A TS score verification method for forecasting cloud area and clear sky area based on satellite data
Through the TS scoring inspection method for predicting cloud areas clear sky areas based on satellite data, using multi-class classifiers and deep learning algorithms, the accuracy problem of the existing meteorological forecasting system in identifying cloud areas and clear sky areas is solved, and efficient clouds are achieved Identification and distinction between zones and clear sky areas improves the accuracy and real-time nature of meteorological forecasts.
Patent Information
- Application Number
- CN202410487709.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-23
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-04-23
Smart Images

Figure CN118260682B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of scoring, and in particular to a TS scoring and verification method for forecasting cloud area and clear sky area based on satellite data. Background Art
[0002] With the rapid development of meteorological science and technology, satellite remote sensing data plays an increasingly important role in weather forecasting. Traditional weather forecasting methods mainly rely on data from ground observation stations. However, due to the uneven distribution of ground observation stations, their data often cannot cover all areas, especially in places where it is difficult to set up stations, such as oceans and remote mountainous areas. Therefore, the introduction of satellite remote sensing data has greatly improved the accuracy and coverage of weather forecasts.
[0003] In weather forecasting, it is crucial to accurately identify and distinguish between cloud areas and clear sky areas. The difference between cloud areas and clear sky areas directly affects the formation, development and dissipation of weather systems, and has a decisive impact on the accuracy of weather forecasts. However, the current weather forecast system often identifies cloud areas and clear sky areas based on simple threshold judgments. This method is often questioned in terms of forecast accuracy when faced with complex and changeable weather systems.
[0004] In order to solve this problem, researchers have proposed a variety of cloud and clear sky identification methods based on satellite data. Most of these methods are based on image processing technology, which processes and analyzes satellite remote sensing images to identify and distinguish between cloud and clear sky areas. Although these methods have improved the accuracy of forecasts to a certain extent, they are difficult to meet the needs of real-time forecasts due to their complex processing and large amount of calculations. Summary of the invention
[0005] The purpose of the present invention is to provide a TS scoring verification method for forecast cloud area and clear sky area based on satellite data.
[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:
[0007] The present invention comprises the following steps:
[0008] Acquire satellite data of a preset area, and pre-process the satellite data;
[0009] Performing satellite cloud detection on the satellite data to identify cloud areas to obtain first observation data, and performing model forecast cloud area identification on the satellite data to obtain second observation data;
[0010] The first observation data and the second observation data are classified to obtain cloud-clear classification data, and TS scoring is performed according to the cloud-clear classification data to obtain a scoring test function; the cloud-clear classification data includes first classification data and second classification data; the first classification data classifies the first observation data using a cloud detection value; the second classification data classifies the second observation data according to water content of hydrometeors;
[0011] A TS scoring verification model for the forecast cloud area and clear sky area is constructed according to the scoring verification function, and the TS scoring verification model for the forecast cloud area and clear sky area is optimized.
[0012] Furthermore, a method for performing satellite cloud detection on the satellite data to identify cloud areas to obtain first observation data, wherein the first observation data is obtained based on satellite data through artificial weather modification model forecast products.
[0013] Furthermore, a method for obtaining second observation data by performing model forecast cloud area identification on the satellite data is provided, wherein the second observation data is obtained according to satellite data through FY4 official cloud top height or cloud detection products.
[0014] Furthermore, the method for obtaining the first classification data includes:
[0015] When the cloud detection value of each grid point in the preset area is 0 or 1, it means that there are clouds or there may be clouds in the preset area; when the cloud top height of each grid point in the preset area is greater than 1km, it means that there are clouds in the preset area;
[0016] When the cloud detection value of each grid point in the preset area is 2 or 3, it means that the preset area is clear or may be clear;
[0017] A multi-class classifier is used to classify the first observation data according to the presence or absence of clouds, and first classified data is output.
