A cloud seeding effect evaluation method for layered hybrid clouds embedded in convective cells

Abstract

The present invention discloses a method for evaluating the cloud seeding effect of a layered hybrid cloud embedded in a convective cell, comprising collecting monitoring data and image data of the embedded convective cell, performing monomer feature extraction on the image data to obtain obvious monomer features, taking the area where the obvious monomer features are located as the first impact area, and vice versa as the non-impact area, calculating the operation impact of the non-impact area using a catalyst diffusion propagation model based on the monitoring data to obtain a second impact area, determining the background area through the echo intensity threshold of the remaining area, constructing a layered hybrid cloud seeding effect function based on the impact area and the background area, constructing a layered hybrid cloud seeding effect evaluation model based on the layered hybrid cloud seeding effect function, and optimizing the layered hybrid cloud seeding effect evaluation model. This method can not only improve the accuracy of the evaluation of the cloud seeding effect of the layered hybrid cloud, but also has good interpretability, and can be directly applied to the cloud seeding effect evaluation system of the layered hybrid cloud.

Description

Technical Field

[0001] The present invention relates to the field of evaluation, and in particular to a method for evaluating the cloud seeding effect of stratified hybrid clouds embedded in convective cells. Background Art

[0002] Weather modification refers to the use of technical means to artificially influence the physical and chemical processes of the atmosphere under appropriate conditions in order to avoid and mitigate meteorological disasters, so as to achieve the goals of increasing rainfall, preventing hail, reducing rain, and eliminating fog, so that local weather phenomena change in a predetermined direction. Generally speaking, the effect of artificial catalysis refers to the changes in the microphysical processes and precipitation processes of the target cloud. Due to the lack of scientific and effective methods and means, the evaluation of artificial catalysis effects has always been one of the technical difficulties that restrict the development of weather modification work.

[0003] At present, the commonly used effect evaluation methods in scientific research and business include physical inspection and numerical simulation inspection. Among them, physical inspection is limited by the coverage and accuracy of cloud precipitation observation instruments, and numerical simulation inspection is often determined by the accuracy of the model's own forecast of cloud precipitation. Therefore, it is still difficult to achieve a scientific and quantitative evaluation of the effect of artificial cloud seeding. For the evaluation of the operation effect of uniform layered clouds and convective cells, an intuitive comparison method is often used. For stratified mixed clouds, especially when there are embedded convective cells, there is a huge natural variation in the process of development and evolution. The difficulty of effect evaluation is to separate the impact of cloud seeding from the natural evolution of the cloud system. The traditional intuitive comparison method has a strong subjectivity in the selection of comparison areas, and sometimes it is even difficult to select a suitable comparison area. Therefore, it is urgent to develop an evaluation method for the artificial catalytic effect of stratified mixed clouds with embedded convective cells. Summary of the invention

[0004] The purpose of the present invention is to provide a method for evaluating the cloud seeding effect of stratified hybrid clouds embedded in convective cells.

[0005] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0006] The present invention comprises the following steps:

[0007] Collecting monitoring data and image data of the embedded convection cell, and preprocessing the monitoring data and the image data; the monitoring data includes radar data and flight detection data;

[0008] Extracting monomer features from the image data to obtain obvious monomer features, and taking the area where the obvious monomer features are located as the first affected area, and vice versa as the non-affected area;

[0009] The operation impact of the non-impacted area is calculated using a catalyst diffusion propagation model according to the monitoring data to obtain a second impacted area, a background area is determined by an echo intensity threshold of the remaining area, the first impacted area and the second impacted area are integrated into an impacted area, and a layered hybrid cloud seeding effect function is constructed according to the impacted area and the background area;

[0010] A layered hybrid cloud seeding cloud effect evaluation model is constructed according to the layered hybrid cloud seeding cloud effect function, the layered hybrid cloud seeding cloud effect evaluation model is optimized, the data to be evaluated is input into the layered hybrid cloud seeding cloud effect evaluation model, and the evaluation result is output.

