Method and device for determining search radius of meteorological data fusion and electronic equipment

Abstract

The embodiment of the invention provides a meteorological data fusion search radius determination method and device and electronic equipment, and the method comprises the steps: obtaining multi-source satellite sea surface temperature data of a target region, and determining the weather information during the collection of the multi-source satellite sea surface temperature data; determining a weather influence factor according to the weather information; by taking preset satellite sea surface temperature data in the starry sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data; determining a target analysis point from the mapped data, and determining a boundary range of a spatio-temporal joint search radius according to a spatio-temporal range of the multi-source satellite sea surface temperature data; and providing the boundary range of the spatio-temporal joint search radius and the weather influence factor to a preset fruit fly algorithm, optimizing the spatio-temporal joint search radius, and determining a target spatio-temporal joint search radius. The information utilization rate is improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of ocean data processing, and particularly relates to a search radius determination method and device for meteorological data fusion and electronic equipment. BACKGROUND

[0002] In multi-source satellite sea surface temperature (SST) data, multi-source refers to data from multiple different satellite sensors or observation platforms. These data sources may include satellites launched by different countries or institutions, different types of sensors, etc., such as polar orbit satellites and geostationary satellites, etc. Satellite SST data refers to the sea surface temperature obtained through satellite remote sensing technology. Satellite remote sensing has the advantages of wide coverage, high observation frequency, and rapid data acquisition, and can provide global sea surface temperature information for studying the interaction between the ocean and the atmosphere, ocean dynamics, climate change, etc. The sea surface temperature has a certain correlation in space, that is, the sea surface temperature of adjacent regions usually has similarity. By setting a search radius, data related to the target point within a certain spatial range can be found, so as to better utilize this spatial correlation to improve the accuracy of the fusion result. That is, the search radius defines the spatial range of data points considered in the fusion process. A larger search radius will include more data points, but may result in overly smooth results; a smaller search radius will retain more details, but may result in overly noisy results. In related technologies, only the time dimension is considered when determining the search radius in the data fusion process, and the influence of weather is not considered, so that the determined search radius is not accurate enough. SUMMARY

[0003] Embodiments of the present disclosure provide a search radius determination method, device and electronic equipment for meteorological data fusion to solve the problem of inaccurate search radius determination.

[0004] Based on the above problems, in a first aspect, a search radius determination method for meteorological data fusion is provided, comprising:

[0005] Obtaining multi-source satellite sea surface temperature data of a target region, and determining weather information at the time of collecting the multi-source satellite sea surface temperature data;

[0006] Determining a weather influence factor according to the weather information; and

[0007] Taking preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data;

[0008] Determining a target analysis point from the mapped data, and determining a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data;

[0009] The boundary range of the spatiotemporal joint search radius and the weather influencing factor are provided to a preset fruit fly algorithm, the spatiotemporal joint search radius is optimized, and a target spatiotemporal joint search radius is determined.

[0010] In conjunction with the first aspect, in a possible implementation, determining a weather influencing factor according to the weather information includes:

[0011] Determining physical quantities corresponding to weather types included in the weather information, and an influence coefficient of each weather type on the search radius;

[0012] Use the influence coefficient as a weight to determine the weighted average of the physical quantities of each weather type; and

[0013] A weather impact factor is determined based on the weighted average.

[0014] In conjunction with the first aspect, in one possible implementation, providing the boundary range of the spatiotemporal joint search radius and the weather influencing factor to a preset fruit fly algorithm, optimizing the spatiotemporal joint search radius, and determining a target spatiotemporal joint search radius includes:

[0015] Determine the initialization parameters of the preset fruit fly algorithm; the initialization parameters include: fruit fly population size, maximum number of iterations, and olfactory search step length; and

[0016] The objective function is constructed with the goal of minimizing the influence of the spatiotemporal joint search radius on multiple preset factors;

[0017] Determining an initial spatiotemporal joint search radius for each fruit fly according to the boundary range of the spatiotemporal joint search radius and a uniformly distributed random number formula;

[0018] Determining an initial spatiotemporal joint search radius constrained by weather according to the weather influencing factors;

[0019] Inputting the initial spatiotemporal joint search radius after weather constraints into the objective function, determining the objective function value, and determining the optimal solution;

[0020] Olfactory search and visual search are performed in multiple iterative operations. In each olfactory search iteration, the current spatiotemporal joint search radius of each fruit fly is randomly adjusted according to the olfactory search step size, and whether the current spatiotemporal joint search radius is updated is determined according to the objective function; in each visual search iteration, the current spatiotemporal joint search radius of each fruit fly is input into the objective function to determine the function value, and the optimal solution is determined based on the function value; according to the preset visual search step size attenuation formula, the spatiotemporal joint search radii corresponding to other fruit flies are made close to the current optimal solution, and a new current spatiotemporal joint search radius is obtained based on the weather influencing factor; until the iteration exit condition is met, the target spatiotemporal joint search radius is obtained.