[0018] Furthermore, the method for obtaining the second classification data includes:
[0019] Calculate the hydrometeoric water content of the second observation:
[0020]
[0021] The water content of the hydromete is F s , the cloud-water mixing ratio is F a , the cloud-rain mixing ratio is The cloud ice mixing ratio is F c , the cloud-snow mixing ratio is F d , the cloud-graupel mixing ratio is F e ;
[0022] When the water content of the hydrometeor is less than 0.001, the initial classification of the second observation data is cloudless; when the water content of the hydrometeor is greater than 0.001, the initial classification of the second observation data is clouded;
[0023] Calculate the Gini index for the second observation after the initial classification:
[0024]
[0025] The Gini index of the zth category in the second observation data is The probability that the i-th second observation data appears in the z-th category is P i ;
[0026] Determine the optimal attribute based on the Gini index, perform secondary classification on the initial classification data with the optimal attribute as the root, and calculate the splitting parameter of the second-order classification:
[0027]
[0028] The splitting parameter is The relative frequency of the second-level classification in category z is E κ , the number of secondary classifications is m;
[0029] Associate the attributes to the child nodes to obtain the attribute value segmentation training set, and calculate the information entropy under the secondary classification:
[0030]
[0031] The training set is ε, and the set of the second observation data belonging to the κth class of objects is L κ , the control factor is Information entropy is
[0032] Classify the training set according to the attribute characteristics to obtain multiple different objects, and calculate the information gain of the second observation data:
[0033]
[0034] The number of training set attribute features is The information gain is The amount of information required for training set classification is ζ, and the amount of information required for training set partitioning is
[0035] Select the attribute with the highest information gain, mark the leaf nodes, get the score of the attribute with high information gain, and all the training set subsets meet the score;
[0036] When the categories at the nodes are the same, create a category number and output the classification result.
[0037] Furthermore, a method for obtaining a scoring test function by performing TS scoring according to the cloud classification data includes:
[0038] When the first classification data indicates that there are clouds and the second classification data indicates that there are no clouds, the forecast is a missed report;
[0039] When the first classification data indicates no cloud and the second classification data indicates cloudy, the forecast is empty;
[0040] When the first classification data is cloud-free and the second classification data is cloud-free, the forecast is correct;
[0041] Given the scoring test function:
[0042]
[0043] The false alarm rate of the kth preset area is U k , the null alarm rate of the kth preset area is V k , the accuracy of the kth preset area is G k , the test function of the k-th preset area is H(k), the natural constant is e, and the actual value of the υ-th value in the k-th preset area is q1(k υ ), the predicted value of the k-th preset area is q2(k υ ), the number of comparison data is The influence coefficient of the υth first classification data in the kth preset area is The adjustment factor of the kth preset area is d k .
[0044] Furthermore, a method for constructing a TS scoring verification model for forecasting cloud area and clear sky area according to the scoring verification function includes:
[0045] The score verification function is used as the objective function of the TS score verification model for forecasting cloud and clear sky areas;
[0046] The TS score verification model for forecasting cloud and clear sky areas includes stratified sampling, clustering algorithm, deep belief network, and deep convolutional neural network algorithm;
[0047] Stratified sampling randomly selects satellite data from a preset area to obtain training data and test data;
[0048] The clustering algorithm iteratively redistributes the training data to the nearest cluster center according to importance, and selects multiple training data with cluster radius less than 0.76 as prediction data;
[0049] The deep belief network automatically learns the hierarchical feature representation of data by pre-training the restricted Boltzmann machine layer by layer to obtain the prediction features;
[0050] The deep convolutional neural network algorithm predicts cloud and clear sky areas by learning the laws of forecast features.
[0051] Further, the method for optimizing the TS score verification model for the forecast cloud area and clear sky area includes:
[0052] Calculate the step size control factor:
[0053]
[0054] The maximum number of iterations is t max , the position constants are c and h, h is less than c, and the step length control factor is χ o , the number of iterations is t, and the particle position is updated using Levy flight. The expression is:
[0055]
[0056] The initial position of the ath particle in the tth iteration is The optimal position of the particle is The dot product is The random number is σ, and the first updated position of the ath particle in the t+1th iteration is Calculate the particle position and velocity:
[0057]
[0058]
[0059] The position of the ath particle in the t+1th iteration is updated twice: The first updated position of the a-th particle in the t-th iteration is The random numbers from 0 to 1 are u1, u2, and u3 respectively, and the global optimal solution for the particle position is pf bt , the local optimal solution of the particle position is yf bt , the velocity of the ath particle in the t+1th iteration is The inertia weight is η, and the speed of the ath particle in the tth iteration is The first learning factor is The second learning factor is
[0060] Update the particle's position:
[0061]
[0062] The first adaptive parameter is The second adaptive parameter is The position of the ath particle updated three times in the t+1th iteration is The speed of the particle is The search angle of the particle is l, the gain coefficient is r, and the control surface angle is γ.
