Method, device and system for predicting meteorological attenuation of a satellite-ground link environment

Abstract

The application relates to the technical field of satellite communication, and discloses a satellite-ground link environment meteorological attenuation prediction method, device and system. The technical scheme comprises the following steps: acquiring current environment data, meteorological prediction results, satellite-ground link parameters and statistical physical attenuation values of a region to be measured; splicing the current environment data and each meteorological prediction result to obtain environment meteorological characteristics; respectively extracting link gain deviation characteristics, statistical physical attenuation characteristics, environment meteorological gain deviation characteristics and environment meteorological basic deviation characteristics according to corresponding parameters; determining statistical physical deviation characteristics according to the product of the three of the link gain deviation characteristics, the environment meteorological gain deviation characteristics and the statistical physical attenuation characteristics; determining link total attenuation characteristics according to the sum of the statistical physical deviation characteristics, the environment meteorological basic deviation characteristics and the statistical physical attenuation characteristics; and mapping each link total attenuation characteristic into a link attenuation prediction value. The technical scheme can more accurately predict the link attenuation prediction value in a future period of time.

Description

Technical Field

[0001] This application relates to the field of satellite communication technology, specifically to a method, device, and system for predicting environmental meteorological attenuation in satellite-to-ground links. Background Technology

[0002] The communication link between the satellite and the ground station needs to pass through the atmosphere. Clouds, fog, rain and other water condensates in the atmosphere can cause power loss and signal attenuation during the communication process.

[0003] To address the issue of attenuation in satellite-to-ground links, the International Telecommunication Union Radiocommunication Sector (ITU-R) has issued a series of recommendations. For example, Recommendation P.618 provides a semi-empirical statistical physical model for designing geo-space telecommunications systems, and Recommendation P.840 provides a semi-empirical statistical physical model for calculating attenuation caused by clouds and fog.

[0004] These semi-empirical statistical physical models are mainly used to process macroscopic statistical data and give macroscopic attenuation values. Clouds, fog, rain and other water condensates in different regions have different morphological characteristics and different degrees of attenuation of radio waves. Therefore, when the models given in the ITU-R recommendations are applied to specific regions, different degrees of deviation often occur.

[0005] In addition, the word "prediction" in existing satellite-to-ground link attenuation prediction technology usually means "estimation". That is, existing technology lacks a scheme to predict the future satellite-to-ground link attenuation, while predicting the satellite-to-ground link attenuation within a certain period of time in the future is conducive to realizing dynamic control of link power. Summary of the Invention

[0006] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0007] This application provides a method, apparatus, and system for predicting meteorological attenuation in satellite-to-ground links. Based on meteorological forecasts from a meteorological prediction model, attenuation anchor points are provided using models from ITU-R recommendations or their derivatives. Then, a neural network attenuation model is used to determine predicted link attenuation values ​​around these anchor points. This approach not only predicts link attenuation values ​​for a future period but also corrects for deviations caused by applying the ITU-R recommendations or their derivatives to the tested area using the experience of the tested region.

[0008] In some embodiments, the satellite-to-ground link environmental meteorological attenuation prediction method includes: The system acquires current environmental data for the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​for the satellite-to-ground link. The area to be measured is the region for which satellite-to-ground link attenuation data is to be measured. The weather forecasts are determined by a weather prediction model based on the current environmental data; the weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather prediction model. Each statistical physical attenuation value is determined by a statistical physical attenuation model based on each weather forecast and satellite-to-ground link parameter; the statistical physical attenuation model is a model from the ITU-R recommendations or a derivative model thereof. The neural network decay model performs the following process: Extract the link gain deviation characteristics of satellite-to-ground link parameters; The current environmental data is spliced ​​with each weather forecast result to obtain one or more environmental meteorological features, and the environmental meteorological gain deviation feature and environmental meteorological basic deviation feature of each environmental meteorological feature are extracted. Extract the statistical physical attenuation characteristics for each statistical physical attenuation value; The statistical physical deviation characteristics are determined by the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. The total link attenuation characteristic is determined by summing each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic. The total attenuation characteristics of each link are mapped to the predicted link attenuation value.

[0009] Optionally, the loss function during the training of the neural network decay model includes: a main loss term, a meteorological true result loss term, and a high-dimensional consistency loss term; among which, The main loss term consists of the difference between the predicted link attenuation value and the actual link attenuation value when the training samples of historical weather forecast results are input. The loss term for actual meteorological results consists of the difference between the predicted link attenuation value and the actual link attenuation value when training samples of historical actual meteorological results are input. The high-dimensional consistency loss term is composed of the L2 norm of the high-dimensional environmental meteorological basic deviation feature obtained when training samples of historical meteorological results are input.

[0010] Correspondingly, in each training process, two forward propagations are performed in parallel: In a forward propagation process, the input data includes historical weather forecast results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value. In another forward propagation process, the input data includes historical meteorological results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value and high-dimensional characteristic environmental meteorological basic deviation features. The loss is calculated based on the output of two forward propagations and the true value of link attenuation. The parameters of the neural network attenuation model are then optimized based on the calculation results.

