A method for predicting long-term solar activity level based on F107 index annual monthly average

Abstract

This invention discloses a method for predicting long-term solar activity levels based on the annual slip-monthly mean of the F107 index, comprising: S1: establishing a solar activity level prediction model; S2: establishing a database of monthly average solar F107 index values; S3: obtaining the target prediction month; S4: using an RBF neural network to predict the annual slip-monthly mean of the solar F107 index for the n months following the target month, and filling it into the database of monthly average solar F107 index values; S5: traversing the database of monthly average solar F107 index values ​​to obtain the monthly average solar F107 index values ​​for the target prediction month and the preceding few months; S6: substituting the monthly average solar F107 index values ​​for the target prediction month and the preceding few months into the solar activity level prediction model; S7: outputting the annual slip-monthly mean of the solar F107 index and its prediction result. This invention can accurately predict solar activity parameters for the current month and the following months, and use them as a reference for obtaining solar activity levels. Furthermore, it provides superior academic and application value as input parameters for forecasting other related parameters.

Description

Technical Field

[0001] This invention relates to the field of solar activity levels in astrophysics, and in particular to a method for predicting long-term solar activity levels. Background Technology

[0002] The solar radio flux at a wavelength of 10.7 cm (2800 MHz) is a key parameter characterizing the level of solar activity, and is known in astrophysics as the F107 index (also called the solar F10.7 index).

[0003] The internationally accepted method uses the 13-month monthly average of the solar F107 index, Fi, to calculate the smoothed monthly average of the solar F107 index for the i-th month (denoted as Fsi): , where Fi is the monthly average of the solar F107 index, which is the value obtained by averaging the daily solar F107 index over the natural month.

[0004] The solar F107 exponential smoothed monthly average (Fsi) is actually calculated based on the value of the median month out of every 13 months. For example, if data from a certain 13-month period is used, the calculated Fsi value is for the 7th month. One drawback of this calculation method is that the latest value is obtained with a time lag of n months. Regardless of how many consecutive months of data are used, this calculation method will objectively introduce some degree of delay. Another drawback is that if the Fsi value for a given month needs to be obtained in real time for subsequent research, the value for that month actually uses measured values ​​from the next 6 months. If this is used as an input factor in a forecast, it's equivalent to incorporating future measured information. In other words, the forecast for the next 6 months using this value includes information from the forecast month itself and even information from after the forecast month, containing a factor of predicting itself, or even predicting the past with the future, which is logically illogical. Furthermore, it fails to reflect real-time information, making it impossible to use Fsi for further academic research and causing delays in related applications.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for predicting long-term solar activity levels based on the annual sliding monthly mean of the F107 index, which can solve the problems existing in the background technology.

[0007] To achieve the above objectives, this invention provides a method for predicting long-term solar activity levels based on the annual slip-monthly average of the F107 index, characterized by comprising: S1: establishing a solar activity level prediction model; S2: establishing a database of monthly average solar F107 index values; S3: obtaining the target prediction month; S4: using an RBF neural network to predict the annual slip-monthly average of the solar F107 index for the n months following the target month, and filling it into the database of monthly average solar F107 index values; S5: traversing the database of monthly average solar F107 index values ​​to obtain the monthly average solar F107 index values ​​for the target prediction month and the preceding several months; S6: substituting the monthly average solar F107 index values ​​for the target prediction month and the preceding several months into the solar activity level prediction model; S7: outputting the annual slip-monthly average of the solar F107 index as the prediction result.

[0008] In one or more embodiments, step S1 includes: S11: inputting the calculation formula for the annual monthly average value of the solar F107 index; S12: determining the input quantity and output quantity; S13: generating input ports and output ports according to the quantity of the input quantity and output quantity.

[0009] In one or more embodiments, in step S11, the formula for calculating the annual slippage monthly average of the solar F107 index is:

[0010] Wherein, i month refers to the target predicted month, Fsyi is the annual average monthly value of the F107 index in the i month, Fi is the monthly average solar F107 index in the i month, Fi-1 is the monthly average solar F107 index one month prior to the i month, and so on, Fi-11 is the monthly average solar F107 index 11 months prior to the i month.

