Ionospheric TEC Determination Method Based on Joint Inversion of Space-Ground Active Measurement and SAR Images

Abstract

The present invention relates to a method and system for determining ionospheric TEC based on joint inversion of space-ground active measurement and SAR image. The method includes: Step S1, establishing a mapping relationship database of the ionospheric scattering characteristics measured by a dual-frequency SAR satellite, the ionospheric penetration characteristics measured by space-ground active measurement, and the ionospheric TEC measured by a GPS receiving station; Step S2, according to the mapping relationship database, using a multi-source amplitude-phase parameter neural network to establish and train a TEC inversion deep learning model, the TEC inversion deep learning model takes the ionospheric amplitude-phase parameters as inputs and outputs the corresponding TEC values; the ionospheric amplitude-phase parameters are the ionospheric scattering signal amplitude-phase parameters and / or the ionospheric transmission signal amplitude-phase parameters; Step S3, performing on-orbit intelligent and rapid inversion of the ionospheric TEC over the globe according to the TEC inversion deep learning model. The present invention establishes a TEC inversion model based on SAR satellite measurement data, which has the advantages of high accuracy and high timeliness, and can achieve high-precision inversion of TEC at any time and space.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing technology, and in particular, to a method for determining ionospheric TEC based on joint inversion of space-ground active measurement and SAR images. Background Art

[0002] In satellite remote sensing, the satellite altitude is usually above 500 km, and the orbital altitude of a geosynchronous orbit SAR reaches 36,000 km. Satellite SAR imaging usually actively emits electromagnetic waves and receives echo signals for processing and imaging. Usually, the signal passes through the ionosphere on the Earth's surface, and both the background ionosphere and ionospheric irregularities will seriously affect the imaging quality of spaceborne SAR.

[0003] The influence of the ionosphere on SAR imaging is mainly reflected in three aspects: First, the "scintillation" effect. Ionospheric scintillation causes a large change in the amplitude of electromagnetic wave transmission, affecting the graphic radiation accuracy and consistency. Second, the "dispersion" effect. The ionosphere will cause the propagation path of electromagnetic waves to deflect and the transmission time delay to be inconsistent, thus seriously causing changes in the frequency and amplitude of remote sensing radar signals, affecting the spatial resolution, geometric distortion, and even image performance such as the peak sidelobe ratio and integrated sidelobe ratio of SAR. Third, the "Faraday rotation". For a linearly polarized remote sensing radar, the Faraday rotation effect will cause a change in the polarization of the electromagnetic wave, resulting in polarization mismatch between transmission and reception, affecting the amplitude of the received signal, and ultimately affecting the quality of the SAR image. The ionosphere will also affect remote sensing systems such as non-imaging microwave radiometers, microwave scatterometers, and radar altimeters in terms of measurement accuracy and consistency.

[0004] Therefore, it is necessary to accurately measure and precisely compensate the ionosphere. Traditional ionospheric measurement methods usually adopt two ways: One is to directly detect the ionosphere through independent equipment, invert the ionospheric information, and then use the obtained ionospheric information to correct and compensate the SAR image. The other is to invert the ionospheric information using the SAR echo characteristics. However, for high-resolution, long integration time, or repeat-pass interferometric imaging, the above methods have problems such as incomplete ionospheric paths, incomplete information, insufficient timeliness, and even large deviations in the space and time of remote sensing signals, making it difficult to meet the requirements of high-precision and high-timeliness ionospheric measurement. New measurement and inversion methods must be adopted. Summary of the Invention

[0005] In order to solve the above technical problems existing in the prior art, the object of the present invention is to provide a method and system for determining ionospheric TEC based on joint inversion of space-ground active measurement and SAR images, which have the advantages of high accuracy and high timeliness, and can achieve high-precision inversion of TEC at any time and space.

[0006] To achieve the above-mentioned invention objective, a method for determining ionospheric TEC based on the joint inversion of space-ground active measurement and SAR images provided by the present invention includes:

[0007] Step S1: Establish a mapping relationship database for the ionospheric scattering characteristics measured by a dual-frequency SAR satellite, the ionospheric penetration characteristics measured by space-ground active measurement, and the ionospheric TEC measured by a GPS receiving station;

[0008] Step S2: According to the mapping relationship database, use a multi-source amplitude-phase parameter neural network to establish and train a TEC inversion deep learning model. The TEC inversion deep learning model takes the ionospheric amplitude-phase parameters as input and outputs the corresponding TEC values; the ionospheric amplitude-phase parameters are the ionospheric scattering signal amplitude-phase parameters and / or the ionospheric transmission signal amplitude-phase parameters;

[0009] Step S3: Perform on-orbit intelligent and rapid inversion of the ionospheric TEC over the globe according to the TEC inversion deep learning model.

[0010] According to a technical solution of the present invention, in the step S1, the mapping relationship database includes Q sample data, and each sample data includes the ionospheric scattering signal amplitude-phase parameters measured by a dual-frequency SAR satellite, the ionospheric transmission signal amplitude-phase parameters measured by space-ground active measurement, and the ionospheric TEC values measured by a GPS receiving station for the ionosphere at the same time and space;

[0011] The sample data of the mapping relationship database is expressed as Among them, is the ionospheric scattering signal amplitude-phase parameter measured by a dual-frequency SAR satellite, is the ionospheric transmission signal amplitude-phase parameter measured by space-ground active measurement, Γ q is the ionospheric TEC value measured by a GPS receiving station; A and B respectively represent the signal amplitudes of the ionospheric scattering signal and the ionospheric transmission signal; respectively represent the signal phases; m represents the signal frequency, m = 1 or 2, when m = 1, it represents the P band, and when m = 2, it represents the L band; n represents the signal polarization mode, n = 1, 2, 3, or 4, when n = 1, it is VV polarization, when n = 2, it is VH polarization, when n = 3, it is HV polarization, and when n = 4, it is HH polarization; q represents the sample number, q = 1, 2, 3,..., Q; d represents the signal propagation direction of the ionospheric transmission signal, d = 1 represents satellite transmission and ground reception, and d = 2 represents ground transmission and satellite reception.

