A Method and System for Measuring the D Layer of the Ionosphere in the Very Low Frequency Region Based on Particle Filtering
By constructing a synchronous observation network and a particle filtering algorithm, the problem of difficult detection of electron density in the D layer of the ionosphere was solved, achieving efficient and low-cost near real-time measurement and providing a high-resolution electron density distribution map of the D layer of the ionosphere.
Patent Information
- Application Number
- CN202311384673.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-10-23
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-10-23
AI Technical Summary
Existing technologies are insufficient for effectively detecting the electron density of the D layer of the ionosphere. Traditional methods are costly and have limited spatial coverage. Traditional VLF remote sensing technology can only estimate the average electron density of a single path, and its spatial coverage is limited.
Based on the particle filtering method, a synchronous observation network is constructed by deploying multiple ground-based very low frequency (VLF) wave receivers. Combining the localization algorithm and the Ferguson ionospheric model, the VLF wave propagation model and particle filtering algorithm are used for data assimilation to achieve near real-time estimation of the electron density of the D layer of the ionosphere.
It enables near real-time, high spatial resolution measurement of electron density in the D layer of the ionosphere over a large area, providing stable and reliable data support, reducing costs and expanding spatial coverage.
Smart Images

Figure CN117452508B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of space exploration, and more specifically, relates to a method and system for measuring the D layer of the ionosphere in the very low frequency region based on particle filtering. Background Technology
[0002] The D layer of the ionosphere is a partially ionized part of the atmosphere at an altitude of 60-100 kilometers above the ground, with an electron density typically ranging from 1 to 1000 cm³. -3 This region of space is affected not only by top-down solar activity but also by bottom-up atmospheric fluctuations. The electron density of the D layer of the ionosphere is crucial for radio communication, radio wave propagation, maritime navigation and positioning, and space environment monitoring. However, this area falls within the blind zone of satellite and balloon exploration; it is too low for satellite instruments and too high for weather balloons. The relatively low electron density of the D layer makes it an ineffective reflector for conventional ionospheric probes, making it a major challenge for space exploration. Currently, the most effective detection method is direct measurement using sounding rockets, but this method has limited time and spatial coverage and is very costly.
[0003] Very Low Frequency (VLF) waves refer to electromagnetic waves with frequencies ranging from 3 to 30 kHz, primarily generated by thunderstorms or by transmitters from man-made stations distributed globally. VLF waves are typically well confined within Earth-ionospheric waveguides, consisting of the ground and the D layer of the ionosphere, enabling long-distance propagation with a typical attenuation coefficient of ~2-3 dB / Mm. VLF waves are of great significance for research in fields such as space science, underwater communication, and maritime navigation and positioning.
[0004] VLF remote sensing technology has long been the primary means of detecting the D layer of the ionosphere. Traditional methods involve observing the amplitude or phase of the signal at distances of hundreds to thousands of kilometers from the source, and then using a very low frequency (VLF) propagation model to inversely estimate the electron density required to generate the observed data (amplitude and phase). This method can only estimate the average electron density along a single propagation path and has a rather limited spatial coverage. Therefore, solving the problem of detecting the D layer of the ionosphere is a current and urgent challenge for researchers. Summary of the Invention
[0005] This invention addresses the technical problems existing in the prior art by providing a method and system for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering. This invention addresses the D-layer measurement problem from the perspective of data assimilation, combining VLF signal observation data and prior estimates of the D-layer of the ionosphere obtained based on empirical models. By utilizing a very low frequency wave propagation model, it achieves near real-time estimation of the D-layer electron density within the network coverage area of the very low frequency wave.