[0011] In one or more embodiments, step S2 includes: S21: obtaining observation data from an authoritative data source; S22: verifying the observation data to ensure that the data contains timestamps and monthly average observation values ​​of the solar F107 index; S23: establishing a correspondence between the timestamps and the monthly average observation values ​​of the solar F107 index; S24: completing the monthly average solar F107 index database based on the correspondence; S25: checking data integrity and replacing any missing values.

[0012] In one or more embodiments, step S4 includes: S41: setting the prediction algorithm of the RBF network and determining the input layer dimension of the RBF neural network; S42: selecting the RBF center point initialization method; S43: determining the number of hidden layer nodes; S44: selecting the radial basis function type; S45: predicting the annual slip monthly average of the solar F107 index and filling the predicted value into the corresponding month of the solar F107 index monthly average database; S46: after the prediction process, adopting a loss optimization strategy to reduce the error of the predicted value and adjusting the prediction algorithm.

[0013] In one or more embodiments, the loss optimization strategy in step S46 includes: S461: Traversing the monthly average database of the solar F107 index, finding the timestamp, and dividing the training set / validation set / test set according to the time order; S462: Performing unsupervised learning; S463: Performing supervised learning; S464: Performing parameter fine-tuning and adjusting the hyperparameters using cross-validation.

[0014] In one or more embodiments, in step S461, the training set calls past data from the monthly average solar F107 index database, and the test set calls current data from the monthly average solar F107 index database.

[0015] Thirdly, this invention provides a system for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index, based on the same concept as the method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index, including:

[0016] The first module is used to establish a predictive model for solar activity levels.

[0017] The second module is used to establish a database of monthly average values ​​of the solar F107 index.

[0018] The first acquisition module is used to acquire the target predicted month;

[0019] The first prediction module uses an RBF neural network to predict the annual slip monthly average of the solar F107 index for n months after the target month, and fills it into the solar F107 index monthly average database.

[0020] The first acquisition module is used to traverse the monthly average value database of the solar F107 index and acquire the monthly average value data of the solar F107 index for the target prediction month and the previous several months.

[0021] The first input module is used to input the target prediction month and the monthly average value of the solar F107 index of the previous several months into the solar activity level prediction model.

[0022] The first output module is used to predict the annual slippage monthly average of the solar F107 index.

[0023] In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method for predicting long-term solar activity levels based on the annual sliding monthly mean of the F107 index.

[0024] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for predicting long-term solar activity levels based on the annual slip-month average of the F107 index.

[0025] Compared with the prior art, the various technical solutions and embodiments provided by the present invention include at least the following technical effects or advantages:

[0026] By inventing a new annual slip-monthly average of the solar F107 index, a more rigorous and academically valuable reference quantity is provided to reduce the delay in solar activity prediction. After a calendar month ends, the measured monthly average of the F107 index is obtained. By querying the measured monthly averages of the F107 index for the preceding 11 months, the average of these 12 monthly averages is calculated and recorded as the annual slip-monthly average of the current month. This value also has two other advantages: firstly, the annual slip-monthly average contains more physical information, as it includes information from the current month itself and the preceding 11 months, totaling 12 months (one year); secondly, this method yields an annual slip-monthly average sequence that is smoother than the original monthly average sequence, thus offering advantages in long-term parameter correlation.

[0027] By setting a programmable calculation program, the calculation and prediction of the annual slip monthly average of the solar F107 index can be arbitrarily edited and adjusted. By setting a loss optimization strategy and utilizing the monthly average database of the solar F107 index, the prediction algorithm can be optimized on a large scale. The monthly average database of the solar F107 index stores all authoritative data that can be queried. By performing simulation calculations at historical nodes, the predicted value can be optimized to approximate the actual value, reducing the error to within the allowable error range of the actual observed value, and achieving an accuracy equal to that of the actual measurement. Attached Figure Description

[0028] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and are intended to explain the invention. They are intended to be seen as illustrative of various combinations of preferred embodiments and do not constitute an undue limitation of the invention. In the drawings:

[0029] Figure 1 A schematic diagram of the overall process for a method to predict long-term solar activity levels based on the annual sliding monthly mean of the F107 index, provided by the present invention.