[0063] In a second aspect, an embodiment of the present application further provides an electronic device, including:
[0064] A processor; and a memory arranged to store computer executable instructions, which when executed cause the processor to perform the method steps described in the first aspect.
[0065] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple applications, the electronic device executes the method steps described in the first aspect.
[0066] The beneficial effects of the present invention are:
[0067] The present invention is a TS scoring inspection method for forecasting cloud areas and clear sky areas based on satellite data. Compared with the prior art, the present invention has the following technical effects:
[0068] The present invention can improve the accuracy of TS scoring inspection of forecast cloud area clear sky area through preprocessing, detection and identification, data classification, acquisition of scoring inspection function, and model construction and model optimization steps, thereby improving the precision of TS scoring inspection of forecast cloud area clear sky area. Optimizing TS scoring inspection of forecast cloud area clear sky area can greatly save resources and improve work efficiency. It can realize automatic TS scoring inspection of forecast cloud area clear sky area and perform scoring correction on TS scoring inspection of forecast cloud area clear sky area in real time, which is of great significance to TS scoring inspection of forecast cloud area clear sky area and can adapt to TS scoring inspection of forecast cloud area clear sky area of different standards and TS scoring inspection requirements of different forecast cloud area clear sky area, and has certain universality. BRIEF DESCRIPTION OF THE DRAWINGS
[0069] Figure 1 A flowchart of the steps of a TS scoring inspection method for forecasting cloud area and clear sky area based on satellite data of the present invention;
[0070] Figure 2 It is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION
[0071] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.
[0072] The present invention provides a TS scoring inspection method for forecasting cloud area and clear sky area based on satellite data, comprising the following steps:
[0073] like Figure 1As shown, in this embodiment, the following steps are included:
[0074] Acquire satellite data of a preset area, and pre-process the satellite data;
[0075] In the actual assessment, satellite data include air temperature, pressure, humidity, wind speed, relative humidity, wind direction and speed;
[0076] The first observation data and the second observation data are classified to obtain cloud-clear classification data, and TS scoring is performed according to the cloud-clear classification data to obtain a scoring test function; the cloud-clear classification data includes first classification data and second classification data; the first classification data classifies the first observation data using a cloud detection value; the second classification data classifies the second observation data according to water content of hydrometeors;
[0077] A TS scoring verification model for the forecast cloud area and clear sky area is constructed according to the scoring verification function, and the TS scoring verification model for the forecast cloud area and clear sky area is optimized.
[0078] In this embodiment, a method for performing satellite cloud detection on the satellite data to identify cloud areas to obtain first observation data is performed. The first observation data is obtained based on satellite data through artificial weather modification model forecast products.
[0079] In this embodiment, a method for obtaining second observation data by performing model forecast cloud area identification on the satellite data is performed, and the second observation data is obtained according to satellite data through FY4 official cloud top height or cloud detection products.
[0080] In this embodiment, the method for obtaining the first classification data includes:
[0081] When the cloud detection value of each grid point in the preset area is 0 or 1, it means that there are clouds or there may be clouds in the preset area; when the cloud top height of each grid point in the preset area is greater than 1km, it means that there are clouds in the preset area;
[0082] When the cloud detection value of each grid point in the preset area is 2 or 3, it means that the preset area is clear or may be clear;
[0083] A multi-class classifier is used to classify the first observation data according to the presence or absence of clouds, and first classified data is output.
