Data processing and analyzing method for multi-source numerical forecasting data based on cloud edge fusion

Abstract

The invention discloses a data processing analysis method for multi-source numerical forecasting data based on cloud edge fusion, and relates to the technical field of data analysis, comprising the following steps: establishing a meteorological multi-source numerical forecasting database; a decoding object of each piece of meteorological general data is created, and a meteorological general data decoding pool is established; fitting matching is carried out based on the meteorological general data decoding pool and the meteorological multi-source numerical forecasting database, and a meteorological multi-source numerical forecasting unified data set is obtained; performing data annotation according to unit time based on the meteorological multi-source numerical prediction unified data set, and constructing a meteorological multi-source numerical prediction unified time sequence data set; carrying out fusion integration on the meteorological multi-source numerical prediction unified time series data set according to a meteorological multi-source numerical prediction mode, and predicting the meteorological rainfall probability of the target area; and on the basis of predicting the meteorological rainfall probability of the target area, generating a meteorological rainfall probability forecast map of the target area. The method has the advantage that the forecasting precision and accuracy of the meteorological data are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion. Background Art

[0002] Existing weather forecasts mainly rely on regression analysis of influencing factors observed within a short time scale and macroscopic estimation of meteorological turbulent fluids using the Navier-Stokes equations to predict future meteorological data. However, due to the unobservability of turbulence in the microscopic state, the actual prediction of future meteorological data is relatively inaccurate. Summary of the invention

[0003] In order to solve the above technical problems, a data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion is provided. This technical solution solves the problem that the existing weather forecast mainly relies on regression analysis of influencing factors observed within a short time scale and macroscopic estimation of meteorological turbulent fluid using the Navier-Stokes equations to predict future meteorological data, but due to the unobservability of turbulence in the microscopic state, the actual prediction of future meteorological data has poor accuracy.

[0004] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0005] The data processing and analysis methods of multi-source numerical forecast data based on cloud-edge fusion include:

[0006] Obtain general meteorological data from several weather forecast systems and establish a meteorological multi-source numerical forecast database;

[0007] Create decoding objects for various meteorological general data and form a meteorological general data decoding pool;

[0008] Based on the meteorological general data decoding pool and the meteorological multi-source numerical forecast database, a unified meteorological multi-source numerical forecast data set is obtained;

[0009] Based on the unified data set of meteorological multi-source numerical forecast, data is annotated according to unit time to construct a unified time series data set of meteorological multi-source numerical forecast;

[0010] The unified time series data set for meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area;

[0011] Based on the predicted meteorological precipitation probability of the target area, a meteorological precipitation probability forecast map of the target area is generated.

[0012] Preferably, fitting and matching the meteorological general data decoding pool with the meteorological multi-source numerical forecast database to obtain a meteorological multi-source numerical forecast unified data set specifically includes:

[0013] Based on the meteorological general data decoding pool, a meteorological general data decoding public class is established with each meteorological general data decoding method;

[0014] The meteorological general data decoding public class is used as the outer loop, and the meteorological multi-source numerical forecast data is used as the inner loop nesting. According to the matching formula, the matching index between each meteorological multi-source numerical forecast data and each meteorological general data decoding method is calculated as the decoding loop stop condition, and the data decoding demand function is established;

[0015] According to the data decoding requirement function, the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method is obtained, a data decoding model is constructed, and a unified meteorological multi-source numerical forecast data set is generated;

[0016] The matching formula is specifically:

[0017]

[0018] In the formula, MatchScore(D i ,C j ) is the i-th meteorological multi-source numerical forecast data, C j is the jth meteorological general data decoding method, is the kth dimension parameter value in the i-th meteorological multi-source numerical forecast data, is the kth dimension parameter value in the jth meteorological general data decoding method, W k is the weight of the parameter value of the kth dimension.

