A method and device for intelligent prediction of cloud and rain meteorological parameters

Abstract

This invention discloses an intelligent prediction method and apparatus for cloud and rain meteorological parameters. The method includes: acquiring a set of historical meteorological environmental measurement data; constructing a cloud and rain parameter fusion prediction model based on the historical measurement data set; collecting a meteorological measurement dataset for the time to be predicted; and processing the meteorological measurement dataset for the time to be predicted using the cloud and rain parameter fusion prediction model to obtain a set of predicted cloud and rain meteorological parameters. The set of predicted cloud and rain meteorological parameters includes predicted precipitation, predicted snowfall, and predicted total cloud cover.

Description

Technical Field

[0001] This invention relates to the fields of meteorological forecasting and data modeling, specifically to an intelligent forecasting method and apparatus for cloud and rain meteorological parameters. Background Technology

[0002] In the field of meteorological parameter forecasting, accurate prediction of cloud and rain meteorological parameters such as precipitation, snowfall, and total cloud cover is crucial for disaster prevention and mitigation, agricultural production planning, and air traffic control. Traditional cloud and rain meteorological parameter forecasting methods often rely on a single forecasting model, such as gray-scale prediction, support vector machines, or deep learning models. Gray-scale prediction methods have limited accuracy when dealing with nonlinear and complex meteorological data; support vector machine methods are limited by kernel function selection and parameter adjustment, making it difficult to adapt to dynamic changes in meteorological data; while deep learning models have powerful feature extraction capabilities, they suffer from problems such as large training data requirements and susceptibility to overfitting. Furthermore, existing technologies typically use a single model independently, failing to fully utilize the advantages of different models, and the processing of meteorological data is relatively fragmented, failing to effectively integrate non-cloud and rain data with cloud and rain data from historical measurements. This results in models that cannot comprehensively capture the intrinsic relationships between meteorological elements, making it difficult to meet the growing demand for high-precision, real-time meteorological forecasting. Summary of the Invention

[0003] This invention primarily addresses the problem that existing meteorological forecasting models cannot fully capture the intrinsic relationships between meteorological elements, making it difficult to meet the ever-growing demand for high-precision, real-time meteorological forecasts. This invention discloses an intelligent forecasting method and device for cloud and rain meteorological parameters.

[0004] In a first aspect, this invention discloses an intelligent prediction method for cloud and rain meteorological parameters, comprising:

[0005] S1, acquire a set of historical meteorological environmental measurement data; the set of historical measurement data includes: a non-cloud and rain data subset and a cloud and rain data subset; the cloud and rain data subset includes precipitation, snowfall, and total cloud cover;

[0006] S2, Based on the historical measurement data set, a cloud and rain parameter fusion prediction model is constructed;

[0007] S3, collect the meteorological measurement dataset for the time to be predicted; the meteorological measurement dataset for the time to be predicted includes the measurement data set of the previous time and the non-cloud and rain data subset of the current time; the previous time is a certain time before the current time;

[0008] S4. Using the cloud and rain parameter fusion prediction model, the meteorological measurement dataset at the time to be predicted is processed to obtain a set of cloud and rain meteorological parameter prediction values; the set of cloud and rain meteorological parameter prediction values ​​includes precipitation prediction values, snowfall prediction values, and total cloud cover prediction values.

[0009] The cloud and rain parameter fusion prediction model constructed based on the historical measurement data set includes:

[0010] S21, using the non-cloud and rain data subset in the historical measurement data set as the input quantity and the cloud and rain data subset in the historical measurement data set as the quantity to be predicted, a prediction model is constructed on the input quantity and the quantity to be predicted to obtain a prediction model set and a model evaluation information set; each prediction model in the prediction model set has a corresponding model evaluation value in the model evaluation information set.

[0011] S22, Based on the model evaluation information set, the prediction model set is fused to obtain the cloud and rain parameter fusion prediction model.