[0006] According to a first aspect of the present invention, a method for measuring the D layer of the ionosphere in the very low frequency region based on particle filtering is provided, comprising the following steps: Step 1: Use a synchronous observation network constructed by multiple ground-based very low frequency (VLF) wave receivers deployed in different locations to receive electromagnetic wave signals from VLF transmitting stations in real time and obtain their amplitude and phase. Step 2: Divide the spatial region to be measured into grids at fixed intervals, use a localization algorithm to determine the weight of each transmitter-receiver path to each grid, and generate a localized grid. Step 3: Use the Ferguson ionospheric model to calculate the initial state parameters of the D layer of the ionosphere at each grid, and sample from a multivariate Gaussian distribution to generate the set of initial states of the D layer of the ionosphere at each grid. Step 4: Under the initial state set of the D layer of the ionosphere, the amplitude and phase of the very low frequency signal of the transmitter at each receiver location are simulated and calculated using the very low frequency wave propagation model based on the transmitter-receiver path information and the D layer parameter set of the ionosphere. Step 5: Based on the error vector of amplitude and phase between the very low frequency observation signal and the simulated signal, use the local particle filter algorithm to optimize and update the D layer parameter set of the ionosphere. Repeat steps 4 and 5 for multiple iterations and record the result of the last iteration. Step 6: Using a two-dimensional interpolation algorithm, spatial interpolation is performed between each grid to obtain high spatial resolution D-layer ionospheric parameter measurement results, and the D-layer electron density distribution of the ionospheric layer is calculated to obtain a two-dimensional D-layer electron density distribution map.
[0007] Based on the above technical solution, the present invention can be further improved as follows: Optionally, the measurement range of the D layer of the ionosphere is determined by a number of propagation paths formed by the very low frequency detection station and the transmitting station, with each propagation path covering a spatial area within a certain range.
[0008] Optionally, the step of using a localization algorithm to determine the weight of each transmitter-receiver path to each grid and generating a localized grid includes: First, calculate the distance from each grid cell to each transmitter-receiver path. If the distance is greater than a set value... If the transmitter-receiver path has a localized weight of 0 for the grid, then the localized weight of the grid is 1 otherwise; wherein the localized grid includes grid coordinates and the weights of all transmitter-receiver paths for the grid.
[0009] Optionally, in step 3, the step of calculating the initial state parameters of the ionospheric D layer at each grid using the Ferguson ionospheric model and sampling to generate the initial state set of the ionospheric D layer at each grid in a multivariate Gaussian distribution includes the following steps: Step 3.1: Use the Ferguson ionospheric model to calculate the h′ and β parameters of the ionospheric D layer at each grid, and use them as the mean of the initial set of the ionospheric D layer h′ and β parameters; Step 3.2: Calculate the spatial covariance matrix Cov between grids using the Caspari-type Gaussian function C0(z, c); Step 3.3, with h ′ and β are the mean values, and Cov is the covariance matrix. The initial state set of the ionosphere is generated by sampling in a multivariate Gaussian distribution.
[0010] Optionally, in step 3.1, the Ferguson ionosphere model formula is expressed as:
[0011]
[0012] In the formula, h F For the artificial height of the ionosphere, β For ionospheric sharpness, Geographical latitude, The zenith angle of the sun. For months.
[0013] Optionally, in step 3.2, the Caspari-type Gaussian function C0(z, c) is expressed as:
[0014] In the formula, and Defined as:
[0015]
[0016] The covariance between mesh A and mesh B is calculated using a Gaussian-like function C0(z, c):
[0017] In the formula, The geographical distance between grids A and B. For length scale, The standard deviation is denoted as .
[0018] Optionally, in step 4, the transmitter-receiver propagation path information includes the very low frequency transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments along the propagation path at fixed distances.
[0019] Optionally, in step 5, the phase error vector Δ is represented as:
[0020] Where Δamp,pathi is the difference between the observed amplitude and the simulated amplitude of the VLF signal on the i-th propagation path, and Δphase,pathi is the difference between the phase of the observed VLF signal and the phase of the simulated VLF signal on the i-th propagation path.
[0021] Optionally, the electron density of the D layer of the ionosphere is expressed as:
[0022] in, N e electron density ,h' β represents the virtual height of the ionosphere, z represents the sharpness of the ionosphere, and z represents the height.