[0011] Optionally, the statistical physical deviation characteristics are determined based on the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic, and its corresponding statistical physical attenuation characteristics, including: ; in, To statistically analyze physical deviation characteristics, To statistically analyze the physical attenuation characteristics, To represent the dot product operation, It is the hyperbolic tangent function. This is a characteristic of link gain deviation. This refers to the environmental meteorological gain bias characteristics.

[0012] Optionally, the statistical physical attenuation model is a model optimized from the model in the ITU-R recommendations. The parameter optimization process includes: Minimize the difference between the historical measured attenuation value and the historical statistical physical attenuation value of the area to be tested.

[0013] Optionally, before determining the statistical physical deviation characteristics based on the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic, and its corresponding statistical physical attenuation characteristics, the method further includes: acquiring the regional characteristics of the area to be tested and extracting the regional gain deviation characteristics of the regional characteristics; wherein, the regional characteristics include one or more of the following: meteorological factors causing satellite-to-ground link attenuation and their rate of change at a preset number of time steps in the area to be tested, the spatial distribution characteristics of the meteorological factors causing satellite-to-ground link attenuation in the area to be tested, the climate characteristics of the area to be tested, the terrain characteristics of the area to be tested, the rate of change of antenna elevation angle, and the temperature and humidity at different altitudes along the link path.

[0014] Correspondingly, the statistical physical deviation characteristics are determined based on the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics, including: determining the statistical physical deviation characteristics based on the product of the link gain deviation characteristics, the regional gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics.

[0015] Optionally, the neural network decay model is a model selected from a model library that is adapted to the region to be tested; The model library contains pre-stored neural network decay models corresponding to multiple regions.

[0016] In some embodiments, the satellite-to-ground link environmental meteorological attenuation prediction device includes an acquisition module and a neural network attenuation model; The acquisition module is used to acquire current environmental data of the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​of the satellite-to-ground link. The area to be measured is the region from which satellite-to-ground link attenuation data is to be measured. The weather forecasts are determined by a weather prediction model based on the current environmental data; the weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather prediction model. Each statistical physical attenuation value is determined by a statistical physical attenuation model based on each weather forecast and satellite-to-ground link parameter; the statistical physical attenuation model is a model in the ITU-R recommendations or a derivative model thereof. The neural network decay model is used to perform the following process: Extract the link gain deviation characteristics of satellite-to-ground link parameters; The current environmental data is spliced ​​with each weather forecast result to obtain one or more environmental meteorological features, and the environmental meteorological gain deviation feature and environmental meteorological basic deviation feature of each environmental meteorological feature are extracted. Extract the statistical physical attenuation characteristics for each statistical physical attenuation value; The statistical physical deviation characteristics are determined by the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. The total link attenuation characteristic is determined by summing each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic. The total attenuation characteristics of each link are mapped to the predicted link attenuation value.

[0017] In some embodiments, the satellite-to-ground link environmental weather attenuation prediction device includes a processor and a memory storing program instructions. The processor is configured to execute the satellite-to-ground link environmental weather attenuation prediction method provided in the foregoing embodiments when executing the program instructions.

[0018] In some embodiments, the satellite-to-ground link environmental meteorological attenuation prediction system includes the satellite-to-ground link environmental meteorological attenuation prediction device provided in the foregoing embodiments.

[0019] The satellite-to-ground link environment meteorological attenuation prediction method, apparatus, and system provided in this application embodiment can achieve the following technical effects: Statistical physical attenuation characteristics serve as the benchmark for calculating its deviation; link gain deviation characteristics are gain factors caused by satellite-to-ground link parameters, such as signal polarization and communication frequency, which can lead to different degrees of deviation; environmental and meteorological gain deviation characteristics are gain factors caused by environmental data and weather forecast results, such as clouds with different water content and fog with different concentrations, which can lead to different degrees of deviation.

[0020] The statistical physical deviation characteristics determined by the product of the above three factors can reflect the deviation of the model or its derivative model in the ITU-R recommendations in the area to be tested.

[0021] Meanwhile, environmental meteorological basic deviation features are extracted from environmental meteorological characteristics to reflect the deviations in the meteorological prediction model itself. The total link attenuation feature is determined based on the sum of each statistical physical deviation feature, the environmental meteorological basic deviation feature, and the statistical physical attenuation feature. This ensures that the final mapped link attenuation prediction value fully considers the deviations in the meteorological forecast results. Even if the overall scheme is based on the meteorological forecast results of the meteorological prediction model, it can still improve the accuracy of the link attenuation prediction value as much as possible.

[0022] In addition, the environmental meteorological characteristics are composed of current environmental data and meteorological forecast results. Data fusion methods include addition, multiplication and splicing. Here, addition or multiplication is not used, but splicing is used, which allows the true value (current environmental data) and the value carrying error (meteorological forecast results) to enter the neural network decay model at the same time.

[0023] Among them, meteorological forecast results are used to guide the extraction process of environmental meteorological gain deviation characteristics and environmental meteorological basic deviation characteristics, while current environmental data are used to provide the true values ​​of other dimensions corresponding to the values ​​carrying errors, and to supplement some parameters to assist the extraction process of environmental meteorological gain deviation characteristics and environmental meteorological basic deviation characteristics.