[0012] The process involves using a non-cloud / rainy data subset from the historical measurement data set as input and a cloud / rainy data subset from the historical measurement data set as the variable to be predicted. A prediction model is constructed on the input and the variable to be predicted, resulting in a prediction model set and a model evaluation information set, including:

[0013] S211, using the non-cloud and rain data subset in the historical measurement data set as input and the cloud and rain data subset in the historical measurement data set as the quantity to be predicted, a dataset to be modeled is constructed; the dataset to be modeled includes the input and the quantity to be predicted.

[0014] S212, The dataset to be modeled is processed using a grayscale prediction method to obtain a first prediction model;

[0015] S213, The dataset to be modeled is processed using the support vector machine method to obtain the second prediction model;

[0016] S214, The dataset to be modeled is processed using deep learning methods to obtain a third prediction model;

[0017] S215, using each prediction model, calculate and process the input quantity in the dataset to be modeled to obtain the corresponding prediction quantity, and evaluate and discriminate the prediction quantity and the quantity to be predicted to obtain the evaluation information of the prediction model.

[0018] S216. Using all the prediction models, construct a set of prediction models; using the evaluation information of all the prediction models, construct a set of model evaluation information.

[0019] The expression for the evaluation and discrimination process is:

[0020]

[0021] Where ReLU(·) is the ReLU activation function, N is the total number of inputs in the dataset to be modeled, pg1 is the evaluation information of the prediction model, and α i and β i These are the predicted values ​​corresponding to the i-th to-be predicted value and the i-th input value in the dataset to be modeled, respectively.

[0022] The expression for the fusion process is:

[0023]

[0024] Among them, pg1 i For the evaluation information of the i-th prediction model, w i f is a preset weighting factor. i f(x) is the expression of the i-th prediction model, x is the input of the prediction model, and f(x) is the cloud and rain parameter fusion prediction model.

[0025] The process of using the cloud and rain parameter fusion prediction model to process the meteorological measurement dataset for the time to be predicted, and obtaining a set of predicted cloud and rain meteorological parameters, includes:

[0026] S41, using the set of measurement data from the preceding time in the meteorological measurement dataset of the time to be predicted, the cloud and rain parameter fusion prediction model is tested to obtain a set of test values;

[0027] S42, using the cloud and rain parameter fusion prediction model and the set of test values, perform prediction processing on the non-cloud and rain data subset of the meteorological measurement dataset at the time to be predicted, and obtain the set of cloud and rain meteorological parameter prediction values.

[0028] The cloud and rain parameter fusion prediction model is tested using the set of measurement data from previous times in the meteorological measurement dataset at the time to be predicted, resulting in a set of test values, including:

[0029] S411, using the cloud and rain parameter fusion prediction model, the non-cloud and rain data subset in the previous time measurement data set of the meteorological measurement dataset at the time to be predicted is processed to obtain the cloud and rain parameter prediction values ​​for the previous time; the cloud and rain parameter prediction values ​​include the prediction values ​​of precipitation, snowfall, and total cloud cover.

[0030] S412, perform difference evaluation processing on the predicted cloud and rain parameter values ​​of the preceding time step and the collected cloud and rain parameter values ​​of the preceding time step to obtain a set of verification values; the set of verification values ​​includes verification values ​​for precipitation, snowfall, and total cloud cover; the collected cloud and rain parameter values ​​include precipitation, snowfall, and total cloud cover.

[0031] The expression for the difference assessment process is:

[0032]

[0033] Where cy is the check value, x k Let y be the k-th predicted value of a class of cloud and rain parameters. k Let K be the kth collected value of a type of cloud and rain parameter, where K is the total number of predicted values ​​of a type of cloud and rain parameter.

[0034] The types of cloud and rain parameters include precipitation, snowfall, and total cloud cover.