[0011] Furthermore, the method of extracting monomer features from the image data to obtain obvious monomer features includes:

[0012] The image data is converted from a Cartesian coordinate system to a spherical coordinate system, and an edge detection algorithm is used to extract the edge of the image data to obtain the image edge;

[0013] Divide the edge into regions and calculate the correlation between the regional edge and the cloud cluster:

[0014]

[0015] The number of edges is K, the number of clouds is Y, the similarity between the kth edge and the yth cloud is S(k,y), the deviation coefficient is p, and the similarity between the regional edge and the cloud is

[0016] The edges of the region are accumulated and averaged, and the image composed of the edges with an average value greater than 0.57 is extracted as a cloud cluster. The convolutional neural network algorithm is used to extract the features of the cloud cluster to obtain the cloud cluster features;

[0017] Sort the cloud features in chronological order, compare the cloud features at different times, and calculate the degree of change of cloud features at different times:

[0018]

[0019] The similarity between the cth cloud cluster characteristics of the wth cloud cluster at the sth moment and the s+1th moment is H w,c (s,s+1), the number of cloud features is The degree of change of the w-th cloud cluster at the s-th moment and the s+1-th moment is The position of the cloud at the s+1th moment is The position of the cloud at the sth moment is The control factor is σ, and the distance between the cth cloud feature at the sth moment and the s+1th moment is The error constant of the w-th cloud is μ w ;

[0020] Cloud features with a variation greater than 0.417 are output as obvious single features.

[0021] Furthermore, the method of calculating the operation impact of the non-impacted area using a catalyst diffusion propagation model according to the monitoring data includes:

[0022] The catalyst diffusion propagation model was simulated using molecular dynamics;

[0023] Calculate the importance of monitoring data:

[0024]

[0025] The importance of the u-th monitoring data is g u , the convection cell is The monitoring data is l, and the uth monitoring data at the τ+1th moment is l u,τ+1 , the uth monitoring data at the τth time is l u,τ , convection cell The probability of occurrence is The probability of occurrence of monitoring data l is q(l), and the probability of convection cell The probability of appearing simultaneously with the monitoring data l is The deviation coefficient is δ, and the wind field influence coefficient is Q;

[0026] The monitoring data with importance greater than 0.407 are taken as key data, and the operational impact of key data in the non-impact area is calculated:

[0027]

[0028] The operation impact of the i-th non-impacted area is The number of key data in the non-affected area is The importance of the κth key data in the i-th non-influence area is g i,κ , the initial κ-th key data value of the i-th non-affected area is χ i,κ , after embedding the convection cell, the k-th key data of the i-th non-affected area is

[0029] The non-influenced area with operation impact greater than 0.657 is output as the second influence area.

[0030] Furthermore, the method for determining the background area by the echo intensity threshold of the remaining area includes:

[0031] The echo intensity threshold is determined by meteorological principles, and the remaining area with an echo intensity less than the threshold is identified as the background area using the threshold segmentation method.

[0032] Furthermore, the method for constructing a layered hybrid cloud seeding effect function according to the influence area and the background area includes:

[0033] The correlation coefficient method is used to screen the monitoring data, obtain the index data, and construct the stratified hybrid cloud seeding effect function, which is expressed as follows:

[0034]

[0035] The effect weight of the jth indicator is θ j , the difference between the jth index before and after the operation in the affected area is ΔR j , the difference between the jth index before and after the background area operation is The natural variation coefficient β of the jth indicator j , the cloud seeding effect function of the accumulative hybrid cloud is

[0036] Furthermore, a method for constructing a layered hybrid cloud seeding effect evaluation model according to the layered hybrid cloud seeding effect function includes:

[0037] Construct the objective function based on the cumulative hybrid cloud seeding effect function:

[0038]

[0039] The objective function is The cloud seeding effect function of the accumulative hybrid cloud is: The actual evaluation value is z, and the predicted evaluation value is z * , the heritability coefficient is α;

[0040] The evaluation model of the cloud seeding effect of the accumulative hybrid cloud includes the bootstrap method, deviation analysis algorithm, hierarchical clustering algorithm, association rule algorithm, and recurrent neural network algorithm;

[0041] The bootstrap method randomly extracts the input data at a ratio of 5:1 to obtain training data and test data;

[0042] The deviation analysis algorithm sorts the training data in time sequence, analyzes the cloud cluster changes on the sorted training data, and obtains the degree of cloud cluster deviation;

[0043] The hierarchical clustering algorithm performs deviation clustering on the training data according to the degree of cloud cluster deviation to obtain cluster data;

[0044] The association rule algorithm selects cluster data according to the correlation degree of cluster data to obtain key data;

[0045] The recurrent neural network algorithm sorts the key data according to time, captures the time variation pattern of the key data after sorting, and evaluates and predicts the test data according to the time variation pattern to obtain the prediction evaluation;

[0046] The parameters of the layered hybrid cloud seeding effect evaluation model are adjusted according to the actual evaluation and predicted evaluation variance of the test data.