[0024] Meanwhile, during the training process of the neural network attenuation model, the auxiliary role played by this true value helps to implicitly correct the errors in the weather forecast results. When the overall scheme is based on the weather forecast results of the weather prediction model, this can still improve the accuracy of the link attenuation prediction value as much as possible.

[0025] In summary, the technical solution of this application can predict the link attenuation value in a relatively accurate manner over a period of time in the future.

[0026] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0027] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrative descriptions and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are considered similar elements, and wherein: Figure 1 This is a flowchart illustrating a method for predicting meteorological attenuation in a satellite-to-ground link environment, as provided in an embodiment of this application. Figure 2This is a flowchart illustrating another method for predicting environmental meteorological attenuation in a satellite-to-ground link, as provided in an embodiment of this application. Figure 3 This is a schematic diagram of a satellite-to-ground link environmental meteorological attenuation prediction device provided in an embodiment of this application; Figure 4 This is a schematic diagram of another satellite-to-ground link environmental weather attenuation prediction device provided in the embodiments of this application. Detailed Implementation

[0028] To provide a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this application. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0029] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0030] Unless otherwise stated, the term "multiple" means two or more.

[0031] In this embodiment, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0032] The term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0033] Before implementing the satellite-to-ground link environmental meteorological attenuation prediction method provided in this application embodiment, data from multiple sources are first collected. Data sources include, but are not limited to: Weather radar data, meteorological satellite data, ground observation station data, and radiosonde data.

[0034] After data collection, the collected data undergoes preprocessing. Preprocessing operations include, but are not limited to: Spatiotemporal alignment, format conversion, missing value filling, and outlier removal.

[0035] The preprocessed data is then subjected to quality control. The quality control process includes consistency checks, accuracy assessments, and labeling of the quality control results.

[0036] Finally, high-quality data were selected to construct a gridded multi-source fusion meteorological dataset.

[0037] Figure 1 This is a flowchart illustrating a method for predicting environmental meteorological attenuation in a satellite-to-ground link, as provided in an embodiment of this application.

[0038] The method for predicting environmental and meteorological attenuation of the satellite-to-ground link can be executed at the ground station or at the corresponding server at the ground station.

[0039] Combination Figure 1 As shown, the methods for predicting environmental meteorological attenuation in satellite-to-ground links include: S101. Obtain current environmental data of the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​of the satellite-to-ground link.

[0040] In this step, the area to be measured is the area where satellite-to-ground link attenuation data is to be measured, and the weather forecast result is determined by the weather prediction model based on the current environmental data. The weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather forecast model.

[0041] Numerical weather prediction models include, but are not limited to: Global / Regional Assimilation and Prediction Enhanced System (GRAPES), Global Forecast System (GFS), Integrated Forecasting System (IFS), Weather Research and Forecasting (WRF), etc.

[0042] Specifically, multi-source fusion meteorological datasets can be fused with the background field of numerical weather prediction models. The initial field of numerical weather prediction models can be optimized through Weather Research and Forecasting Model Data Assimilation (WRFDA), enabling numerical weather prediction models to more accurately capture the current atmospheric state and the occurrence and development of meteorological phenomena such as clouds, fog, and rain.

[0043] The process typically employs a cyclic assimilation strategy, with an assimilation time window of 3–6 hours.

[0044] Artificial intelligence-based weather forecasting models include, but are not limited to: Pangu meteorological model, GraphCast, NowcastNet, Fuxi meteorological model, etc.

[0045] Each statistical physical attenuation value is determined by the statistical physical attenuation model based on each weather forecast result and satellite-to-ground link parameters. The statistical physical attenuation model is the model in the ITU-R Recommendation or a derivative model thereof.

[0046] For example, ITU-R P.840 provides a statistical physical model for cloud and fog attenuation: ; in, The cloud-to-fog ratio attenuation is expressed in decibels (dB / km). This represents the liquid water ratio attenuation coefficient in clouds and fog, with units of (dB / km) / (g / m³). 3 ), This refers to the communication frequency of the satellite-to-ground link, measured in GHz. The temperature of liquid water in clouds and fog, in Kelvin (K). The density of liquid water in clouds and fog, in g / m³. 3 ).

[0047] ; in, This refers to the communication frequency of the satellite-to-ground link, measured in GHz. As a dimensionless intermediate variable, it is used to reduce the liquid water ratio attenuation coefficient in clouds and fog. Simplify expression , and Where is the complex permittivity of water. Let be the real part of the complex permittivity of water. Let be the imaginary part of the complex permittivity of water, and the two are specifically: ; ; in: The static dielectric constant is ; The dielectric constant for the mid-frequency range, ; The dielectric constant is for the high-frequency band. ; The reciprocal temperature parameter, ; The temperature of liquid water in clouds and fog.

[0048] The primary relaxation frequency, The two are secondary relaxation frequencies, and their calculation methods are as follows: ; .

[0049] The statistical physical models given in other ITU-R recommendations will not be elaborated here; those skilled in the art can refer to the corresponding recommendations.

[0050] Derivative models of the models in the ITU-R recommendations include, but are not limited to: regionally modified derivative models, physical-statistical hybrid derivative models, etc.

[0051] The above-mentioned satellite-to-ground link parameters include, but are not limited to: communication frequency, polarization, antenna elevation angle, ground station latitude and longitude, satellite-to-ground link path length, antenna aperture, and radome loss.