[0012] According to a technical solution of the present invention, in the step S1, it specifically includes:

[0013] Step S11: Conduct joint measurements using a ground transceiver station and a dual-frequency SAR satellite to obtain multiple amplitude-phase parameter samples of ionospheric transmission signals, and establish a sub-database of ionospheric transmission signals for space-ground active measurement; the sample data in the sub-database of ionospheric transmission signals for space-ground active measurement is represented as

[0014] Step S12: Use the dual-frequency SAR satellite to measure and obtain multiple amplitude-phase parameter samples of ionospheric scattering signals that are co-spatiotemporal with the ionospheric transmission signals in the space-ground active measurement in Step S11, and establish a sub-database of dual-frequency SAR ionospheric scattering signals; the sample data in the sub-database of dual-frequency SAR ionospheric scattering signals is represented as

[0015] Step S13: Use a GPS receiving station to obtain multiple ionospheric TEC values that are co-spatiotemporal with the ionospheric transmission signals in the space-ground active measurement in Step S11, and establish a sub-database of TEC values; the samples in the sub-database of TEC values are denoted as Γ q , and the measurement accuracy of the samples in the sub-database of TEC values reaches 0.5 TECU;

[0016] Step S14: Establish the mapping relationship database according to the sub-database of ionospheric transmission signals for space-ground active measurement, the sub-database of dual-frequency SAR ionospheric scattering signals, and the sub-database of TEC values.

[0017] According to one technical solution of the present invention, the duration of the space-ground active measurement using a ground transceiver station and a dual-frequency SAR satellite is τ1, and τ1 ≤ 20 ms;

[0018] The duration of the measurement time slot of the space-ground active measurement is τ2, and τ2 ≤ 200 ms;

[0019] The duration of using the dual-frequency SAR satellite to measure ionospheric scattering signals is τ3, and τ3 ≤ 150 ms.

[0020] According to one technical solution of the present invention, in Step S2, it specifically includes:

[0021] Step S21: Construct a multi-source amplitude-phase parameter neural network to establish a TEC inversion deep learning model, and the TEC inversion deep learning model includes a feature extraction network, a feature fusion network, and a TEC regression network;

[0022] Step S22: According to the mapping relationship database, obtain the amplitude-phase parameters s of the ionospheric scattering signals of satellite transmission and ground reception and ground transmission and satellite reception in the sample data 1 and s 3 , and input them into the feature extraction network to obtain the extracted feature map s' 1 =Net(s1 ), s' 3 = Net(s 3 );

[0023] Step S23: According to the mapping relationship database, obtain the amplitude-phase parameters of the ionospheric scattering signal of the dual-frequency SAR satellite's self-transmission and self-reception in the sample data and input them into the feature extraction network to obtain the extracted feature map s' 2 = Net(s 2 );

[0024] Step S24: At the same time, input the feature maps s' 1 , s' 2 and s' 3 extracted by the feature extraction network into the feature fusion network for feature stitching and deep feature fusion;

[0025] Step S25: Perform TEC estimation on the fused deep features through the TEC regression network to obtain the TEC estimation value;

[0026] Step S26: According to the obtained TEC estimation value and the ionospheric TEC value measured by the GPS receiving station in the sample data, estimate the inversion result and perform backpropagation to update the network parameters;

[0027] Step S27: Repeat the above steps S22 to S26 for several rounds of training, save and output the TEC inversion deep learning model;

[0028] In the training stage, set the learning rate to 0.01, the optimizer to SGD, train for 100 rounds, estimate the TEC inversion effect with the mean square error as the loss function and perform backpropagation to update the network parameters of the TEC prediction deep learning module;

[0029] where, Γ k is the TEC value measured by the GPS receiving station and the ground equipment, Γ' k is the TEC estimation value of the intelligent inversion model, k is a variable representing the batch input for a single forward training, and K represents the size of the batch value.

[0030] According to a technical solution of the present invention, the feature extraction network includes a convolutional layer conv1, a first batch normalization layer, a first activation function, a convolutional layer conv2, a second batch normalization layer, and a second activation function connected in series in sequence;

[0031] where the convolutional kernel size of the convolutional layer conv1 is (1, 1), the stride is 1, the padding is 0, the number of input channels is 8, and the number of output channels is set to 64;

[0032] The first activation function and the second activation function are the rectified linear unit relu = max(0, x 1 ), where max() represents the maximum value between 0 and x 1 , and x 1 is an intermediate value generated during the forward process of the neural network;

[0033] The convolution kernel size of the convolutional layer conv2 is (1, 1), the stride is 1, the padding is 0, the number of input channels is 64, and the number of output channels is set to 128;

[0034] The calculation methods of the first batch normalization layer and the second batch normalization layer are x 1 ' = γ * x norm + β, where x norm is the normalized feature, and γ and β are two learnable feature linear transformation parameters, where ∈ is an adjustment factor to prevent the denominator from being 0, and its value is set to 0.0001; is the mean, k is the batch size input for a single forward training, is the variance, and x k represents the intermediate value generated during the forward process of the neural network in the k-th batch of a single forward training;