[0047] Furthermore, the method for optimizing the layered hybrid cloud seeding effect evaluation model includes:

[0048] Calculate the decision variable:

[0049]

[0050] The decision variable is Determine the initial value of the variable The maximum number of iterations is t max , the current number of iterations is t;

[0051] Calculate the fitness value of the searcher:

[0052]

[0053] The fitness value is The target value is The estimated target value is The average target value is The optimization factor is m and the natural constant is e;

[0054] The searcher with the smallest fitness value is selected as the primary searcher. When the absolute value of the judgment variable is less than 1, the searcher position is calculated according to the random walk mechanism:

[0055]

[0056] The position of the vth dimension of the ath searcher in the t+1th iteration is The random perturbation of the roulette wheel of the a-th searcher in the t-th iteration is The elite random perturbation of the a-th searcher in the t-th iteration is

[0057] Otherwise, the search subposition is calculated according to the global search:

[0058]

[0059] The random number between 0 and 1 is The upper bound of the solution space is V max , the lower bound of the solution space is V min , the random numbers from -1 to 1 are r1, r2, r3, r4, r5, r6, and the falling trend is The position of the vth dimension of the tth iteration of the random number r3 is The position of the vth dimension of the tth iteration of the random number r4 is The position of the primary searcher is C, and the average position of the primary searcher is The position of the primary high-quality searcher is C E ;

[0060] Update the position of the primary searcher, the expression is:

[0061]

[0062] The position of the ath primary searcher after update is The position of the b-th searcher is Location The fitness value is Position C a The fitness value is The average position of the primary high-quality searcher is

[0063] Based on the adaptive update searcher position, the expression is:

[0064]

[0065] The position of the a-th searcher in the t+1-th iteration after the update is The control parameter is ρ and the attenuation coefficient is

[0066] Continue iterating until the maximum number of iterations is reached, otherwise update the position of the primary searcher.

[0067] In a second aspect, an embodiment of the present application further provides an electronic device, including:

[0068] 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.

[0069] 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.

[0070] The beneficial effects of the present invention are:

[0071] The present invention is a method for evaluating the cloud seeding effect of stratified hybrid clouds embedded in convective cells. Compared with the prior art, the present invention has the following technical effects:

[0072] The present invention can improve the accuracy of cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells through preprocessing, feature extraction, determining image areas, determining background areas, constructing effect functions, model building and model optimization steps, thereby improving the precision of cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells, optimizing the cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells, greatly saving resources and improving work efficiency, realizing automatic evaluation of cloud seeding effect for stratified mixed cloud, and making real-time correction of influence area for cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells, which is of great significance for cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells, can adapt to cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells of different standards and different requirements for cloud seeding effect evaluation for stratified mixed cloud embedded in convective cells, and has certain universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a flowchart of the steps of a method for evaluating the cloud seeding effect of stratified hybrid clouds embedded in convective cells according to the present invention;

[0074] Figure 2 It is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION

[0075] 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.

[0076] The present invention provides a method for evaluating the cloud seeding effect of a layered hybrid cloud embedded in a convective cell, comprising the following steps:

[0077] like Figure 1 As shown, in this embodiment, the following steps are included:

[0078] Collecting monitoring data and image data of the embedded convection cell, and preprocessing the monitoring data and the image data; the monitoring data includes radar data and flight detection data;

[0079] In the actual evaluation, after the radar data is subjected to data quality control such as eliminating interference signals and ground clutter, the reflectivity data is interpolated from the Cartesian coordinate system to the spherical coordinate system. The interpolation method uses a combination of the nearest neighbor method in the horizontal direction and linear interpolation in the vertical direction.

[0080] Taking an aircraft warm cloud rain reduction test in XX region on September 4, 2016 as an example, Zhoushan S-band network weather radar was used. First, the radar base data was quality controlled and gridded. The horizontal resolution of the data was 0.01°×0.01°, the vertical resolution was 500m, and the time resolution was about 6min.