[0052] Furthermore, the statistical physical attenuation model in this embodiment is a model with optimized parameters from the model in the ITU-R recommendations, belonging to a regionally modified derivative model. The parameter optimization process includes: Minimize the difference between the historical measured attenuation value and the historical statistical physical attenuation value of the area to be tested.

[0053] For example, optimizing the local cloud liquid water density M: ; To optimize the cloud-to-fog ratio attenuation, This refers to the communication frequency of the satellite-to-ground link. The temperature of liquid water in clouds and fog. This represents the cloud-to-fog ratio attenuation. The density of liquid water in the clouds mentioned in the proposal, , For nonlinear optimization coefficients, For the link elevation angle, the aforementioned formula In the default .

[0054] The objective function is: ; in, , For nonlinear optimization coefficients, To determine the total number of samples used in the optimization process, the samples used are from the region to be tested. The cloud-to-fog ratio attenuation is calculated using the model provided in the ITU-R recommendations. To optimize the cloud-to-fog ratio attenuation, This refers to the communication frequency of the satellite-to-ground link. The temperature of liquid water in clouds and fog.

[0055] For optimization methods of other models given in the ITU-R recommendations, please refer to the above implementation; they will not be elaborated on here.

[0056] S102. Extract the link gain deviation characteristics of the satellite-to-ground link parameters.

[0057] This step, as well as subsequent steps, are performed by a neural network decay model.

[0058] Specifically, the link gain deviation characteristics of the satellite-to-ground link parameters are extracted, including: ; in, This is a characteristic of link gain deviation. For activation function, This is the learnable matrix corresponding to the satellite-to-ground link parameters. These are the parameters for the satellite-to-ground link, such as communication frequency, polarization, and antenna elevation. This is the learnable bias corresponding to the satellite-to-ground link parameters.

[0059] S103. Combine the current environmental data with each weather forecast result to obtain one or more environmental meteorological features, and extract the environmental meteorological gain deviation feature and the environmental meteorological basic deviation feature for each environmental meteorological feature.

[0060] Specifically: ; in, As environmental and meteorological characteristics, For current environmental data, This indicates a splicing operation. This is the result of a weather forecast.

[0061] Furthermore, the process of stitching together current environmental data with each weather forecast result to obtain one or more environmental and meteorological characteristics is usually performed in the form of matrix operations, i.e., matrix... It contains one or more weather forecast results, matrix It includes one or more environmental meteorological features.

[0062] Of course, this applies to current environmental data with different numbers of parameters (i.e., different dimensions). With weather forecast results First, a dimensional transformation must be performed to make the two have the same dimensions. The dimensional transformation is a basic linear matrix transformation, which will not be elaborated here.

[0063] Extract the environmental meteorological gain bias characteristics and environmental meteorological baseline bias characteristics for each environmental meteorological feature, including: ; in, This refers to the environmental meteorological gain bias characteristics. For activation function, This is the learnable matrix corresponding to the process of extracting environmental meteorological gain bias features from environmental meteorological characteristics. As environmental and meteorological characteristics, This refers to the learnable bias corresponding to the process of extracting environmental meteorological gain bias features from environmental meteorological characteristics.

[0064] ; in, This represents the basic deviation characteristics of environmental meteorology. For activation function, This is the learnable matrix corresponding to the process of extracting basic environmental and meteorological deviation features from environmental and meteorological characteristics. As environmental and meteorological characteristics, This refers to the learnable bias corresponding to the process of extracting basic environmental and meteorological deviation features from environmental and meteorological characteristics.

[0065] S104. Extract the statistical physical attenuation features for each statistical physical attenuation value.

[0066] Specifically: ; in, To statistically analyze the physical attenuation characteristics, For activation function, To statistically analyze the learnable matrix corresponding to the physical attenuation values, To calculate the physical attenuation value, To calculate the learnable bias corresponding to the physical attenuation value.

[0067] S105. The statistical physical deviation characteristics are determined by multiplying the link gain deviation characteristics, the environmental meteorological gain deviation characteristics, and the corresponding statistical physical attenuation characteristics.

[0068] Optionally, the statistical physical deviation characteristics are determined based on the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic, and its corresponding statistical physical attenuation characteristics, including: ; in, To statistically analyze physical deviation characteristics, To statistically analyze the physical attenuation characteristics, To represent the dot product operation, It is the hyperbolic tangent function. This is a characteristic of link gain deviation. This refers to the environmental meteorological gain bias characteristics.

[0069] The above three factors are not directly multiplied, but rather the link gain deviation characteristics are expressed. conduct Processing, on the characteristics of environmental meteorological gain deviation conduct This is handled to avoid directly causing statistical physical bias characteristics when any gain is zero during training. Setting the gradient to zero avoids gradient vanishing and facilitates smoother training.

[0070] In addition, statistical physical attenuation characteristics It is the baseline; no similar processing is applied to it as described above. When it is zero, the statistical physical deviation characteristics are... It loses its meaning, so it is allowed to be zero at this point.

[0071] S106. Determine the total link attenuation characteristic based on the sum of each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic.

[0072] Specifically: ; in, This represents the total link attenuation characteristic. To statistically analyze physical deviation characteristics, This represents the basic deviation characteristics of environmental meteorology. This is for statistical physical attenuation characteristics.