[0030] Figure 2 The method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index provided by this invention includes the prediction, actual measurement, and error of the first month of the future F107 index from January 2009 to April 2025;

[0031] Figure 3 The present invention provides a method for predicting long-term solar activity levels based on the annual slip monthly mean of the F107 index. The prediction, actual measurement and error of the F107 index for the second month of the future from January 2009 to April 2025 are presented.

[0032] For necessary explanation and distinction, the purple curve represents the measured value, the red curve represents the predicted value, and the dashed line represents the range of measured values ​​with a fluctuation of 10%; the red curve being obscured by the purple curve indicates that the predicted value and the actual value are highly consistent; the blue curves from top to bottom represent the absolute error and the relative error, respectively. Detailed Implementation

[0033] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises", "made for", etc., shall be understood to include the stated elements or components, without excluding other elements or other components.

[0034] This invention defines an innovative annual slip monthly mean of the solar F107 index to reflect solar activity, and through program design, eliminates the forecast delay of the solar F107 index and can predict the future annual slip monthly mean of the solar F107 index, thereby accurately predicting solar activity.

[0035] Example 1:

[0036] This embodiment provides a method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index, including:

[0037] S1: Establish a model for predicting solar activity levels;

[0038] Specifically, this step defines a method for calculating solar activity levels. By inputting variable data into this method, the program can obtain the annual monthly average value of the solar F107 index.

[0039] In a preferred embodiment, step S1 includes:

[0040] S11: Input the calculation formula for the annual monthly average value of the solar F107 index;

[0041] S12: Determine the input and output quantities;

[0042] S13: Generate input ports and output ports based on the number of input and output quantities.

[0043] In a preferred embodiment of this example, the formula for calculating the annual average monthly value of the solar F107 index in step S11 is as follows:

[0044]

[0045] Wherein, i month refers to the target predicted month, Fsyi is the annual average monthly value of the F107 index in the i month, Fi is the monthly average solar F107 index in the i month, Fi-1 is the monthly average solar F107 index one month prior to the i month, and so on, Fi-11 is the monthly average solar F107 index 11 months prior to the i month.

[0046] Specifically, step S1 includes defining a formula, determining input and output variables, and allocating input and output ports to the system. After steps S11 to S13, the formula for calculating the annual monthly average value of the solar F107 index is converted into executable program code. By inputting information into the input port, the accurate calculation result of the formula can be obtained. This calculation result is in data form and can be read and processed by the program.

[0047] As a more preferred implementation, the process is further extended based on the annual average monthly sliding value of the solar F107 index, with different weights assigned to each of the 12 months. The calculation method is shown in the formula:

[0048] .

[0049] In the formula, c0 to c11 represent the weighting coefficients of the months calculated up to the 11 months preceding them, Fsyc is the annual slip monthly average under different weighting coefficients, and the remaining parameters have the same meaning as the calculation formula for the unweighted annual slip monthly average of the solar F107 index. Generally, the weighting coefficient of the month being calculated is the largest, and the weighting coefficient decreases as it moves further away from the month. An initial set of weighting coefficients is given first; for example, c0 to c11 are given as an arithmetic sequence with a sum of 1, such as: [33 / 264, 31 / 264, 29 / 264, 27 / 264, 25 / 264, 23 / 264, 21 / 264, 19 / 264, 17 / 264, 15 / 264, 13 / 264, 11 / 264]. The weighting coefficients can be adjusted in the analysis of the relevant parameter prediction model to seek the optimal coefficient combination to achieve the best correlation. When implementing the loss optimization strategy, the monthly average weights are also optimized to achieve better prediction results.

[0050] S2: Establish a monthly average database of the solar F107 index; wherein, the purpose of establishing the monthly average database of the solar F107 index is to remove useless information and classify and summarize useful information, such as the calculation of the annual monthly average of the solar F107 index requiring time and the monthly average of the solar F107 index as variables, and combine the two and their corresponding relationship into a database.