[0081] Extracting monomer features from the image data to obtain obvious monomer features, and taking the area where the obvious monomer features are located as the first affected area, and vice versa as the non-affected area;

[0082] The operation impact of the non-impacted area is calculated using a catalyst diffusion propagation model according to the monitoring data to obtain a second impacted area, a background area is determined by an echo intensity threshold of the remaining area, the first impacted area and the second impacted area are integrated into an impacted area, and a layered hybrid cloud seeding effect function is constructed according to the impacted area and the background area;

[0083] In the actual assessment, based on the changes in echo parameters in the impact area tracked, and considering the duration of the formation and disappearance of convective cloud cells, the period from 1 to 2 hours before the operation to 1 to 2 hours after the operation was selected as the focus period;

[0084] According to the flight trajectory and the corresponding radar echo distribution, area A is determined to be the effective detection range of the radar, i.e., the background area; area B is the catalyst impact echo range identified by the echo tracking method, i.e., the impact area;

[0085] Constructing a layered hybrid cloud seeding cloud effect evaluation model according to the layered hybrid cloud seeding cloud effect function, optimizing the layered hybrid cloud seeding cloud effect evaluation model, inputting the data to be evaluated into the layered hybrid cloud seeding cloud effect evaluation model, and outputting the evaluation result;

[0086] In the actual evaluation, according to the height of the catalyst layer, considering the diffusion and transmission of the catalyst in the vertical direction, the average value of the echo intensity of the two areas at 1500-2500m was calculated, and the difference before and after the operation was analyzed ( Figure 2 ). During the whole process, the average echo intensity of area A did not change significantly (ref_A=9.1±0.3dBz), and the echo intensity of area B decreased significantly; the echo change in the affected area exceeded that in the background area, and artificial cloud seeding took effect; by calculation, kA=0.004, kB=-0.509, and the echo intensity of the affected area caused by cloud seeding was reduced by about 51.3% compared with the background echo intensity during the period of interest;

[0087] The echo intensity was divided into ≤0dBz, 0-10dBz, 10-20dBz, 20-30dBz, and ≥30dBz, and the frequency distribution changes of echo intensity in the impact area and background area before and after the operation were calculated. The echo intensity in each range of the background area did not change significantly over time, the echo intensity above 20dBz in the impact area decreased significantly, and the echo intensity below 10dBz increased significantly.

[0088] In this embodiment, the method of extracting monomer features from the image data to obtain obvious monomer features includes:

[0089] The image data is converted from a Cartesian coordinate system to a spherical coordinate system, and an edge detection algorithm is used to extract the edge of the image data to obtain the image edge;

[0090] Divide the edge into regions and calculate the correlation between the regional edge and the cloud cluster:

[0091]

[0092] The number of edges is K, the number of clouds is Y, the similarity between the kth edge and the yth cloud is S(k,y), the deviation coefficient is p, and the similarity between the regional edge and the cloud is

[0093] The edges of the region are accumulated and averaged, and the image composed of the edges with an average value greater than 0.57 is extracted as a cloud cluster. The convolutional neural network algorithm is used to extract the features of the cloud cluster to obtain the cloud cluster features;

[0094] Sort the cloud features in chronological order, compare the cloud features at different times, and calculate the degree of change of cloud features at different times:

[0095]

[0096] The similarity between the cth cloud cluster characteristics of the wth cloud cluster at the sth moment and the s+1th moment is H w,c (s,s+1), the number of cloud features is The degree of change of the w-th cloud cluster at the s-th moment and the s+1-th moment is The position of the cloud at the s+1th moment is The position of the cloud at the sth moment is The control factor is σ, and the distance between the cth cloud feature at the sth moment and the s+1th moment is The error constant of the w-th cloud is μ w ;

[0097] Cloud features with a variation greater than 0.417 are output as obvious single features.

