value and the virtual station ZTD value are weighted and fused based on the final ZTD improved value weight and the virtual station ZTD value weight to obtain a unified ZTD observation data set. The ZTD improved value and the virtual station ZTD value are weighted and fused based on the final ZTD improved value weight and the virtual station ZTD value weight to obtain a unified ZTD observation data set. The ZTD observation data set includes the ZTD improved value, the virtual station ZTD value and the corresponding spatial position parameters thereof; the GNSS real station and the virtual station jointly constitute observation points; the spherical cap harmonic analysis method is adopted to solve the adaptive coefficients adapted to the ZTD observation data set and the regional spatial distribution by using known base functions and the ZTD observation data set, including: The spatial position parameters of each observation point in the ZTD observation data set are introduced; the spatial position parameters include the radial distance of the observation point to the earth center and the spherical cap coordinate of the observation point; 7. The method of claim 6, wherein, A group of fixed spherical cap harmonic base functions are determined based on the spherical cap coordinate and the radial distance of the observation point; The observation equation is constructed based on the spherical cap harmonic base function; the adaptive coefficients of the ZTD observation data set and the regional spatial distribution adapted to the observation equation are solved by using the least square method. The continuous ZTD model is constructed based on the adaptive coefficients; the high-resolution and high-precision ZTD map is generated based on the full-region continuous and accurate ZTD data output by the continuous ZTD model, including: The adaptive coefficients obtained by solving are substituted into the group of fixed spherical cap harmonic base functions to form a continuous ZTD mathematical model covering the target region; 8. The method of claim 7, wherein, The spatial position parameters of any position in the target region are input into the continuous ZTD mathematical model to obtain the ZTD value corresponding to the position, and the full-region continuous and accurate ZTD data are output; The high-resolution and high-precision ZTD map is generated based on the full-region continuous and accurate ZTD data. The high-resolution and high-precision ZTD map is generated based on the full-region continuous and accurate ZTD data. ​ 9. An apparatus for high-precision modeling of tropospheric delay under sparse GNSS sites, characterized in that, ​ The ZTD improvement value acquisition unit is configured to calculate an initial ZTD value by using a first preset empirical model, obtain a measured ZTD value by using GNSS measurement, construct a ZTD residual time series based on the initial ZTD value and the measured ZTD value and determine a periodic signal thereof, construct a ZTD residual model based on the periodic signal and obtain a ZTD residual, and obtain a ZTD improvement value of a GNSS real site by combining the initial ZTD value and the ZTD residual; The virtual site ZTD value acquisition unit is configured to generate a virtual site ZTD value in a sparse area of GNSS real sites based on a second preset empirical model in which an error correction term is introduced; The ZTD observation dataset acquisition unit is configured to obtain a unified ZTD observation dataset by using a variance component estimation method to perform weighted fusion on the ZTD improvement value and the virtual site ZTD value; The ZTD map generation unit is configured to use a spherical cap harmonic analysis method to solve an adaptive coefficient which is adapted to the ZTD observation dataset and regional spatial distribution by using a known base function and the ZTD observation dataset, construct a continuous ZTD model based on the adaptive coefficient, and generate a high-resolution and high-precision ZTD map based on full-area continuous and accurate ZTD data output by the continuous ZTD model.

10. An electronic device, comprising: The device comprises a memory and a processor, wherein, The memory is configured to store a program; The processor is coupled with the memory and is configured to execute the program stored in the memory to implement the steps of the high-precision modeling method for the tropospheric delay of sparse GNSS sites according to any one of claims 1 to 8.

Citation Information

Patent Citations

  • Marine precise point positioning method taking observation station baseline length and troposphere constraint into consideration

    CN116299596A

  • Point-plane conversion method and system for real-time troposphere zenith delay

    CN117992706A

  • Real-time ZTD area modeling method based on GNSS and ground meteorological station observation

    CN118840499A

  • GNSS (Global Navigation Satellite System) virtual observation value generation method and system based on real-time precise point positioning

    CN119716933A

  • Improved phase ambiguity resolution using three GNSS signals

    WO2006108227A1