[0019] Preferably, according to the data decoding requirement function, obtaining the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method, constructing a data decoding model, and generating a unified meteorological multi-source numerical forecast data set specifically includes:

[0020] Based on the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method as the root node, and each meteorological general data decoding method as the decoding branch node, a meteorological multi-source numerical forecast data decoding demand decision tree is established, which is connected in series to form a data decoding model;

[0021] Based on the data decoding model, the meteorological multi-source numerical forecast database is used as input, and the meteorological multi-source numerical forecast data is divided according to the maximum gain ratio of the decoding branch nodes to generate a unified meteorological multi-source numerical forecast data set;

[0022] The maximum gain ratio of the meteorological multi-source numerical forecast data for decoding branch nodes is divided into:

[0023]

[0024] In the formula, GR(D,A) is the maximum gain ratio of meteorological multi-source numerical forecast data D to decoding branch node A, IG(D,A) is the information gain of meteorological multi-source numerical forecast data D to decoding branch node A, H(A) is the attribute entropy of meteorological multi-source numerical forecast data D to decoding branch node A, D j The branch nodes for matching the meteorological multi-source numerical forecast data with the jth meteorological general data decoding method are divided into subsets, D is the total set of meteorological multi-source numerical forecast data, log2 is the logarithmic function, and C(A) is the total number of decoding branch nodes A.

[0025] Preferably, the unified time series data set of meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area;

[0026] Based on the unified time series data of meteorological multi-source numerical prediction, the correlation factors are screened according to the meteorological precipitation type of the target area to be predicted, and the time series data of meteorological multi-source numerical prediction influence is obtained;

[0027] Based on the meteorological multi-source numerical prediction affecting time series data, data alignment is performed according to spatial resolution and time resolution, the time resolution is used as the observation window, the meteorological multi-source numerical prediction affecting time series data within the spatial resolution is used as the observation attribute, and the meteorological characteristic time series parameters of the target area to be predicted are obtained through a sliding window; the spatial resolution is the target distance, and the time resolution is hours;

[0028] According to the meteorological characteristic time series parameters of the target area to be predicted, the optimal meteorological precipitation type characteristic time series parameters are screened and integrated for the posterior probability of the unified time series data of the meteorological multi-source numerical prediction of the target area, and a multi-mode meteorological characteristic time series parameter set of the target area to be predicted is obtained;

[0029] According to the multi-mode meteorological characteristic time series parameter set of the target area to be predicted, a meteorological prediction model of the target area is constructed to predict the meteorological precipitation probability of the target area;

[0030] The posterior probability of the meteorological characteristic time series parameters of the target area to be predicted for the unified time series data of the meteorological multi-source numerical prediction of the target area is specifically:

[0031]

[0032] In the formula, P(u,t|x v) is the posterior probability of the characteristic parameter of the vth meteorological precipitation type given the tth unit time of the uth target area to be predicted, P(x v |u,t) is the conditional probability of the u-th target area at the t-th unit time given the v-th meteorological precipitation type characteristic parameter, P(u,t) is the prior probability of the u-th target area to be predicted at the t-th unit time, P(x v ) is the overall probability distribution of characteristic parameters of the vth meteorological precipitation type;

[0033] The meteorological forecast model for the target area is specifically:

[0034]

[0035] In the formula, G U,t is the probability of meteorological precipitation in the tth unit time of the uth target area to be predicted, f() is the indicator function, if x v ≥T, then f(x v ≥T)=1, indicating that precipitation occurred in the area at that time point. Then f(x v ≥T)=0, indicating that there is no precipitation in the area at that time point, T is the precipitation threshold, and m is the set of multi-mode meteorological characteristic time series parameters of the target area to be predicted.