[0035] In the step S24, the feature concatenation method is channel connection, expressed as s” = concat(s' 1 , s' 2 , s' 3 ), where concat is the channel connection function; the deep feature fusion is performed through a deep fusion convolutional layer, expressed as s'” = conv(s”), where conv is the deep fusion convolutional layer, the convolution kernel size is (1, 1), the stride is 1, the padding is 0, the number of input channels is 128, and the number of output channels is set to 64;

[0036] The number of input channels of the deep fusion convolutional layer is the same as the number of channels of the feature map s” obtained after the feature concatenation, where s' 1 and s' 3 are respectively the feature maps of the amplitude-phase parameters of the ionospheric scatter signals received from the satellite and sent from the ground to the satellite. Before feature fusion, there is a probability of z% that they are replaced by Gaussian noise to adapt to the situation where only the feature input of the self-transmitted and self-received signals is available during the on-orbit prediction stage. The Gaussian noise is a normal distribution with a mean of 0 and a variance of 1.

[0037] According to a technical solution of the present invention, in the step S3, in the on-orbit intelligent fast inversion stage, the input of the feature fusion network is the feature map extracted from the ionospheric scattering signal measured by the dual-frequency SAR satellite after passing through the feature extraction network and the randomly generated Gaussian signal, and the feature map extracted after the feature extraction network and the randomly generated Gaussian signal are feature-fused and then deeply feature-fused through the deep fusion convolutional layer. The number of input channels of the deep fusion convolutional layer is the same as the number of channels of the feature map extracted after passing through the feature extraction network.

[0038] According to a technical solution of the present invention, in the step S3, it specifically includes:

[0039] Step S31: Obtain the scattering echo signal of the dual-frequency SAR satellite and process it to obtain the ionospheric scattering amplitude-phase parameters of the scattering echo signal of the dual-frequency SAR satellite;

[0040] Step S32: Perform standardized preprocessing on the extracted ionospheric scattering amplitude-phase parameters of the scattering echo signal of the dual-frequency SAR satellite and input them into the TEC inversion deep learning model;

[0041] Step S33: Perform on-orbit intelligent fast inversion on the ionospheric scattering amplitude-phase parameters of the scattering echo signal of the dual-frequency SAR satellite through the TEC inversion deep learning model to obtain the ionospheric TEC value Γ' over the observation area of the dual-frequency SAR satellite. q 。

[0042] According to an aspect of the present invention, a system for implementing the above-mentioned ionospheric TEC determination method based on space-ground active measurement and SAR image joint inversion includes:

[0043] An FPGA for receiving and obtaining the scattering echo signal of the dual-frequency SAR satellite and preprocessing the scattering echo signal to obtain its ionospheric scattering amplitude-phase parameters;

[0044] A GPU platform for loading and configuring the TEC inversion deep learning model and inversing the ionospheric TEC value over the measurement position of the dual-frequency SAR satellite through the TEC inversion deep learning model.

[0045] The present invention proposes an ionospheric TEC determination method and system based on space-ground active measurement and SAR image joint inversion, which has the following advantages:

[0046] The present invention simultaneously measures the amplitude-phase parameters of electromagnetic wave scattering and the amplitude-phase parameters of electromagnetic wave transmission of the ionosphere by using a dual-frequency SAR satellite and a ground transceiver station, improves the description ability of the propagation characteristics of electromagnetic waves in a complex ionosphere, and uses the amplitude-phase parameters of electromagnetic wave scattering and the amplitude-phase parameters of electromagnetic wave transmission of the ionosphere to train and optimize a large number of measured data samples by using a deep learning network model to obtain a high-precision TEC inversion model. Combining multiple amplitude-phase parameters of scattering signals measured by the dual-frequency SAR satellite, the TEC inversion of any region is realized, the inversion accuracy of TEC over areas where stations cannot be deployed, such as the open sea and remote regions, is significantly improved, and it has the advantages of high accuracy and high timeliness, and can realize the high-precision inversion of TEC at any time and space. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings without creative efforts based on these drawings.

[0048] Figure 1 Schematically showing a flowchart of a method for determining ionospheric TEC based on space-ground active measurement and SAR image joint inversion provided in an embodiment of the present invention;

[0049] Figure 2 Schematically showing a specific flowchart of a method for determining ionospheric TEC based on space-ground active measurement and SAR image joint inversion provided in an embodiment of the present invention;

[0050] Figure 3 Schematically showing a time arrangement diagram of active measurement and SAR imaging according to an embodiment of the present invention;

[0051] Figure 4 Schematically showing a working flowchart of a TEC inversion deep learning model according to an embodiment of the present invention;

[0052] Figure 5 Schematically showing a working flowchart of a feature extraction network according to an embodiment of the present invention;

[0053] Figure 6 Schematically showing a working flowchart of a feature fusion network according to an embodiment of the present invention;

[0054] Figure 7 Schematically showing a flowchart of on-orbit intelligent inversion processing of ionospheric TEC based on a hybrid architecture of GPU and FPGA according to another embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The description of the embodiments of this specification should be combined with the corresponding drawings, which should be part of the complete specification. In the drawings, the shape or thickness of the embodiments may be enlarged, and simplified or convenient labeling may be used. Furthermore, the parts of each structure in the drawings will be described separately. It should be noted that the elements not shown or described in words in the drawings are in forms known to those of ordinary skill in the art.