[0051] In a preferred embodiment of this example, step S2 includes:

[0052] S21: Obtain observation data from authoritative data sources;

[0053] S22: Verify the observation data to ensure that the data includes timestamps and monthly average observations of the solar F107 index;

[0054] S23: Establish the correspondence between the timestamps and the monthly average observations of the solar F107 index;

[0055] S24: Based on the aforementioned correspondence, complete the monthly average value database of the solar F107 index;

[0056] S25: Check data integrity; if missing values ​​are found, replace them.

[0057] Specifically, in order to establish reliable training / validation / test set data support, observation data is obtained from authoritative data sources, and the time is filled into the monthly average value of the solar F107 index corresponding to the monthly average value of the solar F107 index.

[0058] S3: Obtain the target forecast month;

[0059] Specifically, the program receives user input and determines the month to be predicted from the monthly average value database of the solar F107 index.

[0060] S4: Use the RBF neural network to predict the annual slip monthly average of the solar F107 index for the n months following the target month, and fill it into the solar F107 index monthly average database;

[0061] In a preferred embodiment of this example, step S4 includes:

[0062] S41: Set the prediction algorithm for the RBF network and determine the input layer dimension of the RBF neural network;

[0063] S42: Select the RBF center point initialization method;

[0064] S43: Determine the number of hidden layer nodes;

[0065] S44: Select the radial basis function type;

[0066] S45: Predict the annual monthly average value of the solar F107 index and fill the predicted value into the corresponding month of the monthly average value database of the solar F107 index;

[0067] S46: After the prediction process, a loss optimization strategy is adopted to reduce the error of the predicted value and adjust the prediction algorithm.

[0068] Specifically, in order to obtain the predicted results of the monthly average value of the solar F107 index, this implementation method predicts the future trend based on the existing values ​​of the monthly average value database of the solar F107 index.

[0069] In a preferred embodiment of this example, the loss optimization strategy in step S46 includes:

[0070] S461: Traverse the monthly average database of the solar F107 index, find the timestamp, and divide the training set / validation set / test set according to the time order;

[0071] S462: Perform unsupervised learning;

[0072] S463: Conduct supervised learning;

[0073] S464: Perform parameter fine-tuning and adjust hyperparameters using cross-validation.

[0074] In a preferred embodiment of this example, in step S461, the training set calls past data from the monthly average solar F107 index database, and the test set calls current data from the monthly average solar F107 index database.

[0075] Specifically, the training set is used as the corpus, the validation set is used to select better algorithm strategies, and the test set is used to predict the future annual average monthly value of the solar F107 index, and wait for the measured value to be obtained for application in real-world scenarios. A preferred method is to select a specific past time period as the training set and hide its actual results, simulate prediction using the algorithm, and use the actual results as the validation set for optimization. The optimized algorithm is then applied to actual predictions and tested using the latest monthly average of the solar F107 index.

[0076] To avoid using known data from a particular month to predict its own performance, the system hides the actual data for the next n months. Instead, the algorithm uses data from the previous n months to predict the data for the next n months, assuming that the data for the next n months is unavailable. The predicted value is then compared with the actual value, and the algorithm is optimized to make the predicted value closer to the actual value. Through extensive computation, the prediction algorithm can accurately predict the annual average monthly value of the solar F107 index. The loss optimization strategy in this implementation aims to reduce the difference between the predicted and actual values ​​to within the allowable error range of the internationally measured monthly average of the solar F107 index. In other words, even the measured monthly average of the solar F107 index has errors, and the predicted value in this implementation can be optimized to within this allowable error range. Therefore, the predicted value is generally considered by the industry to be equivalent to the measured value, and the monthly average of the solar F107 index calculated based on this is also recognized by the industry as an accurate value.

[0077] S5: Traverse the monthly average value database of the solar F107 index to obtain the monthly average value data of the solar F107 index for the target prediction month and the previous several months.

[0078] S6: Substitute the target prediction month and the monthly average value of the solar F107 index of the previous few months into the solar activity level prediction model.