[0073] S107. Map the total attenuation characteristics of each link to the predicted link attenuation value.

[0074] This step is the process of mapping high-dimensional features to scalars, specifically: ; in, This is the predicted value of link attenuation. Let T be the learnable mapping matrix, and let T denote the transpose. This represents the total link attenuation characteristic. This is a learnable mapping bias.

[0075] In the technical solutions provided in the embodiments of this application, statistical physical attenuation characteristics are used as the benchmark for calculating their deviations; link gain deviation characteristics are gain factors caused by satellite-to-ground link parameters, such as signal polarization and communication frequency, which can lead to different degrees of deviation; environmental meteorological gain deviation characteristics are gain factors caused by environmental data and meteorological forecast results, such as clouds with different water content and fog with different concentrations, which can lead to different degrees of deviation.

[0076] The statistical physical deviation characteristics determined by the product of the above three factors can reflect the deviation of the model or its derivative model in the ITU-R recommendations in the area to be tested.

[0077] Meanwhile, environmental meteorological basic deviation features are extracted from environmental meteorological characteristics to reflect the deviations in the meteorological prediction model itself. The total link attenuation feature is determined based on the sum of each statistical physical deviation feature, the environmental meteorological basic deviation feature, and the statistical physical attenuation feature. This ensures that the final mapped link attenuation prediction value fully considers the deviations in the meteorological forecast results. Even if the overall scheme is based on the meteorological forecast results of the meteorological prediction model, it can still improve the accuracy of the link attenuation prediction value as much as possible.

[0078] In addition, the environmental meteorological characteristics are formed by splicing the current environmental data and the meteorological forecast results. The data fusion methods include addition, multiplication and splicing. Here, we do not use addition or multiplication, but use splicing, which allows the true value (current environmental data) and the value carrying the error (meteorological forecast results) to enter the neural network decay model at the same time.

[0079] Among them, meteorological forecast results are used to guide the extraction process of environmental meteorological gain deviation characteristics and environmental meteorological basic deviation characteristics, while current environmental data are used to provide the true values ​​of other dimensions corresponding to the values ​​carrying errors, and to supplement some parameters to assist the extraction process of environmental meteorological gain deviation characteristics and environmental meteorological basic deviation characteristics.

[0080] Meanwhile, during the training of the neural network attenuation model, the auxiliary role played by this true value helps to implicitly correct the errors in the weather forecast results. When the overall scheme is based on the weather forecast results of the weather prediction model, this can still improve the accuracy of the link attenuation prediction value as much as possible.

[0081] In summary, the technical solution of this application can predict the link attenuation value in a relatively accurate manner over a period of time in the future.

[0082] The foregoing embodiments illustrate the internal structure of the neural network decay model, and the training process will be illustrated below.

[0083] Optionally, the loss function during the training process of the neural network decay model includes: the main loss term, the meteorological true result loss term, and the high-dimensional consistency loss term.

[0084] The main loss term consists of the difference between the predicted link attenuation value and the actual link attenuation value when the training samples of historical weather forecast results are input.

[0085] Specifically: ; in, Main loss item, This represents the total number of samples in a single batch of training. For the first batch of training The predicted link attenuation value for each sample For the first batch of training The actual link attenuation value corresponding to each sample.

[0086] It should be noted that the input samples corresponding to the main loss item include "historical weather forecast results", which form a clear contrast with the "historical actual weather results" in other loss items.

[0087] The input samples corresponding to the main loss term include historical weather forecast results, which enables the link attenuation prediction value output by the neural network attenuation model to continuously approach the true link attenuation value during the actual forecasting process, so as to ensure the accuracy of the final prediction.

[0088] The loss term for actual meteorological results consists of the difference between the predicted link attenuation value and the actual link attenuation value when training samples of historical actual meteorological results are input.

[0089] Specifically: ; in, For the loss item of the actual meteorological results, This represents the total number of samples in a single batch of training. For the first batch of training The predicted link attenuation value for each sample For the first batch of training The actual link attenuation value corresponding to each sample.

[0090] It should be noted that the input sample corresponding to the meteorological actual result loss item includes "historical meteorological actual results," and correspondingly, the link attenuation prediction value. The suffix "_t" is used in conjunction with the predicted link attenuation value when inputting "historical weather forecast results". Make a distinction.

[0091] Because historical meteorological data is input, this loss term can exclude weather errors, i.e., the aforementioned environmental meteorological baseline deviation characteristics. When the value is zero, the total link attenuation characteristic is... Statistical physical deviation characteristics ,Right now, In this case, to fit the input and output, the neural network decay model is forced to continuously adjust. Parts and Some network parameters are used to compensate for the bias of the model in the ITU-R recommendations in the tested area.

[0092] The high-dimensional consistency loss term is composed of the L2 norm of the high-dimensional environmental meteorological basic deviation feature obtained when training samples of historical meteorological results are input.

[0093] Specifically: ; in, For high-dimensional consistency loss terms, As for the basic characteristics of environmental meteorological deviations, it should be noted that the aforementioned basic characteristics of environmental meteorological deviations are based on... This indicates that the training samples include "historical weather data," hence the use of the "_t" suffix for differentiation. This represents the L2 norm.

