An Ionospheric Data Enhancement Method, Device, Equipment and Medium over the Ocean

Abstract

The present invention discloses a method, device, equipment and medium for ionospheric data enhancement over the ocean, belonging to the technical field of ionospheric modeling. The method includes: obtaining vertical total electron content data at several positions of the ionosphere over the ocean through observations by low-Earth orbit satellites; combining the vertical total electron content data at several positions of the ionosphere over the ocean, as well as the longitude, latitude, observation time and space weather state data of the corresponding ionospheric piercing points as input feature data; inputting the input feature data into a first learner to obtain enhanced feature data; and inputting the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean. The present invention enhances ionospheric data over the ocean through an integrated machine learning model, and can obtain virtual observation data of the ionosphere in unobserved areas over the ocean, effectively improving the ionospheric observation coverage over the ocean.

Description

Technical Field

[0001] The present invention belongs to the technical field of ionospheric modeling, and relates to a method, device, equipment and medium for enhancing low-orbit ionospheric data over the ocean. Background Art

[0002] The ionosphere usually refers to the upper atmosphere of the Earth from about 60 km above the ground to the top of the magnetosphere. It is named as the ionosphere because there are a large number of free electrons and ions formed by ionization by X-rays, ultraviolet rays and high-energy particles in this area. Its activity level is gradually increasing as the current solar activity enters a new round of active cycle. At present, ionospheric delay is one of the important error sources in navigation and positioning. At the same time, the intense ionospheric activity has brought serious impacts on the services and applications of satellite navigation systems. In this context, accurately constructing and predicting the spatiotemporal distribution and changing state of the ionospheric total electron content (TEC) is of great significance to mitigate its impact on the performance of high-precision navigation and positioning services.

[0003] In recent years, thanks to the rapid development of multi-system GNSS and low earth orbit (LEO) satellites, the amount of raw ionospheric observation data has increased significantly, providing a rich data source for global ionospheric TEC modeling. However, existing studies have the problem of missing ocean data in the ionospheric modeling part. Since most ionospheric observation data are obtained by ground-based GNSS, the distribution of ionospheric puncture points (IPPs) in the ocean area is basically blank, which seriously distorts the accuracy of ionospheric modeling in some areas. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method, device, equipment and medium for enhancing low-orbit ionospheric data over the ocean, which can improve the observation coverage of ionospheric data over the ocean and solve the problem of missing data over the ocean in ionospheric modeling.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] In a first aspect, the present invention provides a method for enhancing ionospheric data over the ocean, comprising:

[0007] Obtain vertical total electron content data at several locations in the ionosphere over the ocean through low-orbit satellite observations;

[0008] The vertical total electron content data of several locations in the ionosphere above the ocean, as well as the latitude and longitude of the corresponding ionospheric puncture points, observation time and space weather status data are combined as input feature data;

[0009] Inputting the input feature data into a first learner to obtain enhanced feature data;

[0010] Input the input feature data and enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean;

[0011] Wherein, the first learner and the second learner are pre-trained machine learning models.

[0012] Further, obtaining the vertical total electron content data of the ionosphere over the ocean includes: meshing the ionosphere over the ocean; obtaining the vertical total electron content data of each grid of the ionosphere over the ocean.

[0013] Further, the space weather status data includes a solar activity index and a geomagnetic activity index;

[0014] Wherein, the solar activity index includes the F10.7 index; the geomagnetic activity index includes the Dst index.

[0015] Further, the observation time includes the year, day of the year, and local time when data is obtained by low-orbit satellite observation.

[0016] Further, the first learner is a random forest model; the second learner is an XGBoost model.

[0017] Further, when training the first learner and the second learner, historical vertical total electron content data of the ionosphere over the ocean obtained by low-orbit satellite observation is used as a data set, and the data set is divided into a training set, a validation set, and a test set;

[0018] The test set includes data randomly selected from the data set for one day each month and data for N days before and after a geomagnetic storm event within the time range of the data set;

[0019] The training set and the validation set are divided using the K-fold cross-validation method.

[0020] Further, it also includes: performing statistics on the residuals, biases, mean absolute deviations, root mean square errors, and correlation coefficients of the fitting results and true values obtained by the first learner and the second learner through the test set to evaluate the fitting capabilities of the first learner and the second learner.