[0035] A second aspect of this invention discloses an intelligent forecasting device for cloud and rain meteorological parameters, the device comprising:

[0036] Memory containing executable program code;

[0037] A processor coupled to the memory;

[0038] The processor calls the executable program code stored in the memory to execute the intelligent prediction method for cloud and rain meteorological parameters.

[0039] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are invoked by a computer, they are used to execute the intelligent prediction method for cloud and rain meteorological parameters.

[0040] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the intelligent prediction method for cloud and rain meteorological parameters.

[0041] The beneficial effects of this invention are as follows:

[0042] The intelligent prediction method for cloud and rain meteorological parameters proposed in this invention significantly improves the accuracy and reliability of predictions through innovative technical solutions. First, a cloud and rain parameter fusion prediction model is constructed based on historical measurement data, combining grayscale prediction, support vector machines, and deep learning to fully leverage the advantages of different models in handling different types of meteorological data. Grayscale prediction is suitable for situations with limited data and incomplete information; support vector machines excel at handling small samples and nonlinear problems; and deep learning can uncover complex meteorological data features. The fusion of these three methods enables multi-dimensional and multi-level analysis of meteorological data, effectively compensating for the shortcomings of single models. Second, during model construction, specific evaluation and fusion processing expressions dynamically adjust weights based on the predictive performance of each model, making the fusion model more scientific and reasonable, and improving its adaptability to complex meteorological changes. In the prediction stage, the fusion model is tested using measurement data from previous time points, and the test values ​​are used to predict the current data, further optimizing the prediction process and ensuring that the model maintains high prediction accuracy in practical applications, providing more reliable technical support for meteorological decision-making. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention. Detailed Implementation

[0044] To better understand the content of this invention, an embodiment is provided here.

[0045] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention.

[0046] In a first aspect, this invention discloses an intelligent prediction method for cloud and rain meteorological parameters, comprising:

[0047] S1, acquire the historical measurement data set of meteorological environment;

[0048] S2, Based on the historical measurement data set, a cloud and rain parameter fusion prediction model is constructed;

[0049] S3, collect the meteorological measurement dataset for the time to be predicted;

[0050] S4. Using the cloud and rain parameter fusion prediction model, the meteorological measurement dataset at the time to be predicted is processed to obtain a set of cloud and rain meteorological parameter prediction values; the set of cloud and rain meteorological parameter prediction values ​​includes precipitation prediction values, snowfall prediction values, and total cloud cover prediction values.

[0051] The historical measurement data set includes: a non-cloud and rain data subset and a cloud and rain data subset; the non-cloud and rain data subset includes parameter information of typical pressure layers, 10-meter wind field, 2-meter dew point temperature, 2-meter temperature, sea level pressure, surface pressure, total precipitable water content, and total water vapor content; the cloud and rain data subset includes precipitation, snowfall, and total cloud cover.

[0052] The parameters of the typical pressure layer include: temperature, specific humidity, relative humidity, east-west wind speed, north-south wind speed, vertical velocity, and geopotential height.

[0053] The typical pressure layers include altitude layers with pressures of 200hPa, 300hPa, 400hPa, 500hPa, 600hPa, 700hPa, 850hPa, 925hPa, and 1000hPa.

[0054] The meteorological measurement dataset for the time to be predicted includes a set of measurement data from previous times and a subset of non-cloud and rain data for the current time; the previous time is a certain time before the current time; the measurement dataset includes parameter information of a typical pressure layer, 10-meter wind field, 2-meter dew point temperature, 2-meter temperature, sea level pressure, surface pressure, total precipitable water content, total water vapor content, precipitation, snowfall, and total cloud cover;

[0055] The aforementioned "forward for a certain period of time" can be 1 hour or 2 hours.

[0056] The cloud and rain parameter fusion prediction model constructed based on the historical measurement data set includes:

[0057] S21, using the non-cloud and rain data subset in the historical measurement data set as the input quantity and the cloud and rain data subset in the historical measurement data set as the quantity to be predicted, a prediction model is constructed on the input quantity and the quantity to be predicted to obtain a prediction model set and a model evaluation information set; each prediction model in the prediction model set has a corresponding model evaluation value in the model evaluation information set.