[0094] When the training samples include "historical meteorological real results", this environmental meteorological baseline bias characteristic It should be zero. This high-dimensional consistency loss term directly affects... This makes it zero, which is a direct supplement to the loss item of the actual meteorological results.

[0095] Correspondingly, in each training process, two forward propagations are performed in parallel: In a forward propagation process, the input data includes historical weather forecast results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value. In another forward propagation process, the input data includes historical meteorological results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value and high-dimensional characteristic environmental meteorological basic deviation features. The loss is calculated based on the output of two forward propagations and the true value of link attenuation. The parameters of the neural network attenuation model are then optimized based on the calculation results.

[0096] Figure 2 This is a flowchart illustrating another method for predicting environmental meteorological attenuation in satellite-to-ground links provided in this application embodiment.

[0097] The method for predicting environmental and meteorological attenuation of the satellite-to-ground link can be executed at the ground station or at the corresponding server at the ground station.

[0098] Combination Figure 2 As shown, the methods for predicting environmental meteorological attenuation in satellite-to-ground links include: S201. Obtain current environmental data of the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​of the satellite-to-ground link.

[0099] In this step, the area to be measured is the area where satellite-to-ground link attenuation data is to be measured. The weather forecast results are determined by the weather prediction model based on the current environmental data. The weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather forecast model. Each statistical physical attenuation value is determined by the statistical physical attenuation model based on each weather forecast result and satellite-to-ground link parameters. The statistical physical attenuation model is the model in the ITU-R recommendations or its derivative model.

[0100] S202. Extract the link gain deviation characteristics of the satellite-to-ground link parameters.

[0101] This step, as well as subsequent steps, are performed by a neural network decay model.

[0102] S203. Obtain the regional characteristics of the region to be tested, and extract the regional gain deviation features of the regional characteristics.

[0103] Among them, the regional characteristics include one or more of the following: the meteorological factors and their rate of change that cause satellite-to-ground link attenuation in the region under test for a preset number of time steps, the spatial distribution characteristics of the meteorological factors that cause satellite-to-ground link attenuation in the region under test, the climate characteristics of the region under test, the terrain characteristics of the region under test, the rate of change of antenna elevation angle, and the temperature and humidity at different altitudes along the link path.

[0104] Specifically: ; in, This represents the regional gain bias characteristics. For activation function, A learnable matrix corresponding to regional characteristics. A matrix of regional characteristics. Learnable biases corresponding to regional characteristics.

[0105] S204. Combine the current environmental data with each weather forecast result to obtain one or more environmental meteorological features, and extract the environmental meteorological gain deviation feature and the environmental meteorological basic deviation feature for each environmental meteorological feature.

[0106] S205. Extract the statistical physical attenuation features for each statistical physical attenuation value.

[0107] The above steps S202 to S205 are parallel feature extraction processes, and there is no distinction between them in order.

[0108] S206. The statistical physical deviation characteristics are determined by multiplying the link gain deviation characteristics, the regional gain deviation characteristics, the environmental meteorological gain deviation characteristics, and their corresponding statistical physical attenuation characteristics.

[0109] Specifically: ; in, To statistically analyze physical deviation characteristics, To statistically analyze the physical attenuation characteristics, To represent the dot product operation, It is the hyperbolic tangent function. This represents the regional gain bias characteristics. This is a characteristic of link gain deviation. This refers to the environmental meteorological gain bias characteristics.

[0110] S207. Determine the total link attenuation characteristic based on the sum of each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic.

[0111] S208. Map the total attenuation characteristics of each link to the predicted link attenuation value.

[0112] In this embodiment, more dimensional empirical data of the test area are incorporated into the neural network attenuation model to compensate for the feature parameters of specific areas that are ignored by the model in the ITU-R recommendations, thereby further improving the accuracy of the predicted link attenuation values ​​for the test area.

[0113] The foregoing embodiments provide an exemplary illustration of a scenario based on a region to be tested and a neural network decay model.

[0114] In some further application scenarios, a model library is first set up, which pre-stores neural network decay models corresponding to multiple regions.

[0115] Each neural network decay model adopts the structure and training method provided in the aforementioned embodiments.

[0116] For the region to be tested, the neural network decay model is a model selected from the model library that is suitable for the region to be tested.

[0117] Furthermore, a region in the model library can correspond to multiple neural network attenuation models. Different neural network models correspond to different prediction accuracies (and also different computing power requirements), and different neural networks correspond to different bands.

[0118] For the area to be tested, the corresponding neural network attenuation model is selected based on the accuracy requirements and the actual band requirements.

[0119] Figure 3 This is a schematic diagram of a satellite-to-ground link environmental meteorological attenuation prediction device provided in an embodiment of this application.

[0120] This satellite-to-ground link environmental meteorological attenuation prediction device can be implemented through software, hardware, or a combination of both.