[0036] Preferably, based on the unified time series data of meteorological multi-source numerical prediction, the associated factors are screened according to the meteorological precipitation type of the target area to be predicted, and the time series data affecting the meteorological multi-source numerical prediction is obtained, which specifically includes:

[0037] Determine the meteorological precipitation type attention preference influencing factors of the target area to be detected, and establish a meteorological precipitation type preference array of the target area to be detected;

[0038] By using linear algebra, a linear mapping is performed on the meteorological precipitation type preference array of the target area to be detected to obtain the meteorological precipitation type preference vector matrix Ns of the target area to be detected;

[0039]

[0040] Among them, R uv Match the feature vector of the vth meteorological precipitation type of the uth target area to be predicted, z is the total number of target areas to be predicted, and l is the total number of meteorological precipitation types;

[0041] For the unified time series data of meteorological multi-source numerical prediction, the principal component analysis method is used to reduce the data dimension, and the unified time series reduced dimension data of meteorological multi-source numerical prediction is obtained;

[0042] Based on the similarity between the unified time series dimension reduction data of meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected, positive correlation factors are screened to obtain the time series data affecting meteorological multi-source numerical prediction;

[0043] Among them, the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected is specifically:

[0044]

[0045] In the formula, θ(D′ u ,R uv ) is the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction of the u-th target area to be predicted and the v-th meteorological precipitation type preference vector of the u-th target area to be detected.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention proposes a data processing and analysis scheme for multi-source numerical forecast data based on cloud-edge fusion. By integrating the general meteorological data of multiple weather forecast systems, a database is established and a decoding pool is created to unify the decoding format, and then a unified data set is formed by fitting and matching, and a time series data set is constructed according to time series annotation. The meteorological characteristics of the target area are predicted by integrating the multi-source numerical forecast model, and an intuitive precipitation probability forecast map is generated. The beneficial effect is: effectively improving the forecast precision and accuracy of meteorological data. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A flow chart of the method for obtaining a unified data set for multi-source numerical weather forecasting;

[0049] Figure 2 A flow chart of the unified data collection method for multi-source numerical weather forecasting;

[0050] Figure 3 A flow chart of the method for constructing a data decoding model;

[0051] Figure 4 A flow chart of the method for predicting meteorological precipitation probability in the target area;

[0052] Figure 5 Flow chart of the method for obtaining the time series data of multi-source numerical prediction of meteorology;

[0053] Figure 6 It is a schematic diagram of the structure of the electronic device of the present invention;

[0054] Figure 7 It is a schematic diagram of the computer-readable storage medium structure of the present invention. DETAILED DESCRIPTION

[0055] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0056] Reference Figure 1 As shown in the figure, the data processing and analysis method of multi-source numerical forecast data based on cloud-edge fusion includes:

[0057] Obtain general meteorological data from several weather forecast systems and establish a meteorological multi-source numerical forecast database;

[0058] Create decoding objects for various meteorological general data and form a meteorological general data decoding pool;

[0059] Based on the meteorological general data decoding pool and the meteorological multi-source numerical forecast database, a unified meteorological multi-source numerical forecast data set is obtained;

[0060] Based on the unified data set of meteorological multi-source numerical forecast, data is annotated according to unit time to construct a unified time series data set of meteorological multi-source numerical forecast;

[0061] The unified time series data set for meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area;

[0062] Based on the predicted meteorological precipitation probability of the target area, a meteorological precipitation probability forecast map of the target area is generated.

[0063] This solution integrates the meteorological general data of multiple weather forecast systems, establishes a database and creates a decoding pool to unify the decoding format, and then fits and matches to form a unified data set, and constructs a time series data set according to the time series annotation. By integrating the multi-source numerical forecast model, the meteorological characteristics of the target area are predicted, and an intuitive precipitation probability forecast map is generated. The beneficial effect is: effectively improving the forecast precision and accuracy of meteorological data.