[0023] According to a second aspect of the present invention, a particle-filter-based system for measuring the D layer of the ionosphere in the very low frequency region is provided, comprising a very low frequency transmitter and a network receiver; further comprising: The forward simulation module is used to simulate the propagation of very low frequency (VLF) signals along the predetermined transmitter-receiver propagation paths under the initial state of the ionospheric D layer. It obtains the amplitude and phase of the simulated VLF signals at each receiving station. The initial state of the ionospheric D layer includes the virtual height of the ionospheric D layer at various locations within the detection area. h' Along with sharpness β, the transmitter-receiver propagation path information includes the very low frequency transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments at fixed distances along the propagation path, which are simulated using a very low frequency wave propagation model; The observation and receiving module is used by multiple very low frequency (VLF) receivers to receive VLF signals from each transmitter and demodulate them to obtain the amplitude and phase of the VLF signals from each transmitter. The localization module is used to divide the spatial grid and determine the spatial grid within the coverage area of each propagation path; The data assimilation module is used by the particle filter algorithm to update the ionospheric state at the spatial grid within its coverage area based on the error vector Δ on each propagation path; The interpolation module is used for two-dimensional interpolation between ionospheric states at the spatial grid to obtain high spatial resolution ionospheric D-layer detection results.
[0024] Compared with the prior art, the technical effects and advantages of the present invention are as follows: This invention proposes a method and system for measuring the D layer of the ionosphere in very low frequency (VLF) regions based on particle filtering. By constructing a synchronous observation network using multiple ground-based VLF wave receivers (detection stations) deployed at different locations, and using a particle filtering algorithm, continuous remote sensing measurements of the D layer of the ionosphere over a large area are achieved. This enables quasi-real-time estimation of the electron density of the D layer within the VLF wave network coverage area. It can provide stable, reliable, and high spatial resolution quasi-real-time data for the study and related applications of the D layer of the ionosphere. Attached Figure Description
[0025] Figure 1 A block diagram of a particle-filter-based method for measuring the D-layer of the ionosphere in the very low frequency region, provided in an embodiment of the present invention; Figure 2 Spatial grid distribution diagram provided for embodiments of the present invention; Figure 3 This is a schematic diagram of a localized single propagation path provided in an embodiment of the present invention; Figure 4 A schematic diagram of daytime ionospheric prior particles provided in an embodiment of the present invention; Figure 5 The measured daytime “real” ionospheric state, a priori ionospheric state, and ionospheric state estimation results are provided for embodiments of the present invention. Figure 6 The electron density Ne error distribution diagram provided for an embodiment of the present invention. Detailed Implementation
[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0027] Understandably, given the deficiencies in the background technology, this invention proposes a particle filtering-based method for measuring the D-layer of the ionosphere in the very low frequency region. The method includes the following steps: Step 1: Using a synchronous observation network constructed from multiple ground-based very low frequency (VLF) wave receivers deployed in different locations, electromagnetic wave signals from VLF transmitting stations are received in real time, and their amplitude and phase are obtained, denoted as... ; It should be noted that the measurement range of the D layer of the ionosphere is determined by the numerous propagation paths formed by the very low frequency detection station and the transmitting station, with each propagation path covering a certain spatial area within a certain range.
[0028] Step 2: Divide the spatial region to be measured into grids at fixed intervals, use a localization algorithm to determine the weight of each transmitter-receiver path to each grid, and generate a localized grid. Specifically, step 2 includes the following steps: Step 2.1: Convert the latitude and longitude coordinates (lat, lon) in the WGS84 coordinate system to the equidistant conical coordinate system ESRI:54027 coordinates (x, y), and divide the grid according to a fixed geographical distance Δd. The center coordinates of the (i, j)th grid are (x0+jxΔd, y0+ixΔd). Step 2.2: Use a localization algorithm to determine the weight of each transmitter-receiver path to each grid cell: First, calculate the distance from each grid cell to each transmitter-receiver path. If the distance is greater than a set value... If the transmitter-receiver path has a localization weight of 0 for the grid, then the localization weight of the grid is 1.