[0021] In a second aspect, the present invention also provides an apparatus for enhancing ionospheric data over the ocean, the apparatus including:

[0022] A vertical total electron content data acquisition module, configured to obtain vertical total electron content data of several positions of the ionosphere over the ocean through low-orbit satellite observation;

[0023] An input feature data merging module, configured to merge the vertical total electron content data at several positions of the ionosphere over the ocean, as well as the longitude, latitude, observation time, and space weather status data of the corresponding ionospheric piercing points, as input feature data;

[0024] An enhanced feature data acquisition module, configured to input the input feature data into a first learner to obtain enhanced feature data;

[0025] A vertical total electron content data enhancement module, configured to input the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean;

[0026] Wherein, the first learner and the second learner are pre-trained machine learning models.

[0027] In a third aspect, the present invention further provides a computer device, including:

[0028] A memory, configured to store a computer program;

[0029] A processor, configured to execute the computer program to implement the steps of the above-mentioned ionospheric data enhancement method over the ocean.

[0030] In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and the program is executed by a processor to implement the steps of the above-mentioned ionospheric data enhancement method over the ocean.

[0031] Compared with the prior art, the beneficial effects achieved by the present invention:

[0032] The ionospheric data enhancement method over the ocean provided by the present invention performs data enhancement on the ionosphere over the ocean through an ensemble learning model, and can obtain virtual observation data of the ionosphere in unobserved areas over the ocean, effectively improving the ionospheric observation coverage over the ocean, providing more uniform ionospheric observation data for global high-precision and high-resolution ionospheric modeling, and to a certain extent solving the problem of low accuracy of ionospheric modeling over the ocean globally; the present invention selects the F10.7 index representing the solar activity index and the Dst index representing the geomagnetic activity index to construct a data set, fully considering the influence of solar activity and geomagnetic activity on the ionosphere, and putting magnetic storm event data into the data set during data set division, improving the accuracy of virtual observation data of the ionosphere over the ocean during ionospheric disturbance or active periods; using the ensemble learning method, the first learner and the second learner are independently trained respectively and then comprehensively form a stable and effective combined model to obtain a result with high credibility; multiple parameter indicators such as residual, deviation, mean absolute deviation, root mean square error, and correlation coefficient are selected to comprehensively evaluate the generation accuracy and reliability of the virtual observation data of the model, so as to improve the effect of ionospheric data enhancement over the ocean. Description of the Drawings

[0033] Figure 1 Schematic flowchart of a method for enhancing ionospheric data over the ocean provided by an embodiment of the present invention;

[0034] Figure 2 Schematic flowchart of constructing an ensemble learning model in an embodiment of the present invention;

[0035] Figure 3 Schematic diagram of the input and output of a random forest model and an XGBoost model in an embodiment of the present invention;

[0036] Figure 4 Statistical distribution diagram of residuals of the ensemble learning model on the test set in an embodiment of the present invention;

[0037] Figure 5 Statistical distribution diagram of the absolute deviation between the predicted value and the reference true value obtained by the ensemble learning model in an embodiment of the present invention using data under different solar activity conditions as the test set;

[0038] Figure 6 Schematic diagram of the correlation between the predicted value and the reference true value obtained by the ensemble learning model on the test set in an embodiment of the present invention;

[0039] Figure 7 Statistical time series diagram of the mean absolute deviation and root mean square error between the predicted value and the reference true value obtained by the ensemble learning model in an embodiment of the present invention using data during two magnetic storms as the test set;

[0040] Figure 8 Statistical distribution diagram of residuals of the ensemble learning model on another test set in an embodiment of the present invention;

[0041] Figure 9 Schematic diagram of the correlation between the predicted value and the reference true value obtained by the ensemble learning model on another test set in an embodiment of the present invention;

[0042] Figure 10 Schematic structural diagram of an ionospheric data enhancement device over the ocean provided by an embodiment of the present invention;

[0043] Figure 11 Internal structural diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0044] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The same reference numerals in the drawings denote the same or similar components or parts. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. The embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Without conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0045] The term "and / or" in this article is merely a description of the association relationship of associated 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. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0046] Embodiment 1

[0047] As Figures 1 to 9 shown, the embodiment of the present invention provides a method for enhancing ionospheric data over the ocean. Figure 1 is a flowchart of the method for enhancing ionospheric data over the ocean. This flowchart only shows the logical order of the method in this embodiment. On the premise of non-conflict, in other possible embodiments of the present invention, the steps shown or described can be completed in a different Figure 1 order than that shown.