[0056] Any reference to directions and orientations in the description of the embodiments herein is for convenience of description only and should not be construed as any limitation on the scope of protection of the present invention. The following description of the preferred embodiments will involve combinations of features, which may exist independently or in combination. The present invention is not particularly limited to the preferred embodiments. The scope of the present invention is defined by the claims.

[0057] As Figure 1 shown, a method for determining ionospheric TEC based on the joint inversion of space-ground active measurement and SAR images provided by the present invention includes three steps: database building 10, modeling 20, and inversion 30.

[0058] During the database building 10 process, the amplitude-phase parameters of the ionospheric scattering signal are measured using a dual-frequency SAR satellite, and at the same time, the amplitude-phase parameters of the ionospheric transmission signal of the space-ground active measurement are jointly measured using a ground transceiver station and a dual-frequency SAR satellite. The ionospheric TEC value is measured using a GPS receiving station, and a mapping relationship database of the amplitude, phase, and TEC of the complex propagation of electromagnetic waves in the ionosphere is established. The database building 10 includes several parts: the dual-frequency SAR satellite obtains the scattering signal parameters, the dual-frequency SAR satellite and the ground transceiver station jointly obtain the transmission signal parameters, and the GPS receiving station obtains the ionospheric TEC value.

[0059] During the modeling 20 process, considering the complex data such as the measured ionospheric scattering signal and transmission signal, it is impossible to obtain the accurate ionospheric TEC value through an analytical method. Using a multi-source amplitude-phase parameter neural network, with a total of 24 amplitude-phase parameters of satellite transmitting and ground receiving, ground transmitting and satellite receiving, and satellite self-transmitting and self-receiving as inputs, the mapping relationship between the ionospheric scattering signal, transmission signal, and ionospheric TEC is approximated, providing a TEC inversion deep learning model for high-precision measurement of the ionospheric TEC over the global sky. The modeling 20 includes several parts: multi-source feature extraction, multi-source feature fusion, and model training and saving.

[0060] During the inversion 30, for the TEC inversion deep learning model, a hybrid hardware platform of FPGA and GPU is used to load and configure the model. The scattered signal amplitude and phase of the ionosphere in any space of the globe are obtained by a dual-frequency SAR satellite as 8 parameters as the input of the model, and the TEC value of the SAR satellite observation time and space is inverted. The inversion 30 includes several parts: input of the amplitude-phase parameters of the scattered signal of the dual-frequency SAR satellite, configuration of the deep learning model on the hybrid platform, and output of the TEC of the ionosphere above the measurement position of the dual-frequency SAR satellite.

[0061] Compared with the existing ionospheric TEC measurement methods, this method comprehensively considers the complex relationship between the amplitude-phase characteristics of the ionospheric scattered signal, the amplitude-phase characteristics of the penetrating signal, and the TEC, and optimizes the model using a large amount of measured data, having the advantages of high TEC measurement accuracy, good robustness, and large measurement range.

[0062] As Figure 2 shown, the specific process of a method for determining ionospheric TEC based on joint inversion of space-ground active measurement and SAR image of the present invention includes:

[0063] Step S1, establish a mapping relationship database of the ionospheric scattering characteristics measured by the dual-frequency SAR satellite, the ionospheric penetration characteristics measured by the space-ground active measurement, and the ionospheric TEC measured by the GPS receiving station;

[0064] In the step S1, the mapping relationship database contains Q sample data, generally Q≥20000, and each sample data includes 8 parameters of the amplitude-phase parameters of the ionospheric scattered signal measured by the dual-frequency SAR satellite for the ionosphere at the same time and space, 24 parameters of the amplitude-phase parameters of the ionospheric transmitted signal measured by the space-ground active measurement, and the ionospheric TEC value measured by the GPS receiving station;

[0065] The sample data of the mapping relationship database is expressed as Among them, is the amplitude-phase parameter of the ionospheric scattered signal measured by the dual-frequency SAR satellite, is the amplitude-phase parameter of the ionospheric transmitted signal measured by the space-ground active measurement, Γ q is the ionospheric TEC value measured by the GPS receiving station; A and B respectively represent the signal amplitudes of the ionospheric scattered signal and the ionospheric transmitted signal; respectively represent the signal phases; m represents the signal frequency, m = 1 or 2, when m = 1, it represents the P band, when m = 2, it represents the L band; n represents the signal polarization mode, n = 1, 2, 3 or 4, when n = 1, it is VV polarization, when n = 2, it is VH polarization, when n = 3, it is HV polarization, when n = 4, it is HH polarization; q represents the sample number, q = 1, 2, 3,..., Q; d represents the signal propagation direction of the ionospheric transmitted signal, d = 1 represents satellite transmission and ground reception, d = 2 represents ground transmission and satellite reception.

[0066] During the construction of each sample, the TEC is approximately kept constant. The dual-frequency SAR satellite and the space-ground joint detection repeat the measurement of the same ionospheric region H times, and the average of the H measurement results is taken and denoted as a set of sample data.

[0067] In the step S1, it specifically includes:

[0068] Step S11: Use the ground transceiver station and the dual-frequency SAR satellite for joint measurement to obtain multiple amplitude-phase parameter samples of the ionospheric transmission signal, and establish a sub-database of the ionospheric transmission signal for space-ground active measurement; the sample data in the sub-database of the ionospheric transmission signal for space-ground active measurement is expressed as

[0069] During the construction of the q-th sample, the TEC is approximately kept constant. The dual-frequency SAR satellite and the ground station repeat the measurement of the same ionospheric region H times, and the average of the H measurement results is taken and denoted as a set of sample data to improve the signal-to-noise ratio, denoted as where

[0070] During this process, the orbital altitude of the dual-frequency SAR satellite is about 36000 km. The dual frequencies include the P band and the L band. The center frequency of the P band is 435 MHz, and the center frequency of the L band is 1.26 GHz. The polarizations include VV, VH, HV, and HH.