[0021] In combination with the first aspect, in one possible implementation, the objective function is constructed using the following formula:

[0022]

[0023] Among them, R s Characterizes the spatial search radius in the spatiotemporal joint search radius; R t represents the time search radius in the spatiotemporal joint search radius; ω1, ω2, and ω3 represent the weight coefficients of the corresponding items; and W represents the weather influence factor.

[0024] In conjunction with the first aspect, in one possible implementation, determining an initial spatiotemporal joint search radius for each fruit fly according to the boundary range of the spatiotemporal joint search radius and a uniformly distributed random number formula includes:

[0025] The initial space-time joint search radius is determined using the following formula:

[0026] R s =R s，min +(R s，max -R s，min )×rand()

[0027] R t =R t，min +(R t，max -R t，min )×rand()

[0028] Among them, R s Characterizes the spatial search radius in the spatiotemporal joint search radius; R t Characterizes the time search radius in the space-time joint search radius; R s，max Characterizes the maximum value of the spatial search radius; R s，min Characterizes the maximum value of the spatial search radius; R t，max Characterizes the maximum value of the time search radius; R t，min Represents the maximum value of the time search radius; rand() represents the random function.

[0029] With reference to the first aspect, in a possible implementation, in each olfactory search iteration, the current spatiotemporal joint search radius of each fruit fly is randomly adjusted according to an olfactory search step, and it is determined whether to update the current spatiotemporal joint search radius according to the objective function, comprising:

[0030] In each olfactory search iteration, for each fruit fly, the current spatiotemporal joint search radius is randomly adjusted according to an olfactory search step by using the following formula:

[0031] R i,snew = (R si + rand s () × step s ) × W

[0032] R t,snew = (R ti + rand t () × step t ) × W

[0033] wherein R i,snew represents the spatial search radius of the i-th fruit fly after adjustment in the current iteration; R si represents the spatial search radius of the i-th fruit fly before adjustment in the current iteration; rand s represents the random disturbance direction and size corresponding to the spatial search radius in the current iteration; step s represents the olfactory search step for space; R t,snew represents the temporal search radius of the i-th fruit fly after adjustment in the current iteration; R ti represents the temporal search radius of the i-th fruit fly before adjustment in the current iteration; rand t represents the random disturbance direction and size corresponding to the temporal search radius in the current iteration; step t represents the olfactory search step for time; and W represents the weather influence factor

[0034] The spatiotemporal joint search radius before and after adjustment in the current iteration is input into the objective function, respectively, to obtain an objective function value;

[0035] It is determined whether to update the spatiotemporal joint search radius before adjustment using the spatiotemporal joint search radius after adjustment according to the objective function value.

[0036] With reference to the first aspect, in a possible implementation, the spatiotemporal joint search radius corresponding to other fruit flies is caused to approach the current optimal solution according to a preset visual search step attenuation formula, and a new current spatiotemporal joint search radius is obtained based on the weather influence factor, comprising:

[0037] For each of the other fruit flies, the spatial search radius adjustment step size of the current iteration is determined according to the following formula:

[0038]

[0039] Wherein, step i，s represents the spatial search radius adjustment step size of the i th fruit fly in the current iteration; R si represents the initial spatial search radius of the i th fruit fly in the current iteration; step i，t represents the time search radius adjustment step size of the i th fruit fly in the current iteration; R ti represents the initial time search radius of the i th fruit fly in the current iteration; t represents the number of local iterations, and T represents the total number of visual search iterations;

[0040] The spatial search radius of the current iteration is determined by the following formula:

[0041] R simove = (R si + step i，s × rand s ) × W

[0042] R timove = (R ti + step i，t × rand t ) × W

[0043] Wherein, R simove represents the spatial search radius of the i th fruit fly after approaching the current spatial search radius optimal solution; rand s represents the approaching direction and random size of the spatial search radius corresponding to the current iteration; R timove represents the time search radius of the i th fruit fly after approaching the current time search radius optimal solution; rand t represents the approaching direction and random size of the time search radius corresponding to the current iteration; W represents the weather influence factor.

[0044] In combination with the first aspect, in a possible implementation, the multi-element satellite sea surface temperature data includes infrared observation data and microwave observation data;

[0045] Taking the preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as the reference data, the other satellite sea surface temperature data is mapped to the reference data to obtain the mapped data, including:

[0046] The infrared observation data and the microwave observation data are unified to a global coordinate system;

[0047] For infrared observation data and microwave observation data of the same physical point, determining the confidence level of the infrared observation data and microwave observation data;

[0048] Using the confidence as the weight of the corresponding observation data, performing weighted averaging on the infrared observation data and the microwave observation data to obtain the mapped data of the same physical point;

[0049] Taking the infrared observation data position as the reference, for the spatially adjacent physical points in the microwave data, the physical point with the smallest distance to the corresponding microwave observation data is determined from the infrared observation data to obtain the matching physical point pair;

[0050] For each pair of matching physical points, determine the confidence level of the corresponding infrared observation data and microwave observation data;

[0051] The confidence level is used as the weight of the corresponding observation data, and the infrared observation data and the microwave observation data of each pair of matching physical points are weighted averaged to obtain the mapped data of the matching physical point pair.