[0058] S22, Based on the model evaluation information set, the prediction model set is fused to obtain the cloud and rain parameter fusion prediction model.

[0059] The process involves using a non-cloud / rainy data subset from the historical measurement data set as input and a cloud / rainy data subset from the historical measurement data set as the variable to be predicted. A prediction model is constructed on the input and the variable to be predicted, resulting in a prediction model set and a model evaluation information set, including:

[0060] S211, using the non-cloud and rain data subset in the historical measurement data set as input and the cloud and rain data subset in the historical measurement data set as the quantity to be predicted, a dataset to be modeled is constructed; the dataset to be modeled includes input and quantity to be predicted; the quantity to be predicted is also the output quantity corresponding to the input.

[0061] S212, The dataset to be modeled is processed using a grayscale prediction method to obtain a first prediction model;

[0062] S213, The dataset to be modeled is processed using the support vector machine method to obtain the second prediction model;

[0063] S214, The dataset to be modeled is processed using deep learning methods to obtain a third prediction model;

[0064] S215, using each prediction model, calculate and process the input quantity in the dataset to be modeled to obtain the corresponding prediction quantity, and evaluate and discriminate the prediction quantity and the quantity to be predicted to obtain the evaluation information of the prediction model.

[0065] S216. Using all the prediction models, construct a set of prediction models; using the evaluation information of all the prediction models, construct a set of model evaluation information.

[0066] The processing of the dataset to be modeled involves using the input quantity and the quantity to be predicted as input data for the prediction method, and constructing the corresponding prediction model using gray-scale prediction method, support vector machine method, or deep learning method, respectively.

[0067] The expression for the evaluation and discrimination process is:

[0068]

[0069] Where ReLU(·) is the ReLU activation function, N is the total number of inputs in the dataset to be modeled, pg1 is the evaluation information of the prediction model, and α i and β i These are the predicted values ​​corresponding to the i-th predictable value and the i-th input value in the dataset to be modeled, respectively.

[0070] The evaluation and discrimination expression employs the ReLU activation function, which effectively handles nonlinear problems and performs a nonlinear transformation on the deviation between predicted and actual values, enhancing the ability to capture the characteristics of complex meteorological data. Simultaneously, by combining trigonometric and exponential functions, the difference in prediction error is further amplified, allowing the model evaluation information pg1 to more sensitively reflect the performance of the prediction model, thereby accurately selecting the superior prediction model. In the denominator... and The combination of these factors, when the prediction error is large, suppresses outliers by utilizing the periodicity of trigonometric functions and the decaying characteristics of exponential functions. This avoids excessive influence of individual outliers on the model evaluation results, thereby improving the stability and reliability of the evaluation.

[0071] The expression for the fusion process is:

[0072]

[0073] Among them, pg1 i For the evaluation information of the i-th prediction model, w i f is a preset weighting factor. i f(x) is the expression of the i-th prediction model, x is the input of the prediction model, and f(x) is the cloud and rain parameter fusion prediction model.

[0074] In the expression for the fusion process, w i exp(-pg1 i This system dynamically allocates weights based on the evaluation information of each prediction model. Models with better evaluation information receive higher weights and contribute more to the fusion model, enabling the fusion model to adaptively integrate the strengths of different models and achieve optimal prediction results. By fusing the three prediction models through weighted summation, the system fully leverages the advantages of grayscale prediction, support vector machines, and deep learning models in different scenarios, avoiding the limitations of a single model. Different models analyze meteorological data from different perspectives; after fusion, the system can more comprehensively mine data features, improving prediction accuracy and generalization ability.

[0075] When processing the dataset to be modeled using support vector machine, grayscale prediction, and deep learning methods, the input quantities in the dataset to be modeled are used as model inputs, and the quantities to be predicted are used as the expected outputs for modeling.

