A method for constructing a regional polynomial ionospheric oblique delay model

Abstract

The application discloses a kind of construction methods of regional polynomial ionospheric oblique delay model, belong to the ionospheric delay error correction field of global satellite navigation system, comprising: the signal piercing point coordinates are calculated by the longitude and latitude coordinates of reference station, then signal piercing point coordinates, observation time and satellite elevation angle are jointed as independent variable, and STEC is as dependent variable to construct polynomial function model;By synchronously including the space-time dynamic characteristics and satellite line-of-sight geometric characteristics into modeling, the space-time variation law of ionosphere is accurately captured;Using the analytical properties of polynomial, the model parameters can be quickly output, users only need to input their own coordinates, time, satellite elevation angle and satellite azimuth angle to directly calculate the required STEC, significantly simplify the process and improve real-time performance.The present application fills the accuracy and efficiency gap of the prior art in the complex scene, and provides a more reliable ionospheric error correction scheme for regional GNSS high-precision positioning.

Description

Technical Field

[0001] This invention belongs to the field of ionospheric delay error correction for global satellite navigation systems, and specifically relates to a method for constructing a regional polynomial ionospheric slant delay model. Background Technology

[0002] Global Navigation Satellite System (GNSS) has been widely used in many fields. However, ionospheric delay is still one of the main sources of error that restricts the positioning accuracy of GNSS. In satellite navigation and positioning, ionospheric delay error can reach several meters to hundreds of meters, which seriously affects the reliability of positioning results. At present, in order to reduce the impact of this error, the industry has developed a variety of ionospheric delay models, including: (1) using projection functions to convert the total electron content of the oblique path (STEC) to the total electron content of the vertical path (VTEC) to establish a grid model. When using it, the user obtains the signal puncture point VTEC through grid interpolation, and then obtains the required STEC through the inverse transformation of the projection function to realize ionospheric delay correction. However, this method requires multiple conversions between STEC and VTEC, which is not only cumbersome in modeling process and has poor real-time performance, but also generates cumulative projection function error. (2) sending all base station-satellite STEC data directly to the user, and the user performs ionospheric delay correction according to their own positioning algorithm. However, this method requires the transmission of massive amounts of raw STEC data, resulting in high transmission costs, a heavy data processing burden on the user end, and limited coverage of the base station, leading to significant deficiencies in accuracy, real-time performance, and applicability. (3) A function model is established using the latitude and longitude coordinates of the base station and the corresponding STEC data. The model is then sent to the user, who then interpolates the required STEC data based on their own station coordinates. However, this method only uses spatial coordinates as independent variables for modeling. When users interpolate using coordinates, the accuracy is easily insufficient due to the spatiotemporal heterogeneity of the ionosphere (such as day-night variations and regional disturbances) and the accumulation of low elevation angle signal errors. Summary of the Invention

[0003] The purpose of this invention is to address the problems of cumbersome modeling processes, poor real-time performance, and insufficient characterization of spatiotemporal dynamics and geometric features in existing technologies. It provides a method for constructing a regional polynomial ionospheric oblique delay model. First, the coordinates of the signal penetration point are calculated using the latitude and longitude coordinates of a reference station. Then, a polynomial function model is constructed using these signal penetration point coordinates, observation time, and satellite elevation angle as joint independent variables, and STEC as the dependent variable. This eliminates the need for the conversion between STEC and VTEC, fundamentally avoiding the problem of projection function error accumulation. By simultaneously incorporating spatiotemporal dynamic characteristics and satellite line-of-sight geometric features into the modeling, the spatiotemporal variation patterns of the ionospheric region are accurately captured. Utilizing the analytical properties of polynomials, model parameters can be quickly output. Users only need to input their own coordinates, time, satellite elevation angle, and satellite azimuth angle to directly calculate the required STEC, significantly simplifying the process and improving real-time performance. This model effectively fills the gaps in accuracy and efficiency of existing technologies in complex regional scenarios, providing a more reliable ionospheric error correction scheme for high-precision GNSS positioning in regions.