[0084] In this embodiment, the method for obtaining the second classification data includes:
[0085] Calculate the hydrometeoric water content of the second observation:
[0086]
[0087] The water content of the hydromete is Fs , the cloud-water mixing ratio is F a , the cloud-rain mixing ratio is The cloud ice mixing ratio is F c , the cloud-snow mixing ratio is F d , the cloud-graupel mixing ratio is F e ;
[0088] When the water content of the hydrometeor is less than 0.001, the initial classification of the second observation data is cloudless; when the water content of the hydrometeor is greater than 0.001, the initial classification of the second observation data is clouded;
[0089] Calculate the Gini index for the second observation after the initial classification:
[0090]
[0091] The Gini index of the zth category in the second observation data is The probability that the i-th second observation data appears in the z-th category is P i ;
[0092] Determine the optimal attribute based on the Gini index, perform secondary classification on the initial classification data with the optimal attribute as the root, and calculate the splitting parameter of the second-order classification:
[0093]
[0094] The splitting parameter is The relative frequency of the second-level classification in category z is E κ , the number of secondary classifications is m;
[0095] Associate the attributes to the child nodes to obtain the attribute value segmentation training set, and calculate the information entropy under the secondary classification:
[0096]
[0097] The training set is ε, and the set of the second observation data belonging to the κth class of objects is L κ , the control factor is Information entropy is
[0098] Classify the training set according to the attribute characteristics to obtain multiple different objects, and calculate the information gain of the second observation data:
[0099]
[0100] The number of training set attribute features is The information gain is The amount of information required for training set classification is ζ, and the amount of information required for training set partitioning is
[0101] Select the attribute with the highest information gain, mark the leaf nodes, get the score of the attribute with high information gain, and all the training set subsets meet the score;
[0102] When the categories at the nodes are the same, create a category number and output the classification result.
[0103] In this embodiment, the method for obtaining a scoring test function by performing TS scoring according to the cloud classification data includes:
[0104] When the first classification data indicates that there are clouds and the second classification data indicates that there are no clouds, the forecast is a missed report;
[0105] When the first classification data indicates no cloud and the second classification data indicates cloudy, the forecast is empty;
[0106] When the first classification data is cloud-free and the second classification data is cloud-free, the forecast is correct;
[0107] Given the scoring test function:
[0108]
[0109] The false alarm rate of the kth preset area is U k , the null alarm rate of the kth preset area is V k , the accuracy of the kth preset area is G k , the test function of the k-th preset area is H(k), the natural constant is e, and the actual value of the υ-th value in the k-th preset area is q1(k υ ), the predicted value of the k-th preset area is q2(k υ ), the number of comparison data is The influence coefficient of the υth first classification data in the kth preset area is The adjustment factor of the kth preset area is d k .
[0110] In this embodiment, the method for constructing a TS scoring verification model for forecasting cloud area and clear sky area according to the scoring verification function includes:
[0111] The score verification function is used as the objective function of the TS score verification model for forecasting cloud and clear sky areas;
[0112] The TS score verification model for forecasting cloud and clear sky areas includes stratified sampling, clustering algorithm, deep belief network, and deep convolutional neural network algorithm;
[0113] Stratified sampling randomly selects satellite data from a preset area to obtain training data and test data;
[0114] The clustering algorithm iteratively redistributes the training data to the nearest cluster center according to importance, and selects multiple training data with cluster radius less than 0.76 as prediction data;
[0115] The deep belief network automatically learns the hierarchical feature representation of data by pre-training the restricted Boltzmann machine layer by layer to obtain the prediction features;
[0116] The deep convolutional neural network algorithm predicts cloud and clear sky areas by learning the laws of forecast features.
[0117] In this embodiment, the method for optimizing the forecast cloud area clear sky area TS score verification model includes:
[0118] Calculate the step size control factor:
[0119]
[0120] The maximum number of iterations is t max , the position constants are c and h, h is less than c, and the step length control factor is χ o , the number of iterations is t, and the particle position is updated using Levy flight. The expression is:
[0121]
[0122] The initial position of the ath particle in the tth iteration is The optimal position of the particle is The dot product is The random number is σ, and the first updated position of the ath particle in the t+1th iteration is Calculate the particle position and velocity:
[0123]
[0124]
[0125] The position of the ath particle in the t+1th iteration is updated twice: The first updated position of the a-th particle in the t-th iteration is The random numbers from 0 to 1 are u1, u2, and u3 respectively, and the global optimal solution for the particle position is pf bt , the local optimal solution of the particle position is yf bt , the velocity of the ath particle in the t+1th iteration is The inertia weight is η, and the speed of the ath particle in the tth iteration is The first learning factor is The second learning factor is
[0126] Update the particle's position:
[0127]
[0128] The first adaptive parameter is The second adaptive parameter is The position of the ath particle updated three times in the t+1th iteration is The speed of the particle is The search angle of the particle is l, and the gain coefficient is The control surface rotation angle is γ.