[0048] The method for enhancing ionospheric data over the ocean provided in this embodiment can be applied to a terminal and can be executed by an ionospheric data enhancement device over the ocean. This device can be implemented in a software and / or hardware manner and can be integrated in the terminal.

[0049] Referring to Figure 1 , the method of the embodiment of the present invention specifically includes the following steps:

[0050] Step 1: Obtain vertical total electron content data at several positions of the ionosphere over the ocean through observations by low-earth orbit satellites.

[0051] Select sparse low-earth orbit satellite data that can observe the ionosphere over the ocean area, and extract ionospheric observables according to the observation methods of different satellites, that is, vertical total electron content (VTEC) data.

[0052] Taking the Jason-2 satellite as an example, this satellite uses dual-frequency signals for radar altimetry. The main frequency is the Ku band (13.575 GHz), and the auxiliary frequency is the C band (5.3 GHz). The calculation formula for obtaining vertical electron content data is as follows:

[0053] ,

[0054] Among them, is the ionospheric distance correction for the is band frequency, with the unit of GHz.

[0055] Step 2: Combine the vertical total electron content data at several positions in the ionosphere over the ocean, as well as the corresponding longitude, latitude, observation time, and space weather state data of the ionospheric piercing points as input feature data.

[0056] Among them, the vertical total electron content data at several positions in the ionosphere over the ocean needs to be preprocessed before being used as input features, including removing noise and outliers, performing coordinate transformation to unify the coordinate system, and smoothing or interpolating the data according to the sampling frequency required by the analysis.

[0057] Taking the Jason-2 satellite as an example, when selecting Jason-2 observation data over a certain time span and obtaining the vertical total electron content observation values according to the above calculation formula, it is necessary to perform median smoothing on the vertical total electron content data within 10 seconds during the data preprocessing stage to obtain the vertical total electron content observation data over the ocean with a 10s sampling rate.

[0058] The input feature data also includes the observation time, location, and space weather state data, etc. corresponding to the vertical total electron content data obtained by the low-earth orbit satellite at several positions in the ionosphere over the ocean.

[0059] The observation time includes the year, day of the year, and local time when the low-earth orbit satellite observes and obtains the vertical total electron content data at each position; the location is the longitude and latitude of the ionospheric piercing point corresponding to the vertical total electron content data obtained at each position; the space weather state data includes the solar activity index and the geomagnetic activity index. Among them, the solar activity index includes the F10.7 index, and the geomagnetic activity index includes the Dst index.

[0060] Therefore, in the embodiment of the present invention, the preprocessed vertical total electron content observation values, data observation time, longitude and latitude of the ionospheric piercing point, F10.7 index, and Dst index are combined to form a feature vector as the input feature data of the first learner and the first learner of the present invention.

[0061] Step 3: Input the input feature data into the first learner to obtain enhanced feature data.

[0062] Step 4: Input the input feature data and the enhanced feature data into the second learner to obtain the vertical total electron content data of the ionosphere over the ocean.

[0063] Among them, obtaining the vertical total electron content data of the ionosphere over the ocean includes: meshing the ionosphere over the ocean; obtaining the vertical total electron content data of each grid of the ionosphere over the ocean. The purpose of the present invention is to complete the vertical total electron content data of the remaining grids based on the integrated learning model through the vertical total electron content data of some grids of the ionosphere over the ocean, so as to realize the enhancement of the ionosphere data over the ocean.

[0064] The first learner and the second learner constitute an integrated learning model. The idea of constructing the integrated learning model in the present invention is as Figure 2 shown, which mainly includes three parts: construction and division of the data set, training and construction of the model, and result evaluation and verification.

[0065] (1) Construction and division of the data set

[0066] In the present invention, historical data of the vertical total electron content of the ionosphere over the ocean area is obtained through observations by low-earth orbit satellites as the data set. The specific acquisition and preprocessing methods of the historical data of the vertical total electron content are as shown in the above steps 1 and 2, which will not be elaborated here.