[0004] According to one aspect of the present invention, a method for constructing a regional polynomial ionospheric slant delay model is provided, comprising:

[0005] S1. Acquire observation data including total electron content along the oblique path, latitude and longitude coordinates of the base station, observation time, satellite elevation angle, and satellite azimuth angle, and obtain signal puncture point coordinates, normalized time, satellite elevation angle, and total electron content along the oblique path through preprocessing;

[0006] S2, using the coordinates of the signal puncture point Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows:

[0007] ,in, These are the polynomial coefficients;

[0008] S3. Solve and optimize the multivariate polynomial function model using the preprocessed observation data;

[0009] S4. Based on the optimized multivariate polynomial function model, clean the preprocessed observation data: if no abnormal data is removed, output the multivariate polynomial function model optimized in S3.

[0010] Furthermore, the preprocessing includes:

[0011] The latitude and longitude coordinates of the base station are converted into coordinates in the geocentric-earth-fixed coordinate system. Using the coordinates in the geocentric-earth-fixed coordinate system, the satellite azimuth angle, the satellite elevation angle, and the ionospheric height, the coordinates of the signal puncture point are calculated.

[0012] The observation time was converted into normalized time.

[0013] Further, S3 includes:

[0014] Divide the dataset into time windows to generate a continuous time window sequence;

[0015] A design matrix is ​​constructed based on the dataset corresponding to the time window of the same satellite observed from other reference stations;

[0016] The Bayesian optimization algorithm is used to search for the regularization parameter that minimizes the cross-validation loss, which is then used as the optimal ridge regression regularization parameter.

[0017] Based on the design matrix and the optimal ridge regression regularization parameters, the coefficients of the multivariate polynomial function model are obtained by solving the ridge regression function.

[0018] Furthermore, cleaning the preprocessed observation data also includes:

[0019] If abnormal data is removed and the amount of data after removal meets the preset requirements, the coefficients of the optimized multivariate polynomial function model will continue to be solved.

[0020] Furthermore, cleaning the preprocessed observation data also includes:

[0021] If abnormal data is removed and the amount of data after removal does not meet the preset requirements, the current time window will be skipped.

[0022] Furthermore, the method also includes:

[0023] The predicted value of the total electron content of the oblique path is calculated based on the joint independent variables in the dataset;

[0024] The absolute value of the residual is obtained by subtracting the predicted value of the total electron content of the oblique path from the true value of the total electron content of the oblique path in the dataset.

[0025] When the absolute value of the residual is greater than the set value, the true value of the total electron content of the oblique path in the corresponding dataset will be removed as an outlier sample.

[0026] Furthermore, the method also includes:

[0027] When the absolute value of the residual is less than or equal to the set value, the multivariate polynomial function model obtained from the first optimization is used.

[0028] According to one aspect of the present invention, a system for constructing a regional polynomial ionospheric slant delay model is provided, for implementing the method for constructing a regional polynomial ionospheric slant delay model as described in any one of the claims, comprising:

[0029] The preprocessing module is used to acquire observation data including total electron content along the oblique path, latitude and longitude coordinates of the base station, observation time, satellite elevation angle, and satellite azimuth angle, and to obtain signal puncture point coordinates, normalized time, satellite elevation angle, and total electron content along the oblique path through preprocessing.

[0030] The function model building module is used to construct the signal puncture point coordinates. Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows:

[0031] ,in, These are the polynomial coefficients;

[0032] The solution optimization module is used to solve and optimize the multivariate polynomial function model using the preprocessed observation data;

[0033] The data cleaning module is used to clean the preprocessed observation data based on the optimized multivariate polynomial function model: when no abnormal data is removed, it outputs the multivariate polynomial function model obtained by solving the optimization module.

[0034] According to one aspect of the present invention, an electronic device is provided, including a memory and a processor, the memory storing a computer program, wherein the processor executes the computer program to implement the steps of a method for constructing the regional polynomial ionospheric slant delay model.

[0035] According to one aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for constructing the regional polynomial ionospheric slant delay model.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] 1. This invention directly constructs a multinomial function model using the coordinates of the signal puncture point, observation time, and satellite elevation angle as joint independent variables and STEC as the dependent variable. There is no need to perform the conversion between STEC and VTEC. The required STEC can be calculated directly, which avoids the problem of projection function error accumulation from the root, greatly simplifies the modeling process, and significantly improves the real-time performance of the model.