[0076] The process of using deep learning to process the dataset to be modeled to obtain a third prediction model can be achieved by using the dataset to be modeled as the training dataset to train a preset LightGBM model to obtain the third prediction model.

[0077] The process of using the cloud and rain parameter fusion prediction model to process the meteorological measurement dataset for the time to be predicted, and obtaining a set of predicted cloud and rain meteorological parameters, includes:

[0078] S41, using the set of measurement data from the preceding time in the meteorological measurement dataset of the time to be predicted, the cloud and rain parameter fusion prediction model is tested to obtain a set of test values;

[0079] S42, using the cloud and rain parameter fusion prediction model and the set of test values, perform prediction processing on the non-cloud and rain data subset of the meteorological measurement dataset at the time to be predicted, and obtain the set of cloud and rain meteorological parameter prediction values.

[0080] The cloud and rain parameter fusion prediction model is tested using the set of measurement data from previous times in the meteorological measurement dataset at the time to be predicted, resulting in a set of test values, including:

[0081] S411, using the cloud and rain parameter fusion prediction model, the non-cloud and rain data subset in the previous time measurement data set of the meteorological measurement dataset at the time to be predicted is processed to obtain the cloud and rain parameter prediction values ​​for the previous time; the cloud and rain parameter prediction values ​​include the prediction values ​​of precipitation, snowfall, and total cloud cover.

[0082] S412, perform difference evaluation processing on the predicted cloud and rain parameter values ​​at the preceding time and the collected cloud and rain parameter values ​​at the preceding time to obtain a set of verification values; the set of verification values ​​includes verification values ​​for precipitation, snowfall, and total cloud cover; the collected cloud and rain parameter values ​​include precipitation, snowfall, and total cloud cover.

[0083] The expression for the difference assessment process is:

[0084]

[0085] Where cy is the check value, x k Let y be the k-th predicted value of a class of cloud and rain parameters. k Let K be the kth collected value of a type of cloud and rain parameter, where K is the total number of predicted values ​​of a type of cloud and rain parameter.

[0086] The expression for the difference assessment process utilizes the tangent and exponential functions to measure the predicted and collected values ​​from two dimensions: proportional difference and absolute difference. This provides a more comprehensive reflection of the differences between predicted and actual cloud and rainfall parameters, offering a more accurate basis for model validation. Introducing complex numbers not only represents the magnitude of the error but also reflects its phase information, further enriching the meaning of the validation value. This facilitates a more detailed analysis of the deviations in the model's prediction results, providing a more valuable reference for subsequent model optimization and prediction adjustments.

[0087] The types of cloud and rain parameters include precipitation, snowfall, and total cloud cover.

[0088] The method utilizes the cloud and rain parameter fusion prediction model and the set of test values ​​to perform prediction processing on the non-cloud and rain data subset of the meteorological measurement dataset at the time to be predicted, thereby obtaining a set of predicted cloud and rain meteorological parameters, including:

[0089] Using the cloud and rain parameter fusion prediction model, the non-cloud and rain data subset of the meteorological measurement dataset at the time to be predicted is calculated to obtain an initial prediction value set; the initial prediction value set includes the initial prediction values ​​of precipitation, snowfall, and total cloud cover.

[0090] By adding the corresponding elements of the test value set to the initial predicted value set, we obtain the predicted value set of cloud and rain meteorological parameters.

[0091] The non-cloud and rain data subset in the measurement data set includes parameter information of typical pressure layers, 10-meter wind field, 2-meter dew point temperature, 2-meter temperature, sea level pressure, surface pressure, total precipitable water content, and total water vapor content.

[0092] The real-time acquisition of the meteorological measurement dataset can be achieved using micro-rain radar, raindrop spectrometer, fog droplet spectrometer, visibility meter, millimeter-wave radar, dual-polarization Doppler radar, wind profiler radar, microwave radiometer, active lightning positioning imaging system, and ground electric field meter.