[0067] The constructed data set is divided into three parts: a training set, a validation set, and a test set. The training set is used to train the constructed integrated learning model, the validation set evaluates the performance of the model during training and feeds back to adjust the structure and parameters of the model, and the test set is used to predict the final experimental results.

[0068] The data set division rules are as follows: for the test set, it includes two parts: one part is composed of data randomly extracted from the data set for one day each month, and the other part is the data for N days before and after the geomagnetic storm event during the observation time range of the data set. For the training set and the validation set, the K-fold cross-validation method is used for division.

[0069] In the embodiment of the present invention, N is set to 5 and K is selected as 30.

[0070] (2) Training and construction of the model

[0071] In the embodiment of the present invention, the first learner uses a random forest model, and the second learner uses an XGBoost model.

[0072] The training set and the validation set in the data set are input into the integrated learning model, and the relevant parameters of the model are adjusted according to the output prediction values to obtain a relatively stable model for enhancing the observed data over the ocean. The idea of constructing the integrated model is as Figure 3 shown, and the specific implementation steps are as follows:

[0073] The random forest model serves as the first learner. The year, day of the year, local time, longitude, latitude where the piercing point of the vertical total electron content of the ionosphere is located at each moment, as well as the solar activity index of the day and the geomagnetic activity index at that moment are used as the feature vector and input into the random forest model to obtain the predicted value output by the random forest algorithm.

[0074] The XGBoost model serves as the second learner. The predicted value output by the random forest model at each moment in the previous step is used as the enhanced feature data and input into the XGBoost model together with the aforementioned features as the feature vector. The result output by the XGBoost model is the final prediction result. During the training process, the label value is the vertical total electron content calculated from the ocean altimetry satellite data at each moment. The model adjusts the internal parameters of the model using backpropagation based on the residual between the prediction result and the label value.

[0075] Regarding the adjustment and setting of hyperparameters, the grid search algorithm is used in the experiment to find the optimal value for each hyperparameter, and the optimal value of each parameter is determined in turn while fixing other parameters. The final hyperparameters are determined as follows. The parameters of the random forest model are selected as learning_rate = 0.02, n_estimators = 1000, max_features = Auto, random_state = 42. The parameters of the XGBoost model are finally selected as learning_rate = 0.05, n_estimators = 2000, max_depth = 8, min_child_weight = 5.

[0076] After determining the appropriate parameters, a relatively stable ensemble learning model is trained. The random forest model has strong generalization ability and noise resistance. The XGBoost model has a fast training speed, high prediction accuracy, and can prevent overfitting. Using these two models can better achieve the enhancement of ionospheric data over the ocean.

[0077] (3)Result Evaluation and Verification

[0078] The trained ensemble learning model is evaluated and verified through the test set, and the data during the magnetic storm period is analyzed.

[0079] To analyze the relationship between the data generated by fitting the ensemble learning model of the present invention and the reference true value (the vertical total electron content obtained from the low-earth orbit satellite), statistical indicators such as the residual, bias, mean absolute deviation (MAD), root mean square error (RMSE), and correlation coefficient (R) of the ensemble learning model on the test set are counted for evaluation.

[0080] Using the Jason-2 satellite observations as the test set, the distribution of the residuals between the results generated by fitting the model on all test sets and the reference true values is statistically analyzed. The results are as follows Figure 4 shown. It can be found that the residual results basically follow a normal distribution and are relatively concentrated, indicating that the enhanced data generated by model fitting over the ocean can reduce the observation error in the ocean area. Further analyzing the absolute deviation of the ionospheric data enhancement results over the ocean under different solar activity intensities, the obtained results are as follows Figure 5 shown. It can be seen that the distribution of the number of absolute deviations is highly correlated with the degree of solar activity, and the distribution of the quantity size coincides with the solar activity cycle, indicating once again that the results generated by model fitting are reliable.

[0081] Using the Jason-2 satellite observations as the test set, the correlation between the results generated by fitting the ensemble learning model and the reference true values is evaluated. The results are as follows Figure 6 shown. It can be seen that the correlations between the model fitting results and the reference true values are all relatively high, further demonstrating the stability of the ensemble learning model.

[0082] During geomagnetic storms, the data generated using the ensemble learning model of the present invention requires targeted processing, such as introducing a magnification factor method. Using the Jason-2 satellite observations as the test set, Figure 7 shows the statistical time series graphs of the mean absolute deviation and root mean square error of the results generated by fitting the ensemble learning model during two geomagnetic storms. It can be seen that compared with the missing data in the ocean area, the enhanced data generated by the model can be accepted by controlling the error within a limited range.