[0079] S7: Output the annual monthly average of the solar F107 index as the prediction result.

[0080] Specifically, after step S4, the solar activity level prediction model has an accurate solar activity level prediction algorithm. The predicted monthly average value of the solar F107 index is substituted into steps S5, S6, and S7 to predict the annual average value of the solar F107 index. After obtaining the prediction results, those skilled in the art or other related fields can conduct in-depth research based on the annual average value of the solar F107 index, grasp the latest laws of solar activity, and provide more accurate practical advancements for matters such as weather forecasting.

[0081] The actual execution of this embodiment is as follows: Figure 2 , 3As shown, the results of Fsy are adjusted using a radial basis function (RBF) network method. Data from January 1948 to December 2008 is selected and trained using a radial basis function (RBF) network. The observation values ​​D(n-2), D(n-1), and D(n) of the first three months are used as the network's input vectors, and the observation values ​​D(n+1) and D(n+2) of the first and second months after that are used as the output vectors. Then, data from January 2009 to April 2025 is selected for a forecast experiment. The observation values ​​D(n-2), D(n-1), and D(n) of the first three months are used as the network's input vectors, and the output vectors are the predicted values ​​D(n+1) and D(n+2) of the first and second months of the future. This process can be repeated to make forecasts for several months or even longer. Figure 1 and Figure 2 The predicted and measured values ​​of the annual sliding monthly average solar radio flux from January 2009 to April 2025 for the first and second months of the future, along with the absolute and relative errors, are given respectively.

[0082] For the predicted value D(n+1) for the first month, among the 193 predictions, the absolute error of all predictions is within ±7%, of which 187 predictions have an absolute error within ±3.0%, accounting for approximately 97% of the total predictions. All relative errors are within ±4%, of which 181 predictions have a relative error of no more than ±2%, accounting for approximately 94% of the total predictions.

[0083] For the predicted value D(n+2) for the second month, among the 192 predictions, all predictions have an absolute error within ±8%, with 179 predictions having an absolute error within ±5.0%, accounting for approximately 93.2% of the total predictions. All relative errors are within ±7%, with 186 predictions having a relative error within ±5%, accounting for approximately 96.9% of the total predictions.

[0084] This paper forecasts solar activity for the next 1-2 months. If the Fsy index is used as input, and compared with the traditional F107 index smoothed monthly average, the forecast lead time can be advanced by 5-6 months while maintaining the same forecast accuracy.

[0085] For example, taking March 1, 2025 as the forecast starting point, the latest measured value of the Fsy index is for February 2025, while the latest measured value of the smoothed monthly average of the F107 index is for August 2024. Similarly, when forecasting for the next two months, using the Fs index only yields predictions for September and October 2024, which are the 6th and 5th months prior to the forecast starting point (March 2025), failing to meet operational forecasting needs. However, using the Fsy index yields predictions for March and April 2025, providing predictions for one to two months prior to the forecast starting point, thus meeting operational forecasting requirements. Therefore, this scheme has achieved the accuracy standard of forecast error within the acceptable error range of measured values ​​and can serve as a reference for long-term solar activity levels.

[0086] Example 2:

[0087] This embodiment provides a system for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 exponent. Based on the same concept, this embodiment implements a method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 exponent, as described in Embodiment 1, including:

[0088] The first module is used to establish a predictive model for solar activity levels.

[0089] The second module is used to establish a database of monthly average values ​​of the solar F107 index.

[0090] The first acquisition module is used to acquire the target predicted month;

[0091] The first prediction module uses an RBF neural network to predict the annual slip monthly average of the solar F107 index for n months after the target month, and fills it into the solar F107 index monthly average database.

[0092] The first acquisition module is used to traverse the monthly average value database of the solar F107 index and acquire the monthly average value data of the solar F107 index for the target prediction month and the previous several months.

[0093] The first input module is used to input the target prediction month and the monthly average value of the solar F107 index of the previous several months into the solar activity level prediction model.

[0094] The first output module is used to predict the annual slippage monthly average of the solar F107 index.