[0129] Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 2 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.
[0130] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0131] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.
[0132] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a TS score inspection device for the forecast cloud area and clear sky area based on satellite data at the logical level. The processor executes the program stored in the memory and is specifically used to execute any of the aforementioned TS score inspection methods for the forecast cloud area and clear sky area based on satellite data.
[0133] The above application Figure 1The illustrated embodiment discloses a TS scoring inspection method for forecasting cloud areas and clear sky areas based on satellite data, which can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
[0134] The electronic device may also perform Figure 1 A TS scoring verification method for forecasting cloud and clear sky areas based on satellite data is proposed and implemented Figure 1 The functions of the illustrated embodiment will not be described in detail in the embodiments of the present application.
[0135] An embodiment of the present application also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions. When the instructions are executed by an electronic device including multiple applications, any one of the aforementioned satellite data-based forecast cloud area and clear sky area TS score verification methods is executed.
[0136] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0137] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0138] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0140] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0141] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0142] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0143] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0144] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
[0145] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A TS scoring verification method for forecasting cloud area and clear sky area based on satellite data, characterized in that: The following steps are involved: Acquire satellite data of a preset area, and pre-process the satellite data; Performing satellite cloud detection on the satellite data to identify cloud areas to obtain first observation data, and performing model forecast cloud area identification on the satellite data to obtain second observation data; The first observation data and the second observation data are classified to obtain cloud-clear classification data, and TS scoring is performed according to the cloud-clear classification data to obtain a scoring test function; the cloud-clear classification data includes first classification data and second classification data; the first classification data classifies the first observation data using a cloud detection value; the second classification data classifies the second observation data according to water content of hydrometeors; Constructing a TS scoring verification model for the forecast cloud area and clear sky area according to the scoring verification function, and optimizing the TS scoring verification model for the forecast cloud area and clear sky area; The method for obtaining the first classification data includes: When the cloud detection value of each grid point in the preset area is 0 or 1, it means that there are clouds or there may be clouds in the preset area; when the cloud top height of each grid point in the preset area is greater than 1km, it means that there are clouds in the preset area; When the cloud detection value of each grid point in the preset area is 2 or 3, it means that the preset area is clear or may be clear; Using a multi-class classifier to classify the first observation data according to whether there is cloud or not, and outputting first classified data; The method for obtaining the second classification data includes: Calculate the hydrometeoric water content of the second observation: The water content of the hydromete is F s , the cloud-water mixing ratio is F a , the cloud-rain mixing ratio is F l , the cloud ice mixing ratio is F c , the cloud-snow mixing ratio is F d , the cloud-graupel mixing ratio is F e ; When the water content of the hydrometeor is less than 0.001, the initial classification of the second observation data is cloudless; when the water content of the hydrometeor is greater than 0.001, the initial classification of the second observation data is clouded; Calculate the Gini index for the second observation after the initial classification: The Gini index of the zth category in the second observation data is The probability that the i-th second observation data appears in the z-th category is P i ; Determine the optimal attribute based on the Gini index, perform secondary classification on the initial classification data with the optimal attribute as the root, and calculate the splitting parameter of the second-order classification: The splitting parameter is The relative frequency of the second-level classification in category z is E κ , the number of secondary classifications is Associate the attributes to the child nodes to obtain the attribute value segmentation training set, and calculate the information entropy under the secondary classification: The training set is ε, and the set of the second observation data belonging to the κth class of objects is L κ , the control factor is Information entropy is Classify the training set according to the attribute characteristics to obtain multiple different objects, and calculate the information gain of the second observation data: The number of attribute features in the training set is z, and the information gain