[0098] In this embodiment, the method of calculating the operation impact of the non-impacted area using a catalyst diffusion propagation model according to the monitoring data includes:

[0099] The catalyst diffusion propagation model was simulated using molecular dynamics;

[0100] Calculate the importance of monitoring data:

[0101]

[0102] The importance of the u-th monitoring data is g u , the convection cell is The monitoring data is l, and the uth monitoring data at the τ+1th moment is l u,τ+1 , the uth monitoring data at the τth time is l u,τ , convection cell The probability of occurrence is The probability of occurrence of monitoring data l is q(l), and the probability of convection cell The probability of appearing simultaneously with the monitoring data l is The deviation coefficient is δ, and the wind field influence coefficient is Q;

[0103] The monitoring data with importance greater than 0.407 are taken as key data, and the operational impact of key data in the non-impact area is calculated:

[0104]

[0105] The operation impact of the i-th non-impacted area is The number of key data in the non-affected area is The importance of the κth key data in the i-th non-influence area is g i,κ , the initial κ-th key data value of the i-th non-affected area is χ i,κ , after embedding the convection cell, the k-th key data of the i-th non-affected area is

[0106] The non-influenced area with operation impact greater than 0.657 is output as the second influence area.

[0107] In this embodiment, the method for determining the background area by the echo intensity threshold of the remaining area includes:

[0108] The echo intensity threshold is determined by meteorological principles, and the remaining area with an echo intensity less than the threshold is identified as the background area using the threshold segmentation method.

[0109] In this embodiment, the method for constructing a layered hybrid cloud seeding effect function according to the influence area and the background area includes:

[0110] The correlation coefficient method is used to screen the monitoring data, obtain the index data, and construct the stratified hybrid cloud seeding effect function, which is expressed as follows:

[0111]

[0112] The effect weight of the jth indicator is θ j , the difference between the jth index before and after the operation in the affected area is ΔR j , the difference between the jth index before and after the background area operation is The natural variation coefficient β of the jth indicator j , the cloud seeding effect function of the accumulative hybrid cloud is

[0113] In this embodiment, a method for constructing a layered hybrid cloud seeding effect evaluation model according to the layered hybrid cloud seeding effect function includes:

[0114] Construct the objective function based on the cumulative hybrid cloud seeding effect function:

[0115]

[0116] The objective function is The cloud seeding effect function of the accumulative hybrid cloud is: The actual evaluation value is z, and the predicted evaluation value is z * , the heritability coefficient is α;

[0117] The evaluation model of the cloud seeding effect of the accumulative hybrid cloud includes the bootstrap method, deviation analysis algorithm, hierarchical clustering algorithm, association rule algorithm, and recurrent neural network algorithm;

[0118] The bootstrap method randomly extracts the input data at a ratio of 5:1 to obtain training data and test data;

[0119] The deviation analysis algorithm sorts the training data in time sequence, analyzes the cloud cluster changes on the sorted training data, and obtains the degree of cloud cluster deviation;

[0120] The hierarchical clustering algorithm performs deviation clustering on the training data according to the degree of cloud cluster deviation to obtain cluster data;

[0121] The association rule algorithm selects cluster data according to the correlation degree of cluster data to obtain key data;

[0122] The recurrent neural network algorithm sorts the key data according to time, captures the time variation pattern of the key data after sorting, and evaluates and predicts the test data according to the time variation pattern to obtain the prediction evaluation;

[0123] The parameters of the layered hybrid cloud seeding effect evaluation model are adjusted according to the actual evaluation and predicted evaluation variance of the test data.

[0124] In this embodiment, the method for optimizing the layered hybrid cloud seeding effect evaluation model includes:

[0125] Calculate the decision variable:

[0126]

[0127] The decision variable is Determine the initial value of the variable The maximum number of iterations is t max , the current number of iterations is t;

[0128] Calculate the fitness value of the searcher:

[0129]

[0130] The fitness value is The target value is The estimated target value is The average target value is The optimization factor is m and the natural constant is e;

[0131] The searcher with the smallest fitness value is selected as the primary searcher. When the absolute value of the judgment variable is less than 1, the searcher position is calculated according to the random walk mechanism:

[0132]

[0133] The position of the vth dimension of the ath searcher in the t+1th iteration is The random perturbation of the roulette wheel of the a-th searcher in the t-th iteration is The elite random perturbation of the a-th searcher in the t-th iteration is

[0134] Otherwise, the search subposition is calculated according to the global search:

[0135]

[0136] The random number between 0 and 1 is The upper bound of the solution space is V max , the lower bound of the solution space is V min , the random numbers from -1 to 1 are r1, r2, r3, r4, r5, r6, and the falling trend is The position of the vth dimension of the tth iteration of the random number r3 is The position of the vth dimension of the tth iteration of the random number r4 is The position of the primary searcher is C, and the average position of the primary searcher is The position of the primary high-quality searcher is C E ;

[0137] Update the position of the primary searcher, the expression is:

[0138]

[0139] The position of the ath primary searcher after update is The position of the b-th searcher is Location The fitness value is Position C a The fitness value is The average position of the primary high-quality searcher is

[0140] Based on the adaptive update searcher position, the expression is:

[0141]

[0142] The position of the a-th searcher in the t+1-th iteration after the update is The control parameter is ρ and the attenuation coefficient is

[0143] Continue iterating until the maximum number of iterations is reached, otherwise update the position of the primary searcher.