[0029] It should be noted that the localization algorithm defines a region around each transmitter-receiver path. Within this region, the ionospheric state update of the grid will use VLF data observed by the receiver. The propagation characteristics and attenuation coefficient of the VLF wave are closely related to the ionospheric state along the propagation path. The closer the spatial region is to the propagation path, the stronger the correlation with the ionospheric state characterized by the VLF observation data. An excessively large local region will result in the use of some irrelevant observation data for updates, while an excessively small local region will result in the inability to fully utilize relevant VLF observation data. The localization algorithm determines which VLF observation data the grid uses during the state update process.
[0030] Step 3: Use the Ferguson ionospheric model to calculate the initial state parameters of the D layer of the ionosphere at each grid point, denoted as... h ′ and β; and with h ′ and β are the mean values, and the initial state set of the ionospheric D layer at each grid is generated by sampling from a multivariate Gaussian distribution. ; Specifically, step 3 includes the following steps: Step 3.1: Calculate the D layer of the ionosphere at each grid using the Ferguson ionospheric model. h The parameters ′ and β are denoted as h ′ and β As the D layer of the ionosphere h The mean of the initial set of ′ and β parameters; Step 3.2: Calculate the spatial covariance matrix between grids using the Caspari-type Gaussian function C0(z, c). Cov ; The Caspari-class Gaussian function C0(z, c) is expressed as:
[0031] in, and Defined as:
[0032]
[0033] The covariance between mesh A and mesh B is calculated using a Gaussian-like function C0(z, c):
[0034] in, The geographical distance between grids A and B. For length scale, The standard deviation is denoted as .
[0035] Step 3.3, with ′ and β As the mean, with Cov The covariance matrix is used to sample and generate the initial state set of the ionosphere from a multivariate Gaussian distribution.
[0036] It should be noted that in step 3.1, the Ferguson ionospheric model describes the best estimate of the ionospheric state when there is no observation data.
[0037] The Ferguson ionosphere model provides prior particles for ionospheric state estimation. This model describes the best estimate of the ionospheric state when no observational data is available. The Ferguson ionospheric model will... h' Let β be a function of latitude, solar zenith angle, and seasonal parameters. Using a simplified form of the Ferguson ionospheric model as the average value of prior particles, the formula is as follows:
[0038]
[0039] in, h F For the artificial height of the ionosphere, β For ionospheric sharpness, Geographical latitude, The zenith angle of the sun. For months.
[0040] Step 4: Under the initial state set of the ionospheric D layer, use the very low frequency wave propagation model based on the transmitter-receiver path information and the ionospheric D layer... The sets of ′ and β parameters were used to simulate and calculate the amplitude and phase of the very low frequency signal from the transmitter at each receiver location. ; It should be noted that the very low frequency wave propagation model (LWPC, Long-Wavelength Propagation Capability) is based on the input transmitter-receiver propagation path information and the ionospheric virtual height. ′ and ionospheric sharpness β The amplitude and phase of the simulated very low frequency signal at the receiving station location were calculated. This involved dividing the transmitter-receiver path into several uniform waveguide segments, and then passing through each waveguide segment where defined... h' Using the β value, calculate the VLF amplitude and phase at different distances along the propagation path.
[0041] The transmitter-receiver propagation path information includes the very low frequency transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments along the propagation path at fixed distances.
[0042] Step 5: Based on the very low frequency observation signal With analog signals The amplitude and phase error vector Δ is used to optimize and update the D layer of the ionosphere using a local particle filter algorithm. h The set of parameters ′ and β is obtained. Repeat steps 4 and 5 for multiple iterations, where k is the number of iterations, and the result of the last iteration is denoted as k. ; It should be noted that the Local Particle Filter with Gaussian Mixture Extension (LPFGM) algorithm is an improved version of Local Particle Filter (LPF). Specifically, the LPFGM algorithm assimilates the observed data with the Ferguson ionospheric model and updates the electron density distribution across the entire detection area. This method not only provides nonlinear inversion capabilities but also boasts high computational speed and accuracy. The LPFGM particle filter algorithm employs a two-step update strategy: first, it performs ensemble Kalman filtering to reduce the particle ensemble variance; then, it calculates the particle weights and resamples the particle ensemble. This adds a Kalman-based particle update step before the resampling step in the traditional LPF algorithm, making the overall state of the particle swarm closer to the observed state and avoiding weight collapse during resampling.