[0093] The historical measurement data set of the meteorological environment can be obtained from an open-source meteorological environment database.

[0094] The neural network model used in the deep learning method includes a first input module, a first convolution module, a depthwise separable convolution module, a first up-dimensional convolution module, a second up-dimensional convolution module, a third up-dimensional convolution module, a fourth up-dimensional convolution module, a second convolution module, a first pooling module, a third convolution module, and a first fully connected module.

[0095] The input terminal of the first input module of the neural network model is used to receive a subset of non-cloudy / rainy data. The output terminal of the first input module is connected to the input terminal of the first convolutional module of the neural network model. The output terminal of the first convolutional module is connected to the input terminal of the depthwise separable convolutional module of the neural network model. The output terminal of the depthwise separable convolutional module is connected to the input terminal of the first up-dimensional convolutional module of the neural network model. The output terminal of the first up-dimensional convolutional module is connected to the input terminal of the second up-dimensional convolutional module of the neural network model. The output terminal of the second up-dimensional convolutional module is connected to the input terminal of the third up-dimensional convolutional module of the neural network model. The input ends of the blocks are connected; the output end of the third dimensionality-upgrading convolution module of the neural network model is connected to the input end of the fourth dimensionality-upgrading convolution module of the neural network model; the output end of the fourth dimensionality-upgrading convolution module of the neural network model is connected to the input end of the second convolution module of the neural network model; the output end of the second convolution module of the neural network model is connected to the input end of the first pooling module of the neural network model; the output end of the first pooling module of the neural network model is connected to the input end of the third convolution module of the neural network model; the output end of the third convolution module of the neural network model is connected to the input end of the first fully connected module of the neural network model; the first fully connected module is used to output the prediction results of cloud and rain meteorological parameters.

[0096] A second aspect of this invention discloses an intelligent forecasting device for cloud and rain meteorological parameters, the device comprising:

[0097] Memory containing executable program code;

[0098] A processor coupled to the memory;

[0099] The processor calls the executable program code stored in the memory to execute the intelligent prediction method for cloud and rain meteorological parameters.

[0100] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are invoked by a computer, they are used to execute the intelligent prediction method for cloud and rain meteorological parameters.

[0101] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the intelligent prediction method for cloud and rain meteorological parameters.

[0102] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for intelligent prediction of cloud and rain meteorological parameters, characterized in that, include: S1, acquire the historical measurement data set of meteorological environment; The historical measurement data set includes: a non-cloud and rain data subset and a cloud and rain data subset; the cloud and rain data subset includes precipitation, snowfall, and total cloud cover; S2, Based on the historical measurement data set, a cloud and rain parameter fusion prediction model is constructed, including: S21, using the non-cloud / rain data subset from the historical measurement data set as input and the cloud / rain data subset from the historical measurement data set as the variable to be predicted, a prediction model is constructed for the input and the variable to be predicted, resulting in a prediction model set and a model evaluation information set, including: Each prediction model in the prediction model set has a corresponding model evaluation value in the model evaluation information set; S211, using the non-cloud and rain data subset in the historical measurement data set as input and the cloud and rain data subset in the historical measurement data set as the quantity to be predicted, a dataset to be modeled is constructed; the dataset to be modeled includes the input and the quantity to be predicted. S212, The dataset to be modeled is processed using a grayscale prediction method to obtain a first prediction model; S213, The dataset to be modeled is processed using the support vector machine method to obtain the second prediction model; S214, The dataset to be modeled is processed using deep learning methods to obtain a third prediction model; S215, using each prediction model, calculate and process the input quantity in the dataset to be modeled to obtain the corresponding prediction quantity, and evaluate and discriminate the prediction quantity and the quantity to be predicted to obtain the evaluation information of the prediction model. S216, Using all the prediction models, construct a set of prediction models; Using the evaluation information of all the prediction models, construct a set of model evaluation information. S22, Based on the model evaluation information set, the prediction model set is fused to obtain the cloud and rain parameter fusion prediction model; S3, collect the meteorological measurement dataset for the time to be predicted; the meteorological measurement dataset for the time to be predicted includes the measurement data set of the previous time and the non-cloud and rain data subset of the current time; the previous time is a certain time before the current time; S4. Using the cloud and rain parameter fusion prediction model, the meteorological measurement dataset at the time to be predicted is processed to obtain a set of cloud and rain meteorological parameter prediction values; the set of cloud and rain meteorological parameter prediction values ​​includes precipitation prediction values, snowfall prediction values, and total cloud cover prediction values. The expression for the evaluation and discrimination process is: ， in, The ReLU activation function is used, and N is the total number of inputs in the dataset to be modeled. For evaluation information of the predictive model, and These are the predicted values ​​corresponding to the i-th predictable value and the i-th input value in the dataset to be modeled, respectively.