[0038] 2. This invention incorporates spatiotemporal dynamic characteristics (coordinates of the signal penetration point and observation time) and satellite line-of-sight geometric features (satellite elevation angle) into the modeling, which can accurately capture the spatiotemporal variation patterns of the ionospheric region. At the same time, it improves the model's generalization ability through Bayesian regularization optimization and outlier data cleaning, effectively improving GNSS positioning accuracy.

[0039] 3. This invention eliminates the need to transmit massive amounts of raw STEC data. Instead, it simply sends the constructed polynomial function model parameters to the user. The process of calculating the required STEC on the user's end is simple, significantly reducing data transmission costs and the burden of data processing on the user's end. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A schematic diagram illustrating the construction method of the regional polynomial ionospheric slant delay model provided by the present invention.

[0042] Figure 2 A flowchart illustrating the construction process of the regional polynomial ionospheric slant delay model provided by this invention.

[0043] Figure 3 This is a schematic diagram of the error statistics provided by the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Specifically, such as Figure 1 As shown, this embodiment of the invention proposes a method for constructing a regional polynomial ionospheric oblique delay model, including: S1, acquiring observation data including the total electron content of the oblique path, the latitude and longitude coordinates of the reference station, the observation time, the satellite elevation angle, and the satellite azimuth angle, and obtaining the signal puncture point coordinates, normalized time, satellite elevation angle, and total electron content of the oblique path through preprocessing; S2, using the signal puncture point coordinates... Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows: ,in, S3. Solve and optimize the multivariate polynomial function model using the preprocessed observation data; S4. Clean the preprocessed observation data based on the optimized multivariate polynomial function model: if no abnormal data is removed, output the multivariate polynomial function model optimized in S3.

[0046] Specifically, such as Figure 2 As shown, starting from the raw observation data of the base station, STEC data is extracted and system bias (pseudorange hardware delay bias at the satellite and receiver ends) is removed. Data preprocessing is completed by calculating the signal puncture point coordinates and time normalization. Then, a multivariate polynomial function model is constructed using "signal puncture point coordinates - normalized time - elevation angle" as joint independent variables. A Bayesian optimization algorithm is used to search for the optimal ridge regression regularization parameters, and outlier cleaning is employed to improve the model's generalization ability. Finally, the model accuracy is verified and errors are statistically analyzed based on the test set. The specific steps include:

[0047] 1. Nine reference stations were selected. The total electron content (STEC) along the oblique path in the satellite-receiver line-of-sight direction was extracted from the raw observation data of the reference stations using the phase smoothing pseudorange method. The pseudorange hardware delay bias at the satellite and receiver ends was removed. At the same time, the latitude and longitude coordinates of the corresponding reference stations were extracted. The observation time (time, accurate to the second), satellite elevation angle (el), and satellite azimuth angle (az) are used to form the original dataset.

[0048] 2. Data preprocessing:

[0049] 2.1 Calculate the coordinates of the signal puncture point

[0050] First, convert the latitude and longitude coordinates of the base station into XYZ coordinates in the Earth-centered Earth-fixed (ECEF) coordinate system. The formula is as follows:

[0051] (1)

[0052] (2)

[0053] (3)

[0054] (4)

[0055] Among them, the long half-axis meters, first eccentricity squared Flatness N is the radius of curvature. , Latitude ,longitude The radian value is set, and the height h is set to 0.

[0056] Using the coordinates of the reference station in the Earth-centered Earth-fixed coordinate system, the satellite azimuth angle, the satellite elevation angle, and the ionospheric height (set to 450km), the IPPLocation function is called to calculate the coordinates of the signal puncture point. The function's principle formula is as follows:

[0057] (5)

[0058] (6)

[0059] (7)

[0060] When calculating the longitude of the signal puncture point, if the following conditions are met... and or satisfy and ,but

[0061] (8)

[0062] otherwise,

[0063] (9)

[0064] Finally, the latitude and longitude coordinates of the signal puncture point are... Convert to degrees.