[0043] The Ferguson ionosphere model provides prior particles for ionospheric state estimation. The prior particles are generated by sampling from a multivariate Gaussian distribution, with the average value representing the ionospheric state in the Ferguson model. (The sentence about multivariate Gaussian distribution sampling is incomplete and lacks context.) h' The standard deviation of β is 2 km, and the standard deviation of β is 0.04 km. -1 .
[0044] The calculation steps of the local particle filter algorithm are as follows: (1) For the prior particle X b The model observations were calculated using a long-wavelength propagation model.
[0045] (2) Calculate the observed disturbance
[0046] (3) Calculate state disturbance
[0047] For the i-th grid: (4) Select observation data with non-zero localization weights for grid i. (6) Calculate the transition matrix C
[0048] (7) Calculate the state error covariance matrix
[0049] (8) Calculate the particle transformation matrix of Gaussian mixing.
[0050]
[0051] (9) Calculate particle weights and effective particle counts.
[0052]
[0053]
[0054] (10) The transformation matrix is generated by an adaptive resampling algorithm based on the Monte Carlo method. In this embodiment, the effective particle number threshold N0=50.
[0055] (11) Particle state update .
[0056] The above is based on actual observed very low frequency signals With analog signals The error vector Δ between amplitude and phase is expressed as:
[0057] Where Δamp,pathi is the difference between the observed amplitude and the simulated amplitude of the VLF signal on the i-th propagation path, and Δphase,pathi is the difference between the phase of the observed VLF signal and the phase of the simulated VLF signal on the i-th propagation path.
[0058] Step 6: Use a two-dimensional interpolation algorithm to perform interpolation between each grid cell. Spatial interpolation was performed to obtain high spatial resolution measurements of the h′ and β parameters of the ionospheric D layer. The electron density distribution of the ionospheric D layer was calculated to obtain a two-dimensional electron density distribution map of the ionospheric D layer.
[0059] This step specifically includes: using an interpolation imaging algorithm to process the ionospheric state X estimated by the particle filter algorithm. a Spatial interpolation is performed to construct the ionospheric state across the entire geographic region, generating a high spatial resolution map of the electron density distribution in the D layer of the ionosphere.
[0060] It should be noted that the electron density of the D layer of the ionosphere is expressed as a two-parameter exponential function derived by Wait, where the electron density Ne (cm²) is... -3 ) is a function of altitude z (km), given by the following formula:
[0061] in, h' β is the virtual height of the ionosphere (km), and β is the sharpness of the ionosphere (km). -1 By estimating at each spatial grid... h' By using the β value, we can obtain the distribution of electron density at different heights within the grid. (All grid locations) h' The total vector x of β is the state that the filter wants to estimate.
[0062] The present invention will now be described in further detail with reference to specific examples and accompanying drawings. These descriptions are intended to explain the invention and not to limit it.
[0063] The block diagram of the embodiment of the present invention is attached. Figure 1As shown, it mainly includes: the International Geomagnetic Reference Field (IGRF) 13 model. The IGRF 13 model can provide background geomagnetic field information on each transmitter-receiver path of the LWPC model based on the date, time, and geographical coordinates of the waveguide segment of the VLF propagation path.
[0064] The 13th generation of the International Geomagnetic Reference Field, released in 2019, accurately describes the geomagnetic field information required for the LWPC model. In the source-free region above the Earth's surface, the dominant magnetic field originating from the Earth's interior is the negative gradient B of the scalar potential V. V. Function V(r,θ, ,t) represents the spherical harmonic coefficient. , Finite series expansion:
[0065] Where r is the radial distance from the Earth's center, θ, These represent latitude and longitude, respectively, with reference to the Earth's radius a = 6371.2 km. These are the nth and mth order Schmitt semi-normalized Legendre functions.