[0071] Step S12: Use the dual-frequency SAR satellite measurement to obtain multiple amplitude-phase parameter samples of the ionospheric scattering signal that are co-spatiotemporal with the ionospheric transmission signal in the space-ground active measurement in step S11, and establish a sub-database of the dual-frequency SAR ionospheric scattering signal; the sample data in the sub-database of the dual-frequency SAR ionospheric scattering signal is expressed as

[0072] During the construction of the q-th sample, the TEC is approximately kept constant. The dual-frequency SAR satellite repeats the measurement of the same ionospheric region H times, and the average of the H measurement results is taken and denoted as a set of sample data to improve the signal-to-noise ratio, denoted as where

[0073] During this process, the orbital altitude of the dual-frequency SAR satellite is about 36000 km. The dual frequencies include the P band and the L band. The center frequency of the P band is 435 MHz, and the center frequency of the L band is 1.26 GHz. The polarizations include VV, VH, HV, and HH.

[0074] As Figure 3 shown, Figure 3For the time arrangement of space - to - ground active measurement 1021 and SAR imaging measurement 1011. In steps S21 and S22, the duration of the space - to - ground active measurement 1021 carried out by the ground transceiver station and the dual - frequency SAR satellite is τ1, τ1 ≤ 20 ms; the duration of the measurement time slot 1022 of the space - to - ground active measurement is τ2, τ2 ≤ 200 ms; the duration of the SAR imaging measurement 1011 carried out by the dual - frequency SAR satellite for measuring the ionospheric scattering signal is τ3, τ3 ≤ 150 ms.

[0075] Step S13: Use the GPS receiving station to obtain multiple ionospheric TEC values that are co - temporal and co - spatial with the ionospheric transmission signals in the space - to - ground active measurement in step S11, and establish a TEC value sub - database; the samples in the TEC value sub - database are denoted as Γ q , and the measurement accuracy of the samples in the TEC value sub - database reaches 0.5 TECU.

[0076] Step S14: According to the ionospheric transmission signal sub - database of the space - to - ground active measurement, the dual - frequency SAR ionospheric scattering signal sub - database, and the TEC value sub - database, establish the mapping relationship database.

[0077] Step S2: According to the mapping relationship database, use the multi - source amplitude - phase parameter neural network to establish and train a TEC inversion deep - learning model. The TEC inversion deep - learning model takes the ionospheric amplitude - phase parameters as inputs and outputs the corresponding TEC values; the ionospheric amplitude - phase parameters are the ionospheric scattering signal amplitude - phase parameters and / or the ionospheric transmission signal amplitude - phase parameters.

[0078] As Figure 4 shown, the modeling process is mainly to establish a TEC inversion deep - learning model based on the mapping relationship database. The amplitude - phase parameters of the signals of satellite - to - ground receive, ground - to - satellite receive, and self - transmit and self - receive can pass through the feature extraction network 201 to obtain the extracted feature map s’, and then be connected and fused through the feature fusion network 202 to obtain the fused feature s”. Finally, the TEC regression network 203 completes the TEC estimation to obtain the TEC estimation result t’, and performs backpropagation to update the network parameters to realize the training of the TEC inversion deep - learning model.

[0079] In step S2, it specifically includes:

[0080] Step S21: Construct a TEC inversion deep - learning model using the multi - source amplitude - phase parameter neural network. The TEC inversion deep - learning model includes a feature extraction network 201, a feature fusion network 202, and a TEC regression network 203;

[0081] Step S22: According to the mapping relationship database, obtain the ionospheric scattering signal amplitude - phase parameters s of satellite - to - ground receive and ground - to - satellite receive in the sample data1 and s 3 and input it into the feature extraction network 201 to obtain the extracted feature map s' 1 = Net(s 1 ), s' 3 = Net(s 3 );

[0082] Step S23: According to the mapping relationship database, obtain the amplitude-phase parameters s of the ionospheric scattering signal of the dual-frequency SAR satellite's self-transmission and self-reception in the sample data 2 Input it into the feature extraction network 201 to obtain the extracted feature map s' 2 = Net(s 2 );

[0083] In step S22 and step S23, the feature extraction network 201 includes a convolution layer conv12011, a first batch normalization layer 2012, a first activation function 2013, a convolution layer conv22014, a second batch normalization layer 2015, and a second activation function 2016 connected in series in sequence

[0084] Among them, the convolution kernel size of the convolution layer conv1 is (1,1), the stride is 1, the padding is 0, the number of input channels is 8, and the number of output channels is set to 64

[0085] The first activation function and the second activation function are the rectified linear unit relu = max(0, x 1 ), where max() represents the maximum value between 0 and x 1 , and x 1 is the intermediate value generated during the forward process of the neural network

[0086] The convolution kernel size of the convolution layer conv2 is (1,1), the stride is 1, the padding is 0, the number of input channels is 64, and the number of output channels is set to 128

[0087] The calculation methods of the first batch normalization layer 2012 and the second batch normalization layer 2015 are x 1 ' = γ * x norm + β, where x norm is the regularized feature, and γ, β are two learnable feature linear transformation parameters Among them, ∈ is an adjustment factor to prevent the denominator from being 0, and its value is set to 0.0001 is the mean, k is a variable representing the batch input for a single forward training is the variance, K represents the size of the batch value input for a single forward training, and x k represents the intermediate value generated during the forward process of the neural network in the k-th batch of a single forward training

[0088] Step S24: Meanwhile, the feature maps s' 1 , s' 2 , s' 3 extracted by the feature extraction network 201 are input into the feature fusion network 202 for feature concatenation and deep feature fusion to output the fused features.