[0065] Among them, Earth's radius m, ionospheric height m, el_rad are the satellite elevation angles (in radians), and az_rad are the satellite azimuth angles (in radians). For projection coefficients, This is the adjusted angle (used for subsequent geometric calculations). The latitude coordinates of the receiver. This refers to the receiver's longitude coordinates. Note that if the input coordinate parameters are two-dimensional, use them directly; otherwise, call the XYZ2BLH function to convert the ECEF coordinates. Convert to latitude and longitude And the height H, then calculate the coordinates of the signal puncture point.

[0066] Specifically, the principle of the XYZ2BLH function is: longitude Calculate horizontal distance Initial latitude initial height Then iteratively calculate the latitude B and altitude H:

[0067] (10)

[0068] (11)

[0069] (12)

[0070] (13)

[0071] like ,but , Otherwise, the iteration ends. Finally, the rad2deg function in Matlab is used to convert latitude and longitude from radians to degrees.

[0072] 2.2 Convert the observation time into a normalized time parameter (t), the calculation formula is as follows:

[0073] (14)

[0074] 2.3. Dataset Partitioning. For the original dataset, data from one station is selected as the test set, and data from the other stations are used as the training set. The training and test sets are filtered by setting PRN numbers starting with 'C', selecting BeiDou data to serve as the training and test sets respectively.

[0075] 3. Based on the preprocessed training set, establish a system with coordinates of the signal puncture point. The model uses normalized time (t) and satellite elevation angle (el) as joint independent variables, and STEC as the dependent variable in a multivariate polynomial function model. An example of the model expression is as follows:

[0076] (15)

[0077] in, These are the polynomial coefficients.

[0078] 3.1 Dividing the time window: January 11, 2025 and February 12, 2025 are selected as the time range for data processing. Based on the selected time range, a time window sequence is generated at 5-minute intervals to achieve time-based capture of the dynamic characteristics of the ionosphere.

[0079] 3.2 Constructing the Design Matrix: For the training data of each satellite and each time window, based on four fundamental independent variables (latitude and longitude coordinates of the signal penetration point, satellite elevation angle, and normalized time), each row of the design matrix X corresponds to 14 features of a sample, specifically as follows: By expanding the basic independent variables of each sample according to this feature, the design matrix X corresponding to the polynomial terms in Formula 16 below can be constructed.

[0080] (16)

[0081] The design matrix X and the dependent variable (true STEC values) are standardized to eliminate the influence of dimensional differences on parameter estimation. Zero-mean standardization is applied to each characteristic column (14 columns in total) of the design matrix X, using the following formula:

[0082] (17)

[0083] in, Let j be the original column vector of the j-th feature. The mean of this feature. The standard deviation of this feature. This is the standardized design matrix.

[0084] The dependent variable is standardized, and the formula is:

[0085] (18)

[0086] in, The mean of the dependent variable. The standard deviation of the dependent variable. This is the standardized dependent variable vector.

[0087] 3.3 Optimize regularization parameters: Set the parameter search range to... A Bayesian optimization algorithm is employed, targeting the 5-fold cross-validation loss function (ridge_cv_loss), to search for the regularization parameter that minimizes the cross-validation loss, which is then used as the optimal ridge regression regularization parameter. The cross-validation loss function is calculated as follows: the training data is randomly divided into 5 subsets; 4 subsets are selected sequentially as the training set, and the remaining subset as the test set; the model is trained using the training set, and the mean squared error (MSE) between the predicted and actual values ​​is calculated on the test set; this process is repeated, with each subset serving as a test set, and the average of the 5 MSEs is taken as the cross-validation loss, as shown in the formula:

[0088] (19)

[0089] In the formula, Let be the number of samples in the k-th test set. Let i be the true value of the i-th sample in the k-th test set. The predicted value corresponding to the position.

[0090] 3.4 Training the Model: The polynomial coefficients with L2 regularization are solved by calling the `ridge` function in Matlab. The form of the multivariate polynomial function model is as follows:

[0091] (20)

[0092] in, Let be the coefficient vector of the polynomial to be solved. The design matrix is ​​a quadratic polynomial. The ridge function solves an optimization problem by adding an L2 regularization term to the least squares loss function, i.e.:

[0093] (twenty one)

[0094] In the formula, Let i be the true value of the i-th sample. Let be the input feature vector of the i-th sample. To optimize the obtained regularization parameters.