[0095] Example 3:

[0096] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for predicting long-term solar activity levels based on the annual sliding monthly mean of the F107 index, as described in Embodiment 1.

[0097] Example 4:

[0098] A non-transitory computer-readable storage medium storing a computer program thereon, characterized in that, when the computer program is executed by a processor, it implements the method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index as described in Embodiment 1.

[0099] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index, characterized in that, include: S1: Establish a model for predicting solar activity levels; S2: Establish a monthly average database of the solar F107 index; S3: Obtain the target forecast month; S4: Use the RBF neural network to predict the annual slip monthly average of the solar F107 index for the n months following the target month, and fill it into the solar F107 index monthly average database; S5: Traverse the monthly average solar F107 index database to obtain the monthly average solar F107 index data for the target prediction month and the previous few months. S6: Substitute the target prediction month and the monthly average value of the solar F107 index of the previous few months into the solar activity level prediction model. S7: Outputs the annual monthly average of the solar F107 index as the prediction result; Step S1 includes: S11: Input the calculation formula for the annual monthly average value of the solar F107 index; S12: Determine the input and output quantities; S13: Generate input ports and output ports based on the number of input and output quantities; In step S11, the formula for calculating the annual slippage monthly average of the solar F107 index is as follows: ； Where i month refers to the target predicted month, F syi Let F107 be the annualized monthly average of the F107 index for month i. i Let F be the monthly average of the solar F107 index for the i-th month. i-1 The solar F107 index is the monthly average value of the i-th month counting backwards, and so on. i-11 This represents the monthly average of the solar F107 index for the 11 months preceding the i-th month.

2. The method for predicting long-term solar activity level based on F107 index annual mean monthly average value according to claim 1, characterized in that, Step S2 includes: S21: Obtain observation data from authoritative data sources; S22: Verify the observation data to ensure that the data includes timestamps and monthly average observations of the solar F107 index; S23: Establish the correspondence between the timestamps and the monthly average observations of the solar F107 index; S24: Based on the aforementioned correspondence, complete the monthly average value database of the solar F107 index; S25: Check data integrity; if missing values ​​are found, replace them.

3. The method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index as described in claim 2, characterized in that, Step S4 includes: S41: Set the prediction algorithm for the RBF network and determine the input layer dimension of the RBF neural network; S42: Select the RBF center point initialization method; S43: Determine the number of hidden layer nodes; S44: Select the radial basis function type; S45: Predict the annual monthly average value of the solar F107 index and fill the predicted value into the corresponding month of the monthly average value database of the solar F107 index; S46: After the prediction process, a loss optimization strategy is adopted to reduce the error of the predicted value and adjust the prediction algorithm.

4. The method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index as described in claim 3, characterized in that, The loss optimization strategy in step S46 includes: S461: Traverse the monthly average database of the solar F107 index, find the timestamp, and divide the training set / validation set / test set according to the time order; S462: Perform unsupervised learning; S463: Conduct supervised learning; S464: Perform parameter fine-tuning and adjust hyperparameters using cross-validation.

5. A system for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index, based on the method for predicting long-term solar activity levels based on the annual slipping monthly mean of the F107 index as described in any one of claims 1-4, characterized in that... include: The first module is used to establish a predictive model for solar activity levels. The second module is used to establish a database of monthly average values ​​of the solar F107 index. The first acquisition module is used to acquire the target predicted month; The first prediction module uses an RBF neural network to predict the annual slip monthly average of the solar F107 index for n months after the target month, and fills it into the solar F107 index monthly average database. The first acquisition module is used to traverse the monthly average value database of the solar F107 index and acquire the monthly average value data of the solar F107 index for the target prediction month and the previous several months. The first input module is used to input the target prediction month and the monthly average value of the solar F107 index of the previous several months into the solar activity level prediction model. The first output module is used to predict the annual slippage monthly average of the solar F107 index.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a method for predicting long-term solar activity levels based on the annual sliding monthly mean of the F107 index, as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, it implements a method for predicting long-term solar activity levels based on the annual sliding monthly mean of the F107 index, as described in any one of claims 1 to 4.

Images