[0089] In step S24, the way of feature concatenation is channel connection, denoted as s” = concat(s' 1 , s' 2 , s' 3 ), where concat is the channel connection function; deep feature fusion is performed through the deep fusion convolutional layer, denoted as s''' = conv(s”), where conv is the deep fusion convolutional layer, the convolutional kernel size is (1,1), the stride is 1, the padding is 0, the number of input channels is 128, and the number of output channels is set to 64;

[0090] The number of input channels of the deep fusion convolutional layer 2024 is the same as the number of channels of the feature map s” obtained after the feature concatenation, where s' 1 is the feature extraction result 2021 of star - to - ground receive, s' 3 is the feature extraction result 2022 of ground - to - star receive, and it has a probability of z% of being replaced by Gaussian noise before feature fusion to adapt to the situation where only the self - generated and self - received signal features are input in the on - orbit prediction stage, where the noise is a normal distribution with a mean of 0 and a variance of 1.

[0091] In the on - orbit prediction stage, the input of the feature fusion network is the feature map s' 2 extracted from the self - generated and self - received ionospheric signal by the feature extraction network, 1 , and the randomly generated Gaussian signal g' 3 , g' 1 , and feature fusion is performed, s” = concat(g' 2 , s' 3 ), and then deep feature fusion is performed through the deep fusion convolutional layer 2024, s''' = conv(s”), to generate the feature map 2025 s''' after feature fusion. The number of input channels of the convolutional layer is the same as the number of channels of s”, and the network parameters remain unchanged.

[0092] Step S25: Estimate the TEC through the TEC regression network 203 for the fused deep features to obtain the TEC estimation value;

[0093] The TEC regression network 203 consists of two fully - connected layers.

[0094] Step S26: Estimate the inversion result based on the obtained TEC estimation value and the ionospheric TEC value measured by the GPS receiving station in the sample data, and perform backpropagation to update the network parameters;

[0095] Step S27: Repeat the above steps S22 to S26 for several rounds of training, save and output the TEC inversion deep learning model;

[0096] In the training stage, set the learning rate to 0.01, the optimizer to SGD, train for 100 rounds, and estimate the TEC inversion effect with the mean square error as the loss function, and perform backpropagation to update the network parameters of the TEC inversion deep learning model;

[0097] where Γ k is the TEC value measured by the GPS receiving station and the ground equipment, Γ' k is the tec estimation value of the intelligent inversion model, and k is the batch size input for one forward training.

[0098] Step S3: Perform on-orbit intelligent and rapid inversion of the ionospheric TEC over the globe according to the TEC inversion deep learning model.

[0099] The present invention also provides a system for implementing a method for determining ionospheric TEC based on joint inversion of space-ground active measurement and SAR image, including:

[0100] An FPGA, configured to receive and acquire the scattered echo signal of the dual-frequency SAR satellite, and preprocess the scattered echo signal to obtain its ionospheric scattering amplitude-phase parameters;

[0101] A GPU platform, configured to load and configure the TEC inversion deep learning model, and invert the ionospheric TEC value over the measurement position of the dual-frequency SAR satellite through the TEC inversion deep learning model.

[0102] In the step S3, it specifically includes:

[0103] Step S31: Use the FPGA to acquire the scattered echo signal measured by the dual-frequency SAR satellite in orbit in real time and process it to obtain the ionospheric scattering amplitude-phase parameters of the dual-frequency SAR satellite scattered echo signal, denoted as where C represents the signal amplitude of the dual-frequency SAR satellite scattered echo signal, It represents the signal phase, m represents the frequency, where m = 1, 2. When m = 1, it represents the P band, and when m = 2, it represents the L band; n represents the polarization mode, where n = 1, 2, 3, 4. When n = 1, it is VV polarization, when n = 2, it is VH polarization, when n = 3, it is HV polarization, and when n = 4, it is HH polarization. H represents the number of repeated measurements when taking a set of samples, and q represents the sample number, where q = 1, 2, 3, …, Q.

[0104] Step S32: Perform standardized preprocessing on the ionospheric scattering amplitude-phase parameters of the extracted dual-frequency SAR satellite scattered echo signals and input them into the TEC inversion deep learning model;

[0105] According to the input requirements of the intelligent inversion model, perform standardized preprocessing on the scattering echo amplitude-phase parameters of two frequencies and four polarizations, and use the high-speed data interface between the FPGA and the GPU to transfer the eight standardized scattering echo amplitude-phase parameters to the GPU, and the transmission rate is greater than or equal to 10MB / s.

[0106] Step S33: Perform on-orbit intelligent and rapid inversion on the ionospheric scattering amplitude-phase parameters of the dual-frequency SAR satellite scattered echo signals through the TEC inversion deep learning model to obtain the ionospheric TEC value Γ' over the observation area of the dual-frequency SAR satellite q .