is The amount of information required for training set classification is ζ, and the amount of information required for training set partitioning is Select the attribute with the highest information gain, mark the leaf nodes, get the score of the attribute with high information gain, and all the training set subsets meet the score; When the categories at the nodes are the same, create a category number and output the classification result; The method for optimizing the TS score verification model for the forecast cloud area and clear sky area includes: Calculate the step size control factor: The maximum number of iterations is t max , the position constants are c and h, h is less than c, and the step length control factor is χ o , the number of iterations is t, and the particle position is updated using Levy flight. The expression is: The initial position of the ath particle in the tth iteration is The optimal position of the particle is The dot product is The random number is σ, and the first updated position of the ath particle in the t+1th iteration is Calculate the particle position and velocity: The position of the ath particle in the t+1th iteration is updated twice: The first updated position of the a-th particle in the t-th iteration is The random numbers from 0 to 1 are u1, u2, and u3 respectively, and the global optimal solution for the particle position is pf bt , the local optimal solution of the particle position is yf bt , the velocity of the ath particle in the t+1th iteration is The inertia weight is η, and the speed of the ath particle in the tth iteration is The first learning factor is The second learning factor is Update the particle's position: The first adaptive parameter is The second adaptive parameter is The position of the ath particle updated three times in the t+1th iteration is The speed of the particle is The search angle of the particle is l, and the gain coefficient is The control surface rotation angle is γ.
2. According to the TS scoring verification method for forecasting cloud area and clear sky area based on satellite data in claim 1, it is characterized in that: A method for performing satellite cloud detection on the satellite data to identify cloud areas to obtain first observation data, wherein the first observation data is obtained based on satellite data through artificial weather influence model forecast products.
3. According to the TS scoring inspection method for forecasting cloud area and clear sky area based on satellite data in claim 1, it is characterized in that: A method for obtaining second observation data by performing model forecast cloud area identification on the satellite data, wherein the second observation data is obtained according to satellite data through FY4 official cloud top height or cloud detection products.
4. According to the TS scoring inspection method for forecasting cloud area and clear sky area based on satellite data in claim 1, it is characterized in that: The method for obtaining a scoring test function by performing TS scoring according to the cloud classification data includes: When the first classification data indicates that there are clouds and the second classification data indicates that there are no clouds, the forecast is a missed report; When the first classification data indicates no cloud and the second classification data indicates cloudy, the forecast is empty; When the first classification data is cloud-free and the second classification data is cloud-free, the forecast is correct; Given the scoring test function: The false alarm rate of the kth preset area is U k , the null alarm rate of the kth preset area is V k , the accuracy of the kth preset area is G k , the test function of the k-th preset area is H(k), the natural constant is e, and the actual value of the υ-th value in the k-th preset area is q1(k υ ), the predicted value of the k-th preset area is q2(k υ ), the number of comparison data is The influence coefficient of the υth first classification data in the kth preset area is The adjustment factor of the kth preset area is d k .
5. According to the method for checking TS scores of cloud area and clear sky area forecast based on satellite data in claim 1, it is characterized in that: The method for constructing a TS scoring verification model for a forecast cloud area and clear sky area according to the scoring verification function comprises: The score verification function is used as the objective function of the TS score verification model for forecasting cloud and clear sky areas; The TS score verification model for forecasting cloud and clear sky areas includes stratified sampling, clustering algorithm, deep belief network, and deep convolutional neural network algorithm; Stratified sampling randomly selects satellite data from a preset area to obtain training data and test data; The clustering algorithm iteratively redistributes the training data to the nearest cluster center according to importance, and selects multiple training data with cluster radius less than 0.76 as prediction data; The deep belief network automatically learns the hierarchical feature representation of data by pre-training the restricted Boltzmann machine layer by layer to obtain the prediction features; The deep convolutional neural network algorithm predicts cloud and clear sky areas by learning the laws of forecast features.
6. An electronic device comprising: processor; as well as A memory arranged to store computer executable instructions, which, when executed, cause the processor to perform the method according to any one of claims 1 to 5.
7. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of application programs, enables the electronic device to execute the method according to any one of claims 1 to 5.