[0064] Reference Figure 2 As shown, based on the meteorological general data decoding pool and the meteorological multi-source numerical forecast database, the meteorological multi-source numerical forecast unified data set is obtained, which specifically includes:

[0065] Based on the meteorological general data decoding pool, a meteorological general data decoding public class is established with each meteorological general data decoding method;

[0066] The meteorological general data decoding public class is used as the outer loop, and the meteorological multi-source numerical forecast data is used as the inner loop nesting. According to the matching formula, the matching index between each meteorological multi-source numerical forecast data and each meteorological general data decoding method is calculated as the decoding loop stop condition, and the data decoding demand function is established;

[0067] According to the data decoding requirement function, the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method is obtained, a data decoding model is constructed, and a unified meteorological multi-source numerical forecast data set is generated;

[0068] The matching formula is specifically:

[0069]

[0070] In the formula, MatchScore(D i ,C j ) is the i-th meteorological multi-source numerical forecast data, C j is the jth meteorological general data decoding method, is the kth dimension parameter value in the i-th meteorological multi-source numerical forecast data, is the kth dimension parameter value in the jth meteorological general data decoding method, W k is the weight of the parameter value of the kth dimension.

[0071] Reference Figure 3 As shown, according to the data decoding requirement function, the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method is obtained, a data decoding model is constructed, and a unified meteorological multi-source numerical forecast data set is generated, which specifically includes:

[0072] Based on the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method as the root node, and each meteorological general data decoding method as the decoding branch node, a meteorological multi-source numerical forecast data decoding demand decision tree is established, which is connected in series to form a data decoding model;

[0073] Based on the data decoding model, the meteorological multi-source numerical forecast database is used as input, and the meteorological multi-source numerical forecast data is divided according to the maximum gain ratio of the decoding branch nodes to generate a unified meteorological multi-source numerical forecast data set;

[0074] The maximum gain ratio of the meteorological multi-source numerical forecast data for decoding branch nodes is divided into:

[0075]

[0076] In the formula, GR(D,A) is the maximum gain ratio of meteorological multi-source numerical forecast data D to decoding branch node A, IG(D,A) is the information gain of meteorological multi-source numerical forecast data D to decoding branch node A, H(A) is the attribute entropy of meteorological multi-source numerical forecast data D to decoding branch node A, D j The branch nodes for matching the meteorological multi-source numerical forecast data with the jth meteorological general data decoding method are divided into subsets, D is the total set of meteorological multi-source numerical forecast data, log2 is the logarithmic function, and C(A) is the total number of decoding branch nodes A.

[0077] It is understandable that since there are many types of encoding formats in meteorological multi-source numerical forecast data, the conventional matching method will have too many categories for decoding selection, and it is impossible to make decisions based on the data itself. Therefore, the matching formula is embedded in the self-iterative function to determine the optimal decoding method, and a decision tree for decoding requirements of meteorological multi-source numerical forecast data is constructed, which are connected in series as a data decoding model to achieve fast and unified decoding of data.

[0078] Reference Figure 4 As shown, the unified time series data set of meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area;

[0079] Based on the unified time series data of meteorological multi-source numerical prediction, the correlation factors are screened according to the meteorological precipitation type of the target area to be predicted, and the time series data of meteorological multi-source numerical prediction influence is obtained;

[0080] Based on the meteorological multi-source numerical prediction affecting time series data, data alignment is performed according to spatial resolution and time resolution, the time resolution is used as the observation window, the meteorological multi-source numerical prediction affecting time series data within the spatial resolution is used as the observation attribute, and the meteorological characteristic time series parameters of the target area to be predicted are obtained through a sliding window; the spatial resolution is the target distance, and the time resolution is hours;

[0081] According to the meteorological characteristic time series parameters of the target area to be predicted, the optimal meteorological precipitation type characteristic time series parameters are screened and integrated for the posterior probability of the unified time series data of the meteorological multi-source numerical prediction of the target area, and a multi-mode meteorological characteristic time series parameter set of the target area to be predicted is obtained;

[0082] According to the multi-mode meteorological characteristic time series parameter set of the target area to be predicted, a meteorological prediction model of the target area is constructed to predict the meteorological precipitation probability of the target area;

[0083] The posterior probability of the meteorological characteristic time series parameters of the target area to be predicted for the unified time series data of the meteorological multi-source numerical prediction of the target area is specifically:

[0084]