[0095] 3.5 Cleaning Training Data: Calculate the STEC predicted values ​​using the training data, subtract them from the true values, and remove outlier samples with an absolute residual value greater than 1 TECU, retaining only valid training data. If the amount of training data after cleaning still meets the requirement (greater than or equal to 60, to ensure model accuracy), then re-standardize the data and optimize the regularization parameters again, calling the ridge function to solve for the final model coefficients. If it is found that the training data has not been removed after cleaning, then directly use the model coefficients obtained during the first model training. If there is insufficient data after removal, then skip this time window.

[0096] 4. The test set is processed using standardized parameters consistent with the training data. The test set design matrix is ​​then used. The standardized test data is substituted into the model to calculate the predicted value. Then, the model is de-standardized according to Formula 18 to obtain the true value of the total electron content of the oblique path corresponding to the test set, which is the final prediction result.

[0097] 5. Calculate the prediction residuals for each satellite in each time window ( ), and statistically analyze the average error and root mean square error (RMSE) of each satellite within a given time range, such as Figure 3 As shown, the error analysis results for each satellite are stored as text files and mat files.

[0098] Formula for calculating average error:

[0099] (twenty two)

[0100] Formula for calculating root mean square error:

[0101] (twenty three)

[0102] Where n is the number of residuals, For the true value, These are predicted values.

[0103] The implementation of the various embodiments of the present invention is based on programmed processing through a device with processor functionality. Therefore, in practical engineering, the technical solutions and functions of the various embodiments of the present invention are encapsulated into various modules. Based on this reality, and building upon the above embodiments, the embodiments of the present invention provide a system for constructing a regional polynomial ionospheric slant delay model. This system is used to execute a method for constructing a regional polynomial ionospheric slant delay model as described in the above method embodiments.

[0104] The system includes: a preprocessing module for acquiring observation data including total electron content along the oblique path, latitude and longitude coordinates of the base station, observation time, satellite elevation angle, and satellite azimuth angle, and obtaining signal puncture point coordinates, normalized time, satellite elevation angle, and total electron content along the oblique path through preprocessing; and a function model construction module for using the signal puncture point coordinates... Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows: ,in, The coefficients are polynomials; the solution optimization module is used to solve and optimize the multivariate polynomial function model using the preprocessed observation data; the data cleaning module is used to clean the preprocessed observation data based on the optimized multivariate polynomial function model: if no abnormal data is removed, the optimized multivariate polynomial function model obtained in the solution optimization module is output.

[0105] The regional polynomial ionospheric oblique delay model construction system provided in this invention addresses the problems of cumbersome modeling processes, poor real-time performance, and insufficient characterization of spatiotemporal dynamics and geometric features in existing technologies. It employs several modules, first calculating the signal penetration point coordinates using the latitude and longitude coordinates of a reference station. Then, using these signal penetration point coordinates, observation time, and satellite elevation angle as joint independent variables, and STEC as the dependent variable, a polynomial function model is constructed. This eliminates the need for the conversion between STEC and VTEC, fundamentally avoiding the problem of projection function error accumulation. By simultaneously incorporating spatiotemporal dynamic characteristics and satellite line-of-sight geometric features into the modeling, it accurately captures the spatiotemporal variation patterns of the ionospheric region. Utilizing the analytical properties of polynomials, model parameters can be quickly output; users only need to input their own coordinates, time, satellite elevation angle, and satellite azimuth angle to directly calculate the required STEC, significantly simplifying the process and improving real-time performance.

[0106] This invention also provides an electronic device, including a memory and a processor. The memory stores computer-executable instructions, and the processor executes the computer-executable instructions to implement a method for constructing a regional polynomial ionospheric slant delay model as proposed in the above embodiments.

[0107] This invention also provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, this program effectively addresses the shortcomings of existing technologies in terms of accuracy and efficiency in complex regional scenarios, providing a more reliable ionospheric error correction scheme for high-precision GNSS positioning in regional areas.