[0144] 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.

[0145] 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.

[0146] 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.

[0147] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a device for evaluating the cloud seeding effect of stratified hybrid cloud embedded in the convective cell at the logical level. The processor executes the program stored in the memory and is specifically used to execute any of the aforementioned methods for evaluating the cloud seeding effect of stratified hybrid cloud embedded in the convective cell.

[0148] The above application Figure 1 The embodiment shown discloses a method for evaluating the cloud seeding effect of a layered hybrid cloud embedded in a convection cell, 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.

[0149] The electronic device may also perform Figure 1 A cloud seeding effect evaluation method for layered hybrid cloud embedded in convective cells 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.

[0150] 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 application programs, any one of the aforementioned methods for evaluating the cloud seeding effect of layered hybrid cloud embedded in convective cells is executed.

[0151] 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.

[0152] 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.

[0153] 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.

[0154] 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.

[0155] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0156] 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.

[0157] 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.

[0158] 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.

[0159] 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.

[0160] 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 method for evaluating the cloud seeding effect of layered hybrid clouds embedded in convective cells, characterized in that: The following steps are involved: Collecting monitoring data and image data of the embedded convection cell, and preprocessing the monitoring data and the image data; the monitoring data includes radar data and flight detection data; Extracting monomer features from the image data to obtain obvious monomer features, taking the area where the obvious monomer features are located as the first impact area, and vice versa as the non-impact area; comprising: The image data is converted from a Cartesian coordinate system to a spherical coordinate system, and an edge detection algorithm is used to extract the edge of the image data to obtain the image edge; Divide the edge into regions and calculate the correlation between the regional edge and the cloud cluster: The number of edges is K, the number of clouds is Y, and the similarity between the kth edge and the yth cloud is , the deviation coefficient is p, and the similarity between the edge of the region and the cloud is ; The edges of the region are accumulated and averaged, and the image consisting of edges with an average value greater than 0.57 is extracted as a cloud. The convolutional neural network algorithm is used to extract the features of the cloud clusters and obtain the cloud cluster features; Sort the cloud features in chronological order, compare the cloud features at different times, and calculate the degree of change of cloud features at different times: The similarity of the cth cloud cluster characteristics between the wth cloud cluster at the sth moment and the s+1th moment is , the number of cloud features is , the degree of change of the wth cloud cluster at the sth moment and the s+1th moment is , the position of the cloud at the s+1th moment is , the position of the cloud at the sth moment is The regulatory factor is , the distance between the cth cloud feature at the sth moment and the s+1th moment is , the error constant of the w-th cloud is ; The cloud features with a variation greater than 0.417 are output as obvious single features; According to the monitoring data, the operation impact of the non-impacted area is calculated by using a catalyst diffusion propagation model to obtain a second impacted area, the background area is determined by an echo intensity threshold of the remaining area, the first impacted area and the second impacted area are integrated into an impacted area, and a layered hybrid cloud seeding effect function is constructed according to the impacted area and the background area; including: The correlation coefficient method is used to screen the monitoring data, obtain the index data, and construct the stratified hybrid cloud seeding effect function, which is expressed as follows: The effect weight of the jth indicator is , the difference of the jth index before and after the operation in the impact area is , the difference between the jth index before and after the background area operation is , the natural variation coefficient of the jth index , the cloud seeding effect function of the accumulative hybrid cloud is ; Constructing a layered hybrid cloud seeding cloud effect evaluation model according to the layered hybrid cloud seeding cloud effect function, optimizing the layered hybrid cloud seeding cloud effect evaluation model, inputting the data to be evaluated into the layered hybrid cloud seeding cloud effect evaluation model, and outputting the evaluation result; comprising: Construct the objective function based on the cumulative hybrid cloud seeding effect function: The objective function is , the cloud seeding effect function of the accumulative hybrid cloud is , the actual evaluation value is z, and the predicted evaluation value is The heritability coefficient is ; The evaluation model of the cloud seeding effect of the accumulative hybrid cloud includes the bootstrap method, deviation analysis algorithm, hierarchical clustering algorithm, association rule algorithm, and recurrent neural network algorithm; The bootstrap method randomly extracts the input data at a ratio of 5:1 to obtain training data and test data; The deviation analysis algorithm sorts the training data in time sequence, analyzes the cloud cluster changes on the sorted training data, and obtains the degree of cloud cluster deviation; The hierarchical clustering algorithm performs deviation clustering on the training data according to the degree of cloud cluster deviation to obtain cluster data; The association rule algorithm selects cluster data according to the correlation degree of cluster data to obtain key data; The recurrent neural network algorithm sorts the key data according to time, captures the time variation pattern of the key data after sorting, and evaluates and predicts the test data according to the time variation pattern to obtain the prediction evaluation; The parameters of the layered hybrid cloud seeding effect evaluation model are adjusted according to the actual evaluation and predicted evaluation variance of the test data.