[0066] Figure 1 In the middle, the ground conductivity parameter , The VLF ground conductivity distribution map can be obtained from the VLF ground conductivity distribution map. The VLF ground conductivity distribution map can provide the LWPC model with ground conductivity parameters for each transmitter-receiver path based on the date, time, and geographical coordinates of the waveguide segment of the VLF propagation path. , .
[0067] This invention uses a simulation experiment to demonstrate the measurement results of the D layer of the ionosphere under daytime conditions. The simulation experiment uses six receivers located in central and northern China to construct a very low frequency (VLF) synchronous observation network. It measures signals from three transmitters in southern China at frequencies of 18.2, 19.8, and 22.2 kHz, with a total of 18 transmitter-receiver paths, to obtain the actual VLF observation signals. .
[0068] To ensure the rigor of the simulation experiments, we used a Longwave Mode Propagator (LMP), which differs from the longwave propagation model, to generate VLF signal observation data. (See attached image.) Figure 5As shown in Day(a) and Day(d), the “real” ionosphere measured in this embodiment is generated by superimposing Fourier perturbations on the Ferguson ionosphere model:
[0069]
[0070]
[0071]
[0072]
[0073]
[0074] First, the geographical range for ionospheric measurements is defined as 73°E–135°E, 16°N–54°N. A spatial grid is then created in the ESRI 54027 coordinate system with a step size of 300 km. The grid distribution is shown in the attached figure. Figure 2 As shown.
[0075] Secondly, a localization algorithm is used to determine all grids within a 600 km radius of the propagation path that use VLF observation data from that path during state updates. This observation data is then used in the local particle filter algorithm to estimate the ionospheric state within each grid. The localization region for a single path is shown in the attached figure. Figure 3 As shown, in this embodiment, we determine that during state updates, all grids within a 600 km range of the propagation path use the VLF observation data of that path.
[0076] Then, using the ionospheric state defined by the Ferguson ionospheric model as the mean, we take... h' The standard deviation of β is 2 km, and the standard deviation of β is 0.04 km. -1 Based on the fifth-order piecewise rational function proposed by Caspari, a spatial covariance matrix is generated. One hundred random samples are performed within a multivariate Gaussian distribution to serve as the initial set of ionospheric states. In this embodiment, the particle set size is 100. Figure 4 The ionospheric states of three random particles are shown, where (a), (b), and (c) are... (d), (e), and (f) are In this embodiment, the length scale =600 km, the covariance between grids more than 1200 km apart is 0. This covariance matrix has a Gaussian-like spatial correlation, giving the prior ionosphere a reasonable spatial structure. Its ionospheric state mean is attached. Figure 5 As shown in Day(b) and Day(e).
[0077] A long-wavelength propagation model is used to simulate the propagation of VLF signals beneath the prior ionosphere. The background geomagnetic field and ground conductivity parameters required for simulating VLF signal propagation are provided by the IGRF 13 model and the VLF ground conductivity distribution map, respectively. The long-wavelength propagation model uses the background geomagnetic field and ground conductivity provided by the IGRF 13 model and the VLF ground conductivity distribution map as inputs, assuming the Earth is spherical. In this example, the long-wavelength propagation model must simulate all combinations of particle and transmitter-receiver paths, i.e., run 100 × 18 = 1800 times, where 18 is the number of transmitter-receiver paths and 100 is the particle set size. To reduce the complexity of running the long-wavelength propagation model, in this embodiment, the propagation path is divided into uniform waveguide segments of 50 km length, and a first-order polynomial interpolator is used to interpolate on the spatial grid to obtain the values at each uniform waveguide segment. h' and β value.
[0078] Finally, the local particle filter algorithm is used to assimilate the VLF observation data and the prior ionospheric state, updating the ionospheric state across the entire detection area. The local particle filter algorithm can iterate using multiple observations; this embodiment shows the ionospheric estimation results after three iterations, as shown in the appendix. Figure 5 As shown in Day(c) and Day(f).