[0107] The ground measurement and control station uploads the trained and optimized ionospheric TEC intelligent inversion model to the on-orbit processing module on the satellite through the space-ground measurement and control link, and the upload rate is greater than or equal to 128kb / s. Under the control of the FPGA, the ionospheric TEC intelligent inversion model is configured into the GPU;

[0108] Using the ionospheric TEC intelligent inversion model configured by the GPU and the eight received scattering echo amplitude-phase parameters perform on-orbit intelligent and real-time inversion on the ionospheric TEC to obtain the ionospheric TEC value Γ' over the observation area of the dual-frequency SAR satellite q .

[0109] It should be noted that in this article, the term "including", "containing" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. Without more limitations, the element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or terminal device including the said element.

[0110] Finally, it should be noted that the above description is the preferred embodiment of the present invention. It should be pointed out that although the preferred embodiments of the present invention have been described, for those skilled in the art of this technology, once the basic creative concept of the present invention is known, without departing from the principle described in the present invention, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. A method for determining ionospheric TEC based on joint inversion of satellite-ground active measurement and SAR image, characterized in that: include: Step S1, establishing a mapping relationship database of ionospheric scattering characteristics measured by dual-frequency SAR satellites, ionospheric penetration characteristics measured by satellite-ground active measurement, and ionospheric TEC measured by GPS receiving stations; Step S2, according to the mapping relationship database, using a multi-source amplitude-phase parameter neural network, establishing and training a TEC inversion deep learning model, wherein the TEC inversion deep learning model takes ionospheric amplitude-phase parameters as input and outputs corresponding TEC values; the ionospheric amplitude-phase parameters are ionospheric scattering signal amplitude-phase parameters and / or ionospheric transmission signal amplitude-phase parameters; Step S3: Perform on-orbit intelligent rapid inversion of the ionospheric TEC over the globe according to the TEC inversion deep learning model.

2. The ionospheric TEC determination method based on satellite-ground active measurement and SAR image joint inversion according to claim 1 is characterized in that: In step S1, the mapping relationship database includes Q sample data, each sample data includes the amplitude and phase parameters of the ionospheric scattering signal measured by the dual-frequency SAR satellite for the simultaneous and spatial ionosphere, the amplitude and phase parameters of the ionospheric transmission signal measured by the satellite-ground active measurement, and the ionospheric TEC value measured by the GPS receiving station; The sample data of the mapping relationship database is represented as in, are the amplitude and phase parameters of the ionospheric scattering signal measured by the dual-frequency SAR satellite, is the amplitude and phase parameters of the ionospheric transmission signal measured by satellite-to-ground active measurement, Γ q is the ionospheric TEC value measured by the GPS receiving station; A and B represent the signal amplitudes of the ionospheric scattering signal and the ionospheric transmission signal, respectively; They respectively represent the signal phase; m represents the signal frequency, m=1 or 2, m=1 represents P band, and m=2 represents L band; n represents the polarization mode of the signal, n=1, 2, 3 or 4, n=1 is VV polarization, n=2 is VH polarization, n=3 is HV polarization, and n=4 is HH polarization; q represents the sample number, q=1,2,3,…,Q; d represents the signal propagation direction of the ionospheric transmission signal, d=1 represents satellite transmission and ground reception, and d=2 represents ground transmission and satellite reception.

3. The ionospheric TEC determination method based on satellite-ground active measurement and SAR image joint inversion according to claim 2 is characterized in that: In the step S1, it specifically includes: Step S11, using the ground transceiver station and the dual-frequency SAR satellite to perform joint measurement, obtain multiple ionospheric transmission signal amplitude and phase parameter samples, and establish an ionospheric transmission signal sub-database for satellite-ground active measurement; the sample data in the ionospheric transmission signal sub-database for satellite-ground active measurement is expressed as Step S12, using dual-frequency SAR satellite measurement, obtain multiple ionospheric scattering signal amplitude and phase parameter samples that are simultaneously and spatially identical to the ionospheric transmission signal measured by the satellite-ground active measurement in step S11, and establish a dual-frequency SAR ionospheric scattering signal sub-database; the sample data in the dual-frequency SAR ionospheric scattering signal sub-database is expressed as Step S13, using the GPS receiving station to obtain multiple ionospheric TEC values ​​at the same time and space as the ionospheric transmission signal measured by the satellite-ground active measurement in step S11, and establish a TEC value sub-database; the samples in the TEC value sub-database are recorded as Γ q , the measurement accuracy of samples in the TEC value sub-database reaches 0.5TECU; Step S14: establishing the mapping relationship database according to the ionospheric transmission signal sub-database of the satellite-ground active measurement, the dual-frequency SAR ionospheric scattering signal sub-database and the TEC value sub-database.

4. The method for determining ionospheric TEC based on satellite-ground active measurement and SAR image joint inversion according to claim 3, characterized in that: The duration of satellite-to-ground active measurement using a ground transceiver station and a dual-frequency SAR satellite is τ1, τ1≤20ms; The duration of the measurement time slot of the satellite-to-ground active measurement is τ2, τ2≤200ms; The duration of the ionospheric scattering signal measured by the dual-frequency SAR satellite is τ3, τ3≤150ms.