[0083] To verify the generalization ability of the ensemble model of the present invention, the ensemble learning model is used to verify the observation data of the Jason-3 satellite in the same series as the above Jason-2 satellite as the test set. The obtained results are as follows Figure 8 and Figure 9 shown. It can be seen that the residual results also basically follow a normal distribution, and the correlation between the data enhancement results generated by the ensemble model and the reference true values is relatively high, indicating that the data enhancement model of the present invention has strong generalization ability.

[0084] Embodiment 2

[0085] Based on the same inventive concept as Embodiment 1, the present invention also provides an ionospheric data enhancement device over the ocean for implementing the above ionospheric data enhancement method over the ocean. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in the following embodiments of the ionospheric data enhancement device over the ocean can refer to the limitations on the ionospheric data enhancement method over the ocean in the above text, and will not be repeated here.

[0086] AsFigure 10 As shown in the figure, an ionospheric data enhancement device over the ocean according to an embodiment of the present invention includes:

[0087] A vertical total electron content data acquisition module for acquiring vertical total electron content data at several positions in the ionosphere over the ocean through observations by low-earth orbit satellites;

[0088] An input feature data merging module for merging the vertical total electron content data at several positions in the ionosphere over the ocean, as well as the longitude, latitude, observation time, and space weather state data of the corresponding ionospheric piercing points, as input feature data;

[0089] An enhanced feature data acquisition module for inputting the input feature data into a first learner to obtain enhanced feature data;

[0090] A vertical total electron content data enhancement module for inputting the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean;

[0091] Wherein, the first learner and the second learner are pre-trained machine learning models.

[0092] Embodiment 3

[0093] An embodiment of the present invention further provides a computer device, which may be a server, and its internal structure diagram may be as shown in Figure 11 the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with external terminals through a network connection. When the computer program is executed by the processor, it realizes the steps of the ionospheric data enhancement method over the ocean in the foregoing embodiments.

[0094] Those skilled in the art can understand that Figure 11 the structure shown in the figure is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0095] Example 4

[0096] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the following method are implemented:

[0097] Obtain vertical total electron content data at several positions in the ionosphere over the ocean through observations by low-earth orbit satellites;

[0098] Merge the vertical total electron content data at several positions in the ionosphere over the ocean, as well as the corresponding longitude, latitude, observation time, and space weather state data of the ionospheric piercing points, as input feature data;

[0099] Input the input feature data into a first learner to obtain enhanced feature data;

[0100] Input the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean;

[0101] Wherein, the first learner and the second learner are pre-trained machine learning models.

[0102] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0103] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0104] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the flow Figure 1 one or more flows and / or blocks Figure 1 specified in the blocks.

[0105] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the flow Figure 1 one or more flows and / or blocks Figure 1 specified in the blocks.

[0106] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope of the present invention as protected by the claims. All of these are within the protection scope of the present invention.

Claims

1. An ionospheric data enhancement method over the ocean, characterized in that, Comprising: Obtaining vertical total electron content data at several positions in the ionosphere over the ocean through observations by low-Earth orbit satellites; Combining the vertical total electron content data at several positions in the ionosphere over the ocean, as well as the longitude, latitude, observation time, and space weather status data of the corresponding ionospheric piercing points, as input feature data; Inputting the input feature data into a first learner to obtain enhanced feature data; Inputting the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data of the ionosphere over the ocean; Wherein, the first learner and the second learner are pre-trained machine learning models; The first learner is a random forest model; the second learner is an XGBoost model; During the training of the first learner and the second learner, historical vertical total electron content data of the ionosphere over the ocean is obtained through observations by low-Earth orbit satellites as a data set, and the data set is divided into a training set, a validation set, and a test set; The test set includes data randomly selected one day per month from the data set and data for N days before and after geomagnetic storm events within the time range of the data set; The training set and the validation set are divided using the K-fold cross-validation method; Wherein, the vertical total electron content data at several positions in the ionosphere over the ocean needs to be pre-processed before being used as input features, including removing noise and outliers, performing coordinate transformation to unify the coordinate system, and smoothing or interpolating the data according to the sampling frequency required for analysis; Using the random forest model as the first learner, inputting the year, day of the year, local time, longitude, latitude where the ionospheric vertical total electron content piercing point is located at each moment, as well as the solar activity index on the current day and the geomagnetic activity index at the current moment as feature vectors into the random forest model to obtain the predicted values output by the random forest algorithm; Using the XGBoost model as the second learner, inputting the predicted values output by the random forest model at each moment in the previous step as enhanced feature data together with the input features into the XGBoost model as feature vectors, and the result output after passing through the XGBoost model is the final prediction result; During the training process, the label value is the vertical total electron content calculated from ocean altimetry satellite data at each moment, and the XGBoost model adjusts the internal parameters of the random forest model using backpropagation based on the residuals between the prediction result and the label value.