[0108] Finally, it should be noted that the above specific embodiments are merely representative examples of the present invention. Obviously, the present invention is not limited to the above specific embodiments and many variations are possible. Any simple modifications, equivalent changes, and alterations made to the above specific embodiments based on the technical essence of the present invention should be considered within the protection scope of the present invention.

Claims

1. A method for constructing a regional polynomial ionospheric slant delay model, characterized in that, The method comprises the following steps: S1, obtaining observation data comprising slant path total electron content, latitude and longitude coordinates of a reference station, observation time, satellite elevation angle and satellite azimuth angle, and obtaining signal piercing point coordinates, normalized time, satellite elevation angle and slant path total electron content through preprocessing; S2, using the coordinates of the signal puncture point Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows: ;in, It consists of 14 features; S3, solving and optimizing the multi-element polynomial function model by using the preprocessed observation data; S4, based on the optimized multi-element polynomial function model, cleaning the preprocessed observation data: when no abnormal data is removed, outputting the multi-element polynomial function model obtained by optimization in S3.

2. The method of claim 1, wherein the method further comprises: The preprocessing comprises: Converting the latitude and longitude coordinates of the reference station into coordinates in the geocentric and geodetic coordinate system, and calculating the signal piercing point coordinates by using the coordinates in the geocentric and geodetic coordinate system, the satellite azimuth angle, the satellite elevation angle and the ionospheric height; Converting the observation time into normalized time.

3. The method of claim 1, wherein the method further comprises: The S3 comprises: Dividing a time window based on a data set, and generating a continuous time window sequence; Based on the data set corresponding to the time window of the same satellite observed by other reference stations, a design matrix is constructed; Using a Bayesian optimization algorithm to search for a regularization parameter that minimizes the cross-validation loss, as the optimal ridge regression regularization parameter; Based on the design matrix and the optimal ridge regression regularization parameter, the coefficients of the multi-element polynomial function model are solved by using a ridge regression function.

4. The method of claim 1, wherein the method further comprises: The cleaning of the preprocessed observation data further comprises: When abnormal data is removed and the data quantity after removal meets the preset requirement, the coefficients of the multi-element polynomial function model are continuously solved and optimized.

5. The method of claim 3, wherein the method further comprises: The cleaning of the preprocessed observation data further comprises: When abnormal data is removed and the data quantity after removal does not meet the preset requirement, the current time window is skipped.

6. The method of claim 1, wherein, The method further comprises: Based on the joint independent variables in the data set, a predicted value of the slant path total electron content is calculated; The predicted value of the slant path total electron content is subtracted from the true value of the slant path total electron content in the data set to obtain a residual absolute value; When the residual absolute value is greater than a set value, the true value of the slant path total electron content in the corresponding data set is removed as an abnormal sample.

7. The method of claim 6, wherein the method further comprises: The method further comprises: When the residual absolute value is less than or equal to the set value, the multi-element polynomial function model obtained by the first solving and optimization is used.

8. A system for constructing a regional polynomial ionospheric slant delay model, characterized by, A method for constructing a regional polynomial ionospheric slant delay model according to any one of claims 1-7, comprising: A preprocessing module for obtaining observation data comprising slant path total electron content, latitude and longitude coordinates of a reference station, observation time, satellite elevation angle and satellite azimuth angle, and obtaining signal piercing point coordinates, normalized time, satellite elevation angle and slant path total electron content through preprocessing; The function model building module is used to construct the signal puncture point coordinates. Normalized time Satellite elevation angle As a joint independent variable, the total electron content along the oblique path is used. A multivariate polynomial function model is constructed for the dependent variable, and the expression of the multivariate polynomial function model is as follows: ; wherein, are 14 features; A solving and optimization module for solving and optimizing the multi-element polynomial function model by using the preprocessed observation data; A data cleaning module for cleaning the preprocessed observation data based on the optimized multi-element polynomial function model: when no abnormal data is removed, outputting the multi-element polynomial function model obtained by optimization in the solving and optimization module. 9.An electronic device comprising a memory and a processor, the memory storing a computer program, wherein, The computer program is executed by the processor to implement the steps of the method for constructing the regional polynomial ionospheric slant delay model according to any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method for constructing the regional polynomial ionospheric slant delay model according to any one of claims 1-7.

Images