[0085] In the formula, P(u,t|x v ) is the posterior probability of the characteristic parameter of the vth meteorological precipitation type given the tth unit time of the uth target area to be predicted, P(x v |u,t) is the conditional probability of the u-th target area at the t-th unit time given the v-th meteorological precipitation type characteristic parameter, P(u,t) is the prior probability of the u-th target area to be predicted at the t-th unit time, P(x v ) is the overall probability distribution of characteristic parameters of the vth meteorological precipitation type;

[0086] The meteorological forecast model for the target area is specifically:

[0087]

[0088] In the formula, G U,t is the probability of meteorological precipitation in the tth unit time of the uth target area to be predicted, f() is the indicator function, if x v ≥T, then f(x v ≥T)=1, indicating that precipitation occurred in the area at that time point. Then f(x v ≥T)=0, indicating that there is no precipitation in the area at that time point, T is the precipitation threshold, and m is the set of multi-mode meteorological characteristic time series parameters of the target area to be predicted.

[0089] This scheme predicts the probability of meteorological precipitation in the target area by fusing multi-source numerical forecast data. By using the unified time series data based on meteorological multi-source numerical forecast, the related influencing factors are screened and data alignment is performed, and the meteorological characteristic time series parameters are extracted based on spatial and temporal resolution. By screening and fusing the optimal meteorological characteristic time series parameters, a multi-mode meteorological characteristic time series parameter set is constructed, and then the meteorological forecast model is trained to predict the meteorological precipitation probability in the target area. The beneficial effect is that it can accurately integrate multi-source information and improve the accuracy of meteorological precipitation probability prediction through data processing and feature screening.

[0090] Reference Figure 5 The above-mentioned method is based on the unified time series data of meteorological multi-source numerical prediction, and according to the meteorological precipitation type of the target area to be predicted, the associated factors are screened to obtain the meteorological multi-source numerical prediction influencing time series data, which specifically includes:

[0091] Determine the meteorological precipitation type attention preference influencing factors of the target area to be detected, and establish a meteorological precipitation type preference array of the target area to be detected;

[0092] By using linear algebra, a linear mapping is performed on the meteorological precipitation type preference array of the target area to be detected to obtain the meteorological precipitation type preference vector matrix Ns of the target area to be detected;

[0093]

[0094] Among them, R uv Match the feature vector of the vth meteorological precipitation type of the uth target area to be predicted, z is the total number of target areas to be predicted, and l is the total number of meteorological precipitation types;

[0095] For the unified time series data of meteorological multi-source numerical prediction, the principal component analysis method is used to reduce the data dimension, and the unified time series reduced dimension data of meteorological multi-source numerical prediction is obtained;

[0096] Based on the similarity between the unified time series dimension reduction data of meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected, positive correlation factors are screened to obtain the time series data affecting meteorological multi-source numerical prediction;

[0097] Among them, the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected is specifically:

[0098]

[0099] In the formula, θ(D′ u ,R uv ) is the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction of the u-th target area to be predicted and the v-th meteorological precipitation type preference vector of the u-th target area to be detected.

[0100] Furthermore, the method or system according to the embodiment of the present application can also be Figure 6 The electronic device architecture shown in FIG. Figure 6As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, and the like. The storage device in the electronic device 500, such as ROM503 or hard disk 507, can store the data processing and analysis method of multi-source numerical forecast data based on cloud-edge fusion provided by the present application, including: obtaining general meteorological data of several weather forecast systems and establishing a meteorological multi-source numerical forecast database; creating decoding objects for each general meteorological data and forming a general meteorological data decoding pool; fitting and matching the general meteorological data decoding pool with the meteorological multi-source numerical forecast database to obtain a unified meteorological multi-source numerical forecast data set; based on the unified meteorological multi-source numerical forecast data set, data is annotated according to unit time to construct a unified meteorological multi-source numerical prediction time series data set; the unified meteorological multi-source numerical prediction time series data set is integrated and integrated according to the meteorological multi-source numerical forecast mode to predict the meteorological precipitation probability of the target area; based on the predicted meteorological precipitation probability of the target area, a meteorological precipitation probability forecast map of the target area is generated. The electronic device 500 may also include a user interface 508. Of course, Figure 6 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 6 One or more components of an electronic device are shown.