2. The method for enhancing ionospheric data over the ocean according to claim 1, characterized in that, Obtaining the vertical total electron content data of the ionosphere over the ocean includes: gridifying the ionosphere over the ocean; obtaining the vertical total electron content data of each grid in the ionosphere over the ocean.

3. The method for enhancing ionospheric data over the ocean according to claim 1, wherein The space weather status data includes the solar activity index and the geomagnetic activity index; Wherein, the solar activity index includes the F10.7 index; the geomagnetic activity index includes the Dst index.

4. The method for enhancing ionospheric data over the ocean according to claim 1, wherein The observation time includes the year, day of the year, and local time when the low-Earth orbit satellite observes and obtains data.

5. The method for enhancing ionospheric data over the ocean according to claim 1, wherein Also comprising: Statistically analyzing the residuals, biases, mean absolute deviations, root mean square errors, and correlation coefficients of the fitting results and the true values obtained by the first learner and the second learner through the test set to evaluate the fitting capabilities of the first learner and the second learner.

6. An apparatus for enhancing low-orbit ionospheric data over the ocean, characterized in that, Comprising: A vertical total electron content data acquisition module for obtaining vertical total electron content data at several positions in the ionosphere over the ocean through observations by low-orbit satellites; An input feature data merging module for merging the vertical total electron content data at several positions in the ionosphere over the ocean, as well as the longitude, latitude, observation time, and space weather status data of the corresponding ionospheric piercing points as input feature data; An enhanced feature data acquisition module for inputting the input feature data into a first learner to obtain enhanced feature data; A vertical total electron content data enhancement module for inputting the input feature data and the enhanced feature data into a second learner to obtain the vertical total electron content data in the ionosphere over the ocean; Wherein, the first learner and the second learner are pre-trained machine learning models; The first learner is a random forest model; the second learner is an XGBoost model; During the training of the first learner and the second learner, historical vertical total electron content data in the ionosphere over the ocean is obtained through observations by low-orbit satellites as a data set, and the data set is divided into a training set, a validation set, and a test set; The test set includes data randomly selected one day per month from the data set and data for N days before and after geomagnetic storm events during the time range of the data set; The training set and the validation set are divided using the K-fold cross-validation method; Wherein, the vertical total electron content data at several positions in the ionosphere over the ocean needs to be pre-processed before being used as input features, including removing noise and outliers, performing coordinate transformation to unify the coordinate system, and smoothing or interpolating the data according to the sampling frequency required by the analysis; The random forest model, as the first learner, inputs the year, day of the year, local time, longitude, latitude where the ionospheric vertical total electron content piercing point is located at each moment, as well as the solar activity index on the current day and the geomagnetic activity index at the current moment as feature vectors into the random forest model to obtain the predicted values output by the random forest algorithm; The XGBoost model, as the second learner, inputs the predicted values output by the random forest model at each moment in the previous step as enhanced feature data together with the input features into the XGBoost model as feature vectors, and the result output after passing through the XGBoost model is the final prediction result; During the training process, the label value is the vertical total electron content calculated from the ocean altimetry satellite data at each moment, and the XGBoost model adjusts the internal parameters of the random forest model using backpropagation based on the residuals between the prediction result and the label value.

7. A computer device, characterized in that, Including: A memory for storing computer programs; A processor for executing the computer program to implement the steps of the method for enhancing low-orbit ionospheric data over the ocean according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method for enhancing low-orbit ionospheric data over the ocean according to any one of claims 1 to 5.

Images