[0121] Combination Figure 3 As shown, the satellite-to-ground link environmental meteorological attenuation prediction device includes an acquisition module 31 and a neural network attenuation model 32; The acquisition module 31 is used to acquire the current environmental data of the area to be measured, one or more future weather forecast results, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​of the satellite-to-ground link. The area to be measured is the area where the satellite-to-ground link attenuation data is to be measured. The weather forecast results are determined by a weather prediction model based on the current environmental data. The weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather prediction model. Each statistical physical attenuation value is determined by a statistical physical attenuation model based on each weather forecast result and satellite-to-ground link parameters. The statistical physical attenuation model is a model in the ITU-R recommendations or a derivative model thereof. The neural network decay model 32 is used to perform the following process: Extract the link gain deviation characteristics of satellite-to-ground link parameters; The current environmental data is spliced ​​with each meteorological forecast result to obtain one or more environmental meteorological features, and the environmental meteorological gain deviation feature and environmental meteorological basic deviation feature of each environmental meteorological feature are extracted. Extract the statistical physical attenuation characteristics for each statistical physical attenuation value; The statistical physical deviation characteristics are determined by the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. The total link attenuation characteristic is determined by summing each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic. The total attenuation characteristics of each link are mapped to the predicted link attenuation value.

[0122] The exemplary descriptions and specific limitations of the features and steps in the aforementioned embodiments of the satellite-to-ground link environmental meteorological attenuation prediction method also apply to this satellite-to-ground link environmental meteorological attenuation prediction device, and will not be repeated here.

[0123] In some embodiments, the satellite-to-ground link environmental weather attenuation prediction device includes a processor and a memory storing program instructions. The processor is configured to execute the satellite-to-ground link environmental weather attenuation prediction method provided in the foregoing embodiments when executing the program instructions.

[0124] In some embodiments, the satellite-to-ground link environmental meteorological attenuation prediction system includes the satellite-to-ground link environmental meteorological attenuation prediction device provided in the foregoing embodiments.

[0125] Figure 4 This is a schematic diagram of another satellite-to-ground link environmental weather attenuation prediction device provided in the embodiments of this application.

[0126] Combination Figure 4As shown, the satellite-to-ground link environmental meteorological attenuation prediction device includes: The processor 41 and memory 42 may also include a communication interface 43 and a bus 44. The processor 41, communication interface 43, and memory 42 can communicate with each other via the bus 44. The communication interface 43 can be used for information transmission. The processor 41 can call logical instructions in the memory 42 to execute the satellite-to-ground link environmental meteorological attenuation prediction method provided in the foregoing embodiments.

[0127] Furthermore, the logical instructions in the aforementioned memory 42 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0128] The memory 42, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this application. The processor 41 executes functional applications and data processing by running the software programs, instructions, and modules stored in the memory 42, thereby implementing the methods in the above-described method embodiments.

[0129] The memory 42 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 42 may include high-speed random access memory and may also include non-volatile memory.

[0130] This application provides a computer-readable storage medium storing computer-executable instructions, which are configured to execute the satellite-to-ground link environmental meteorological attenuation prediction method provided in the foregoing embodiments.

[0131] This application provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to perform the satellite-to-ground link environmental meteorological attenuation prediction method provided in the aforementioned embodiment.

[0132] The aforementioned computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0133] The technical solutions of this application embodiment can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in this application embodiment. The aforementioned storage medium can be a non-transitory storage medium, including: USB flash drive, portable hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, and other media capable of storing program code; it can also be a transient storage medium.

[0134] The foregoing description and accompanying drawings fully illustrate embodiments of this application to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operations may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Additionally, when used in this application, the terms “comprise” and its variations “comprises” and / or “comprising” refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Unless otherwise specified, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes that element. In this document, each embodiment may focus on describing the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, then the relevant parts can be referred to the description of the method section.

[0135] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0136] The methods and products (including but not limited to devices and equipment) disclosed in the embodiments herein can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units may be merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed units may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms. Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, i.e., they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to implement this embodiment according to actual needs. Furthermore, the functional units in the embodiments of this application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0137] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A method for predicting environmental meteorological attenuation in satellite-to-ground links, characterized in that, include: The system acquires current environmental data for the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​for the satellite-to-ground link. The area to be measured is the region for which satellite-to-ground link attenuation data is to be measured. The weather forecasts are determined by a weather prediction model based on the current environmental data; the weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather prediction model. Each statistical physical attenuation value is determined by a statistical physical attenuation model based on each weather forecast and satellite-to-ground link parameter; the statistical physical attenuation model is a model from the ITU-R recommendations or a derivative model. The neural network decay model performs the following process: Extract the link gain deviation characteristics of the satellite-to-ground link parameters; the link gain deviation characteristics are the gain factors caused by the satellite-to-ground link parameters. The current environmental data is spliced ​​with each weather forecast result to obtain one or more environmental meteorological features. The environmental meteorological gain deviation feature and the environmental meteorological basic deviation feature of each environmental meteorological feature are extracted. The environmental meteorological gain deviation feature is the gain factor caused by the environmental data and weather forecast results, and the environmental meteorological basic deviation feature reflects the deviation of the weather prediction model itself. Extract the statistical physical attenuation characteristics for each statistical physical attenuation value; the statistical physical attenuation characteristics serve as the benchmark for calculating the statistical physical deviation characteristics; The statistical physical deviation characteristics are determined by the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. The total link attenuation characteristic is determined by summing each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic. The total attenuation characteristics of each link are mapped to the predicted link attenuation value.

2. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to claim 1, characterized in that, The loss function during the training of the neural network decay model includes: the main loss term, the meteorological true result loss term, and the high-dimensional consistency loss term; among them, The main loss term consists of the difference between the predicted link attenuation value and the actual link attenuation value when the training samples of historical weather forecast results are input. The loss term for actual meteorological results consists of the difference between the predicted link attenuation value and the actual link attenuation value when training samples of historical actual meteorological results are input. The high-dimensional consistency loss term is composed of the L2 norm of the high-dimensional environmental meteorological basic deviation feature obtained when training samples of historical meteorological results are input.

3. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to claim 2, characterized in that, During each training session, two forward propagations are performed in parallel: In a forward propagation process, the input data includes historical weather forecast results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value. In another forward propagation process, the input data includes historical meteorological results, historical environmental data, and historical statistical physical attenuation values, and the output is the link attenuation prediction value and high-dimensional characteristic environmental meteorological basic deviation features. The loss is calculated based on the output of two forward propagations and the true value of link attenuation. The parameters of the neural network attenuation model are then optimized based on the calculation results.

4. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to claim 1, characterized in that, The statistical physical deviation characteristics are determined by multiplying the link gain deviation characteristics, the gain deviation characteristics of each environmental meteorological environment, and their corresponding statistical physical attenuation characteristics. These characteristics include: ； in, To statistically analyze physical deviation characteristics, To statistically analyze the physical attenuation characteristics, To represent the dot product operation, It is the hyperbolic tangent function. This is a characteristic of link gain deviation. This refers to the environmental meteorological gain bias characteristics.

5. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to claim 1, characterized in that, The statistical physical attenuation model is a parameter-optimized version of the model in the ITU-R recommendations. The parameter optimization process includes: Minimize the difference between the historical measured attenuation value and the historical statistical physical attenuation value of the area to be tested.

6. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to claim 1, characterized in that, Before determining the statistical physical deviation characteristics based on the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic, and its corresponding statistical physical attenuation characteristics, the process also includes: acquiring the regional characteristics of the area to be tested and extracting the regional gain deviation characteristics of the regional characteristics; wherein, the regional characteristics include one or more of the following: meteorological factors causing satellite-to-ground link attenuation and their rate of change at a preset number of time steps in the area to be tested, the spatial distribution characteristics of the meteorological factors causing satellite-to-ground link attenuation in the area to be tested, the climate characteristics of the area to be tested, the terrain characteristics of the area to be tested, the rate of change of antenna elevation angle, and the temperature and humidity at different altitudes along the link path; The statistical physical deviation characteristics are determined by multiplying the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. This includes determining the statistical physical deviation characteristics by multiplying the link gain deviation characteristics, the regional gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics.

7. The method for predicting environmental meteorological attenuation of satellite-to-ground links according to any one of claims 1 to 6, characterized in that, The neural network decay model is a model selected from the model library that is adapted to the region to be tested; The model library contains pre-stored neural network decay models for multiple regions.

8. A satellite-to-ground link environmental meteorological attenuation prediction device, characterized in that, include: The acquisition module is used to acquire current environmental data of the area to be measured, one or more future weather forecasts, satellite-to-ground link parameters, and one or more statistical physical attenuation values ​​of the satellite-to-ground link. The area to be measured is the region from which satellite-to-ground link attenuation data is to be measured. The weather forecasts are determined by a weather prediction model based on the current environmental data; the weather prediction model is a numerical weather prediction model or an artificial intelligence-based weather prediction model. Each statistical physical attenuation value is determined by a statistical physical attenuation model based on each weather forecast and satellite-to-ground link parameter; the statistical physical attenuation model is a model in the ITU-R recommendations or a derivative model thereof. A neural network decay model is used to perform the following process: Extract the link gain deviation characteristics of the satellite-to-ground link parameters; the link gain deviation characteristics are the gain factors caused by the satellite-to-ground link parameters. The current environmental data is spliced ​​with each weather forecast result to obtain one or more environmental meteorological features. The environmental meteorological gain deviation feature and the environmental meteorological basic deviation feature of each environmental meteorological feature are extracted. The environmental meteorological gain deviation feature is the gain factor caused by the environmental data and weather forecast results, and the environmental meteorological basic deviation feature reflects the deviation of the weather prediction model itself. Extract the statistical physical attenuation characteristics for each statistical physical attenuation value; the statistical physical attenuation characteristics serve as the benchmark for calculating the statistical physical deviation characteristics; The statistical physical deviation characteristics are determined by the product of the link gain deviation characteristics, each environmental meteorological gain deviation characteristic and its corresponding statistical physical attenuation characteristics. The total link attenuation characteristic is determined by summing each statistical physical deviation characteristic, environmental meteorological basic deviation characteristic, and statistical physical attenuation characteristic. The total attenuation characteristics of each link are mapped to the predicted link attenuation value.

9. A satellite-to-ground link environmental meteorological attenuation prediction device, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute the satellite-to-ground link environmental meteorological attenuation prediction method as described in any one of claims 1 to 7 when executing the program instructions.

10. A satellite-to-ground link environmental meteorological attenuation prediction system, characterized in that, Includes the satellite-to-ground link environmental weather attenuation prediction device as described in claim 8 or 9.

Images