2. The intelligent prediction method for cloud and rain meteorological parameters as described in claim 1, characterized in that, The expression for the fusion process is: ， in, This provides evaluation information for the i-th prediction model. The preset weighting factor, Let x be the expression for the i-th prediction model, and let x be the input to the prediction model. This is a cloud and rain parameter fusion prediction model.

3. The intelligent prediction method for cloud and rain meteorological parameters as described in claim 1, characterized in that, The process of using the cloud and rain parameter fusion prediction model to process the meteorological measurement dataset for the time to be predicted, and obtaining a set of predicted cloud and rain meteorological parameters, includes: S41, using the set of measurement data from the preceding time in the meteorological measurement dataset of the time to be predicted, the cloud and rain parameter fusion prediction model is tested to obtain a set of test values; S42, using the cloud and rain parameter fusion prediction model and the set of test values, perform prediction processing on the non-cloud and rain data subset of the meteorological measurement dataset at the time to be predicted, and obtain the set of cloud and rain meteorological parameter prediction values.

4. The intelligent prediction method for cloud and rain meteorological parameters as described in claim 3, characterized in that, The cloud and rain parameter fusion prediction model is tested using the set of measurement data from previous times in the meteorological measurement dataset at the time to be predicted, resulting in a set of test values, including: S411, using the cloud and rain parameter fusion prediction model, the non-cloud and rain data subset in the previous time measurement data set of the meteorological measurement dataset at the time to be predicted is processed to obtain the cloud and rain parameter prediction values ​​for the previous time; the cloud and rain parameter prediction values ​​include the prediction values ​​of precipitation, snowfall, and total cloud cover. S412, perform difference evaluation processing on the predicted cloud and rain parameter values ​​of the preceding time step and the collected cloud and rain parameter values ​​of the preceding time step to obtain a set of verification values; the set of verification values ​​includes verification values ​​for precipitation, snowfall, and total cloud cover; the collected cloud and rain parameter values ​​include precipitation, snowfall, and total cloud cover.

5. The intelligent prediction method for cloud and rain meteorological parameters as described in claim 4, characterized in that, The expression for the difference assessment process is: ， Where cy is the check value. This is the k-th predicted value of a class of cloud and rain parameters. Let K be the kth collected value of a type of cloud and rain parameter, where K is the total number of predicted values ​​of a type of cloud and rain parameter. The types of cloud and rain parameters include precipitation, snowfall, and total cloud cover.

6. An intelligent forecasting device for cloud and rain meteorological parameters, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the intelligent prediction method for cloud and rain meteorological parameters as described in any one of claims 1 to 5.

7. A computer-storable medium, characterized in that, The computer storage medium stores computer instructions, which, when invoked by the computer, are used to execute the intelligent prediction method for cloud and rain meteorological parameters as described in any one of claims 1 to 5.

8. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the intelligent prediction method for cloud and rain meteorological parameters as described in any one of claims 1 to 5.

Images