5. The method for determining ionospheric TEC based on satellite-ground active measurement and SAR image joint inversion according to claim 3, characterized in that: In the step S2, it specifically includes: Step S21, constructing and utilizing the multi-source amplitude and phase parameter neural network to establish a TEC inversion deep learning model, wherein the TEC inversion deep learning model includes a feature extraction network, a feature fusion network and a TEC regression network; Step S22, according to the mapping relationship database, obtain the amplitude and phase parameters s1 and s3 of the ionospheric scattering signal of satellite-to-ground and ground-to-satellite in the sample data, and input them into the feature extraction network to obtain the extracted feature graphs s'1=Net(s1), s'3=Net(s3); Step S23, according to the mapping relationship database, the amplitude and phase parameters of the ionospheric scattering signal spontaneously transmitted and received by the dual-frequency SAR satellite in the sample data are obtained and input into the feature extraction network to obtain the extracted feature graph s'2=Net(s2); Step S24, simultaneously inputting the feature graphs s'1, s'2 and s'3 extracted by the feature extraction network into the feature fusion network for feature splicing and deep feature fusion; Step S25, performing TEC estimation on the fused deep features through the TEC regression network to obtain a TEC estimation value; Step S26, based on the obtained TEC estimation value and the ionospheric TEC value measured by the GPS receiving station in the sample data, the inversion result is estimated and back-propagated to update the network parameters; Step S27, repeating the above steps S22 to S26, training for several rounds, saving and outputting the TEC inversion deep learning model; In the training phase, the learning rate is set to 0.01, the optimizer is SGD, and the training is repeated for 100 rounds. estimating the TEC inversion effect for the loss function, and back-propagating and updating the network parameters of the TEC inversion deep learning model; Among them, Γ k is the TEC value measured by the GPS receiving station and ground equipment, Γ' k is the TEC estimate of the intelligent inversion model, k is a variable representing the batch input for a single forward training, and K represents the batch value size input for a single forward training.

6. The method for determining ionospheric TEC based on satellite-ground active measurement and SAR image joint inversion according to claim 5, characterized in that: The feature extraction network includes a convolutional layer conv1, a first batch of normalization layers, a first activation function, a convolutional layer conv2, a second batch of normalization layers, and a second activation function connected in series in sequence; The convolution kernel size of the convolution layer conv1 is (1,1), the step size is 1, the padding is 0, the number of input channels is 8, and the number of output channels is set to 64; The first activation function and the second activation function are linear rectification functions relu=max(0,x1), wherein max() represents the maximum value between 0 and x1, and x1 is an intermediate value generated in the forward process of the neural network; The convolution kernel size of the convolution layer conv2 is (1,1), the step size is 1, the padding is 0, the number of input channels is 64, and the number of output channels is set to 128; The calculation method of the first batch of normalized layers and the second batch of normalized layers is x1'=γ*x norm +βwhere x norm is the feature after regularization, γ and β are two learnable feature linear transformation parameters, Among them, ∈ is an adjustment factor to prevent the denominator from being zero, and its value is set to 0.0001; is the mean, k is the batch size input for a single forward training, is the variance, x k Represents the intermediate value generated during the forward process of the kth batch of neural networks in a single forward training; In the step S24, the feature splicing method is channel connection, expressed as s"=concat(s'1,s'2,s'3), where concat is a channel connection function; the deep feature fusion is performed through a deep fusion convolution layer, expressed as s"'=conv(s"), where conv is a deep fusion convolution layer, the convolution kernel size is (1,1), the step size is 1, the padding is 0, the number of input channels is 128, and the number of output channels is set to 64; The number of input channels of the deep fusion convolution layer is consistent with the number of channels of the feature map s" obtained after the feature splicing, where s'1 and s'3 are the feature maps of the amplitude and phase parameters of the ionospheric scattering signals of satellite-to-ground and ground-to-satellite transmission and reception, respectively. Before feature fusion, there is a z% probability that they are replaced by Gaussian noise to adapt to the input of only spontaneous and self-received signal features in the on-orbit prediction stage, where Gaussian noise is a normal distribution with a mean of 0 and a variance of 1.

7. The method for determining ionospheric TEC based on satellite-ground active measurement and SAR image joint inversion according to claim 6, characterized in that: In step S3, in the on-orbit intelligent rapid inversion stage, the input of the feature fusion network is the feature map extracted from the ionospheric scattering signal measured by the dual-frequency SAR satellite after passing through the feature extraction network and the randomly generated Gaussian signal, and the feature map extracted after the feature extraction network and the randomly generated Gaussian signal are feature fused and then deep feature fusion is performed through the deep fusion convolution layer, and the number of input channels of the deep fusion convolution layer is consistent with the number of channels of the feature map extracted after passing through the feature extraction network.

8. The method for determining ionospheric TEC based on satellite-ground active measurement and SAR image joint inversion according to claim 7, characterized in that: In the step S3, it specifically includes: Step S31, obtaining and processing the dual-frequency SAR satellite scattered echo signal to obtain the ionospheric scattering amplitude and phase parameters of the dual-frequency SAR satellite scattered echo signal; Step S32, performing standardized preprocessing on the ionospheric scattering amplitude and phase parameters of the dual-frequency SAR satellite scattered echo signal extracted and inputting them into the TEC inversion deep learning model; Step S33: Perform on-orbit intelligent rapid inversion of the ionospheric scattering amplitude and phase parameters of the dual-frequency SAR satellite scattered echo signal through the TEC inversion deep learning model to obtain the ionospheric TEC value Γ' above the dual-frequency SAR satellite observation area. q .

9. A system for implementing the ionospheric TEC determination method based on satellite-ground active measurement and SAR image joint inversion as claimed in any one of claims 1 to 8, characterized in that: include: FPGA, used for receiving and acquiring scattered echo signals of dual-frequency SAR satellites, and preprocessing the scattered echo signals to obtain ionospheric scattering amplitude and phase parameters thereof; The GPU platform is used to load and configure the TEC inversion deep learning model, and to invert the TEC value of the ionosphere above the measurement position of the dual-frequency SAR satellite through the TEC inversion deep learning model.

Images