[0079] The following analysis examines the measurement error of the electron density in the D layer of the ionosphere in this embodiment. Based on the "real" ionosphere in the simulation experiment and the ionosphere measured in this embodiment, the electron density Ne at an altitude of 70 km is calculated using the two-parameter exponential function derived by Wait. The error is shown in the appendix. Figure 6 As shown. The maximum error is 2.8 cm. -3 The average error is 0.22 cm. -3 .
[0080] Therefore, the method provided by the embodiments of the present invention can effectively obtain the electron density of the D layer of the ionosphere over a large area, with an electron density error of less than 3 cm. -3 This technology is expected to provide stable, reliable, and high-resolution near real-time data for ionospheric research and related engineering applications.
[0081] Finally, this embodiment of the invention also provides a very low frequency (VLF) ionospheric D-layer measurement system, including a VLF transmitter and a network receiver, for use in the aforementioned VLF ionospheric D-layer measurement method based on particle filtering. The system further includes: The forward simulation module is used to simulate the propagation of very low frequency (VLF) signals along the predetermined transmitter-receiver propagation paths under the initial state of the ionospheric D layer. It obtains the amplitude and phase of the simulated VLF signals at each receiving station. The initial state of the ionospheric D layer includes the virtual height of the ionospheric D layer at various locations within the detection area. h' Along with sharpness β, the transmitter-receiver propagation path information includes the very low frequency transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments at fixed distances along the propagation path, which are simulated using a very low frequency wave propagation model; The observation and receiving module is used by multiple very low frequency (VLF) receivers to receive VLF signals from each transmitter and demodulate them to obtain the amplitude and phase of the VLF signals from each transmitter. The localization module is used to divide the spatial grid and determine the spatial grid within the coverage area of each propagation path; The data assimilation module is used by the particle filter algorithm to update the ionospheric state at the spatial grid within its coverage area based on the error vector on each propagation path; The interpolation module is used for two-dimensional interpolation between ionospheric states at the spatial grid to obtain high spatial resolution ionospheric D-layer detection results.
[0082] It is understood that the particle-filter-based very low frequency region ionospheric D-layer measurement system described in the embodiments of the present invention and the particle-filter-based very low frequency region ionospheric D-layer measurement method described above can be referred to in correspondence with each other.
[0083] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering, characterized in that, Includes the following steps: Step 1: Use a synchronous observation network constructed by multiple ground-based very low frequency (VLF) wave receivers deployed in different locations to receive electromagnetic wave signals from VLF transmitting stations in real time and obtain their amplitude and phase. Step 2: Divide the spatial region to be measured into grids at fixed intervals, use a localization algorithm to determine the weight of each transmitter-receiver path to each grid, and generate a localized grid. Step 3: Use the Ferguson ionospheric model to calculate the initial state parameters of the D layer of the ionosphere at each grid, and sample from a multivariate Gaussian distribution to generate the set of initial states of the D layer of the ionosphere at each grid. Step 4: Under the initial state set of the D layer of the ionosphere, the amplitude and phase of the very low frequency signal of the transmitter at each receiver location are simulated and calculated using the very low frequency wave propagation model based on the transmitter-receiver path information and the D layer parameter set of the ionosphere. Step 5: Based on the error vector of amplitude and phase between the very low frequency observation signal and the simulated signal, use the local particle filter algorithm to optimize and update the D layer parameter set of the ionosphere. Repeat steps 4 and 5 for multiple iterations and record the result of the last iteration. Step 6: Using a two-dimensional interpolation algorithm, spatial interpolation is performed between each grid to obtain high spatial resolution D-layer ionospheric parameter measurement results, and the D-layer electron density distribution of the ionospheric layer is calculated to obtain a two-dimensional D-layer electron density distribution map.
2. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, The measurement range of the D layer of the ionosphere is determined by numerous transmitter-receiver paths formed by very low frequency detection stations and transmitting stations, with each transmitter-receiver path covering a spatial area within a certain range.
3. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, The process of using a localization algorithm to determine the weight of each transmitter-receiver path to each grid and generating a local grid includes: Calculate the distance from each grid cell to each transmitter-receiver path. If the distance is greater than a set value... If the transmitter-receiver path has a localized weight of 0 for the grid, then the localized weight of the grid is 1 otherwise; wherein the localized grid includes grid coordinates and the weights of all transmitter-receiver paths for the grid.
4. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, Step 3, which involves calculating the initial state parameters of the D layer of the ionosphere at each grid using the Ferguson ionosphere model and sampling to generate the initial state set of the D layer of the ionosphere at each grid using a multivariate Gaussian distribution, includes the following steps: Step 3.1: Use the Ferguson ionospheric model to calculate the h′ and β parameters of the ionospheric D layer at each grid, and use them as the mean of the initial set of the ionospheric D layer h′ and β parameters; Step 3.2: Calculate the spatial covariance matrix Cov between grids using the Caspari-type Gaussian function C0(z, c); Step 3.3, with h ′ and β are the mean values, and Cov is the covariance matrix. The initial state set of the ionosphere is generated by sampling in a multivariate Gaussian distribution.
5. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 4, characterized in that, In step 3.1, the Ferguson ionosphere model formula is expressed as: in, h F For the artificial height of the ionosphere, β For ionospheric sharpness, Geographical latitude, The zenith angle of the sun. For months.
6. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 4, characterized in that, In step 3.2, the Caspari-type Gaussian function C0(z, c) is expressed as: In the formula, and Defined as: The covariance between mesh A and mesh B is calculated using a Gaussian-like function C0(z, c): In the formula, The geographical distance between grids A and B. For length scale, The standard deviation is denoted as .
7. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, In step 5, the phase error vector Δ is represented as: Where Δamp,pathi is the difference between the observed amplitude and the simulated amplitude of the VLF signal on the i-th propagation path, and Δphase,pathi is the difference between the phase of the observed VLF signal and the phase of the simulated VLF signal on the i-th propagation path.
8. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, In step 4, the transmitter-receiver propagation path information includes the very low frequency transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments along the propagation path at fixed distances.
9. The method for measuring the D-layer of the ionosphere in the very low frequency region based on particle filtering according to claim 1, characterized in that, In step 6, the calculated electron density distribution of the D layer of the ionosphere is expressed as follows: in, N e electron density ,h' β represents the virtual height of the ionosphere, z represents the sharpness of the ionosphere, and z represents the height.
10. A particle-filter-based very low frequency (VLF) ionospheric D-layer measurement system, used to execute the particle-filter-based VLF ionospheric D-layer measurement method of claim 1, comprising a VLF transmitter and a network receiver; characterized in that: Also includes: The forward simulation module is used to simulate the propagation of very low frequency (VLF) signals along the set transmitter-receiver propagation paths under the initial state of the D layer of the ionosphere. It obtains the amplitude and phase of the simulated VLF signals at each receiving station. The initial state of the D layer of the ionosphere includes the virtual height h' and sharpness β of the D layer of the ionosphere at each location in the detection area. The transmitter-receiver propagation path information includes the VLF transmitter coordinates, transmitter frequency, receiver coordinates, and coordinates of segments at fixed distances along the propagation path. The simulation is performed using a VLF wave propagation model. The observation and receiving module is used by multiple very low frequency (VLF) receivers to receive VLF signals from each transmitter and demodulate them to obtain the amplitude and phase of the VLF signals from each transmitter. The localization module is used to divide the spatial grid and determine the spatial grid within the coverage area of each propagation path; The data assimilation module is used by the particle filter algorithm to update the ionospheric state at the spatial grid within its coverage area based on the error vector on each propagation path; The interpolation module is used for two-dimensional interpolation between ionospheric states at the spatial grid to obtain high spatial resolution ionospheric D-layer detection results.
Citation Information
Patent Citations
Ionized layer chromatography technology and ionized layer delay correction method based on multi-scale subdivision
CN104007479A
Inversion method of ionosphere height
CN108549114A