[0101] Figure 7 Schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 7 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, the data processing and analysis method of multi-source numerical forecast data based on cloud-edge fusion according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0102] When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are performed.

[0103] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion, characterized in that: include: Obtain general meteorological data from several weather forecast systems and establish a meteorological multi-source numerical forecast database; Create decoding objects for various meteorological general data and form a meteorological general data decoding pool; Based on the meteorological general data decoding pool and the meteorological multi-source numerical forecast database, a unified meteorological multi-source numerical forecast data set is obtained; Based on the unified data set of meteorological multi-source numerical forecast, data is annotated according to unit time to construct a unified time series data set of meteorological multi-source numerical forecast; The unified time series data set for meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area; Based on the predicted meteorological precipitation probability of the target area, a meteorological precipitation probability forecast map of the target area is generated.

2. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 1 is characterized in that: Based on the fitting and matching of the meteorological general data decoding pool and the meteorological multi-source numerical forecast database, the unified meteorological multi-source numerical forecast data set is obtained, which specifically includes: Based on the meteorological general data decoding pool, a meteorological general data decoding public class is established with each meteorological general data decoding method; The meteorological general data decoding public class is used as the outer loop, and the meteorological multi-source numerical forecast data is used as the inner loop nesting. According to the matching formula, the matching index between each meteorological multi-source numerical forecast data and each meteorological general data decoding method is calculated as the decoding loop stop condition, and the data decoding demand function is established; According to the data decoding requirement function, the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method is obtained, a data decoding model is constructed, and a unified meteorological multi-source numerical forecast data set is generated.

3. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 2 is characterized in that: The matching formula is specifically: In the formula, MatchScore(D i ,C j ) is the i-th meteorological multi-source numerical forecast data, C j is the jth meteorological general data decoding method, is the kth dimension parameter value in the i-th meteorological multi-source numerical forecast data, is the kth dimension parameter value in the jth meteorological general data decoding method, W k is the weight of the parameter value of the kth dimension.

4. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 3 is characterized in that: According to the data decoding demand function, the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method is obtained, a data decoding model is constructed, and a unified meteorological multi-source numerical forecast data set is generated, which specifically includes: Based on the meteorological multi-source numerical forecast data corresponding to the meteorological general data decoding method as the root node, and each meteorological general data decoding method as the decoding branch node, a meteorological multi-source numerical forecast data decoding demand decision tree is established, which is connected in series to form a data decoding model; Based on the data decoding model, the meteorological multi-source numerical forecast database is taken as input, and the meteorological multi-source numerical forecast data is divided according to the maximum gain ratio of the decoding branch nodes to generate a unified meteorological multi-source numerical forecast data set.

5. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 4 is characterized in that: The maximum gain ratio of the decoding branch nodes based on the meteorological multi-source numerical forecast data is divided into: In the formula, GR(D,A) is the maximum gain ratio of meteorological multi-source numerical forecast data D to decoding branch node A, IG(D,A) is the information gain of meteorological multi-source numerical forecast data D to decoding branch node A, H(A) is the attribute entropy of meteorological multi-source numerical forecast data D to decoding branch node A, D j The branch nodes for matching the meteorological multi-source numerical forecast data with the jth meteorological general data decoding method are divided into subsets, D is the total set of meteorological multi-source numerical forecast data, log2 is the logarithmic function, and C(A) is the total number of decoding branch nodes A.

6. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 5 is characterized in that: The unified time series data set for meteorological multi-source numerical prediction is integrated according to the meteorological multi-source numerical prediction model to predict the meteorological precipitation probability in the target area; Based on the unified time series data of meteorological multi-source numerical prediction, the correlation factors are screened according to the meteorological precipitation type of the target area to be predicted, and the time series data of meteorological multi-source numerical prediction influence is obtained; Based on the meteorological multi-source numerical prediction impact time series data, data alignment is performed according to spatial resolution and temporal resolution. The temporal resolution is used as the observation window, and the meteorological multi-source numerical prediction impact time series data within the spatial resolution is used as the observation attribute. The meteorological characteristic time series parameters of the target area to be predicted are obtained through a sliding window. The spatial resolution is the target distance, and the temporal resolution is hours; According to the meteorological characteristic time series parameters of the target area to be predicted, the optimal meteorological precipitation type characteristic time series parameters are screened and integrated for the posterior probability of the unified time series data of the meteorological multi-source numerical prediction of the target area, and a multi-mode meteorological characteristic time series parameter set of the target area to be predicted is obtained; According to the multi-mode meteorological characteristic time series parameter set of the target area to be predicted, a meteorological prediction model for the target area is constructed to predict the meteorological precipitation probability of the target area.

7. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 6 is characterized in that: The posterior probability of the meteorological characteristic time series parameters of the target area to be predicted for the unified time series data of the meteorological multi-source numerical prediction of the target area is specifically: In the formula, P(u,t|x v ) is the posterior probability of the characteristic parameter of the vth meteorological precipitation type under the tth unit time of the uth target area to be predicted, P(x v |u,t) is the conditional probability of the u-th target area at the t-th unit time given the v-th meteorological precipitation type characteristic parameter, P(u,t) is the prior probability of the u-th target area to be predicted at the t-th unit time, P(x v ) is the overall probability distribution of characteristic parameters of the vth meteorological precipitation type; The meteorological forecast model for the target area is specifically: In the formula, G U,t is the probability of meteorological precipitation in the tth unit time of the uth target area to be predicted, f() is the indicator function, if x v ≥T, then f(x v ≥T)=1, indicating that precipitation occurred in the area at that time point. Then f(x v ≥T)=0, indicating that there is no precipitation in the area at that time point, T is the precipitation threshold, and m is the set of multi-mode meteorological characteristic time series parameters of the target area to be predicted.

8. The data processing and analysis method for multi-source numerical forecast data based on cloud-edge fusion according to claim 7 is characterized in that: Based on the unified time series data of meteorological multi-source numerical prediction, the associated factors are screened according to the meteorological precipitation type of the target area to be predicted, and the time series data of meteorological multi-source numerical prediction influencing are obtained, including: Determine the meteorological precipitation type attention preference influencing factors of the target area to be detected, and establish the meteorological precipitation type preference array of the target area to be detected; By using linear algebra, a linear mapping is performed on the meteorological precipitation type preference array of the target area to be detected to obtain the meteorological precipitation type preference vector matrix Ns of the target area to be detected; Among them, R uv Match the feature vector of the vth meteorological precipitation type of the uth target area to be predicted, z is the total number of target areas to be predicted, and l is the total number of meteorological precipitation types; For the unified time series data of meteorological multi-source numerical prediction, the data dimension reduction is carried out according to the principal component analysis method to obtain the unified time series dimension reduction data of meteorological multi-source numerical prediction; Based on the similarity between the unified time series dimension reduction data of meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected, positive correlation factors are screened to obtain the time series data affecting meteorological multi-source numerical prediction; Among them, the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction and each element in the meteorological precipitation type preference vector matrix of the target area to be detected is specifically: In the formula, θ(D′ u ,R uv ) is the similarity between the unified time series dimension reduction data of the meteorological multi-source numerical prediction of the u-th target area to be predicted and the v-th meteorological precipitation type preference vector of the u-th target area to be detected.

9. An electronic device, characterized in that: include: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the data processing and analysis method for multi-source numerical forecast data of cloud-edge fusion as described in any one of claims 1-8.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the data processing and analysis method for cloud-edge fused multi-source numerical forecast data described in any one of claims 1-8 is implemented.

Images