2. According to claim 1, a method for evaluating the cloud seeding effect of layered hybrid clouds embedded in convective cells is characterized in that: The method for calculating the operation impact of the non-impacted area using a catalyst diffusion propagation model according to the monitoring data includes: The catalyst diffusion propagation model was simulated using molecular dynamics; Calculate the importance of monitoring data: The importance of the u-th monitoring data is , the convection cell is The monitoring data is , No. The uth monitoring data at time is , No. The uth monitoring data at time is , convection cell The probability of occurrence is , monitoring data The probability of occurrence is , convection cell and monitoring data The probability of simultaneous occurrence is , the coefficient of deviation is , the wind field influence coefficient is ; The monitoring data with importance greater than 0.407 are taken as key data, and the operational impact of the key data in the non-impact area is calculated: The operation impact of the i-th non-impacted area is , the number of key data in the non-affected area is , the i-th non-affected area The importance of key data is , the initial i-th non-affected area The key data value is , after embedding the convection cell, the i-th non-affected area The key data is ; The non-influenced area with operation impact greater than 0.657 is output as the second influence area.

3. The method for evaluating the cloud seeding effect of layered hybrid clouds embedded in convective cells according to claim 1 is characterized in that: The method for determining the background area by using the echo intensity threshold of the remaining area includes: The echo intensity threshold is determined by meteorological principles, and the remaining area with an echo intensity less than the threshold is identified as the background area using the threshold segmentation method.

4. According to claim 1, a method for evaluating the cloud seeding effect of layered hybrid clouds embedded in convective cells is characterized in that: The method for optimizing the layered hybrid cloud seeding effect evaluation model comprises: Calculate the decision variable: The decision variable is , the initial value of the variable is determined to be , the maximum number of iterations is , the current number of iterations is t; Calculate the fitness value of the searcher: The fitness value is , the target value is , the estimated target value is The average target value is , the optimization factor is m, and the natural constant is e; The searcher with the smallest fitness value is selected as the primary searcher. When the absolute value of the judgment variable is less than 1, the searcher position is calculated according to the random walk mechanism: The position of the vth dimension of the ath searcher in the t+1th iteration is , the random perturbation of the roulette wheel of the a-th searcher in the t-th iteration is , the elite random perturbation of the a-th searcher in the t-th iteration is ; Otherwise, the search subposition is calculated according to the global search: The random number between 0 and 1 is , the upper bound of the solution space is , the lower bound of the solution space is , the random numbers from -1 to 1 are , , , , , , the falling trend is , random number The position of the vth dimension at the tth iteration is , random number The position of the vth dimension at the tth iteration is , the position of the primary searcher is , the average position of the primary searcher is , the position of the primary high-quality searcher is ; Update the position of the primary searcher, the expression is: The position of the ath primary searcher after update is , the position of the b-th searcher is ,Location The fitness value is ,Location The fitness value is The average position of the primary high-quality searcher is ; Based on the adaptive update searcher position, the expression is: The position of the a-th searcher in the t+1-th iteration after the update is , the control parameters are , the attenuation coefficient is ; Continue iterating until the maximum number of iterations is reached, otherwise update the position of the primary searcher.

Images