[0052] In a second aspect, a device for determining a search radius for meteorological data fusion is provided, comprising:

[0053] an acquisition module, configured to acquire multi-source satellite sea surface temperature data of a target area and determine weather information when the multi-source satellite sea surface temperature data is collected;

[0054] A weather factor determination module is used to determine weather influencing factors based on the weather information; and

[0055] a data fusion module, configured to use the preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, map other satellite sea surface temperature data to the reference data, and obtain mapped data;

[0056] a boundary determination module, configured to determine a target analysis point from the mapped data and determine a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data;

[0057] The search radius determination module is used to provide the target analysis point position, the boundary range of the spatiotemporal joint search radius and the weather influencing factor to the preset fruit fly algorithm, optimize the spatiotemporal joint search radius, and determine the target spatiotemporal joint search radius.

[0058] In a third aspect, an electronic device is provided, comprising a processor, a memory, and a bus, the memory storing machine readable instructions executable by the processor, when the computer device is running, the processor and the memory communicate through the bus, the machine readable instructions are executed by the processor to perform the steps of the weather data fusion search radius determination method as described in the first aspect or any of the embodiments in combination with the first aspect.

[0059] The beneficial effects of the embodiments of the present disclosure include:

[0060] The embodiments of the present disclosure provide a weather data fusion search radius determination method, device and electronic device, comprising: obtaining multi-source satellite sea surface temperature data of a target area, and determining weather information at the time of collecting the multi-source satellite sea surface temperature data; determining a weather influence factor according to the weather information; taking preset satellite sea surface temperature data in the satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data; determining a target analysis point from the mapped data, and determining a boundary range of a spatio-temporal joint search radius according to a spatio-temporal range of the multi-source satellite sea surface temperature data; providing the boundary range of the spatio-temporal joint search radius and the weather influence factor to a preset fruit fly algorithm, optimizing the spatio-temporal joint search radius, and determining a target spatio-temporal joint search radius. The weather data fusion search radius determination method provided by the present disclosure considers the influence of time and weather on the search radius, so that the obtained search radius is more accurate, and the information utilization rate is improved. BRIEF DESCRIPTION OF DRAWINGS

[0061] Figure 1 A flowchart of a weather data fusion search radius determination method provided by the embodiments of the present disclosure is provided;

[0062] Figure 2 A structural schematic diagram of a weather data fusion search radius determination device provided by the embodiments of the present disclosure is provided;

[0063] Figure 3 A structural schematic diagram of an electronic device provided by the embodiments of the present disclosure is provided. DETAILED DESCRIPTION

[0064] The embodiments of the present disclosure provide a weather data fusion search radius determination method, device and electronic device, the preferred embodiments of the present disclosure are described below in conjunction with the drawings of the specification, it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure, and are not used to limit the present disclosure. And in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0065] The embodiments of the present disclosure provide a method for generating a quality label of marine data, as Figure 1 Shown, including:

[0066] S101, acquiring multi-source satellite sea surface temperature data for a target area, and determining weather information when the multi-source satellite sea surface temperature data is collected;

[0067] S102, determining a weather influencing factor according to the weather information;

[0068] S103, using preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data;

[0069] S104, determining a target analysis point from the mapped data, and determining a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data;

[0070] S105: providing the boundary range of the spatiotemporal joint search radius and the weather influencing factor to a preset fruit fly algorithm, optimizing the spatiotemporal joint search radius, and determining a target spatiotemporal joint search radius.

[0071] In the disclosed embodiments, multi-source satellite sea surface temperature data can include infrared and microwave observations. Weather conditions have a significant impact on the acquisition, transmission, and interpretation of multi-source satellite data. Different weather conditions directly affect the observation range, accuracy, and reliability of microwave and infrared sensors, necessitating changes in the search radius setting during data fusion. On clear or partly cloudy days, atmospheric transparency is high, and sensors are less susceptible to interference from clouds, precipitation, and other factors, resulting in high-quality observation data and a wide coverage area. Microwave sensors such as radar can effectively detect distant targets (e.g., hundreds of kilometers) on clear days with minimal signal attenuation. Infrared radiation is less absorbed by the atmosphere, clearly capturing temperature differences between the surface and clouds, resulting in high positioning accuracy. Because both types of data are highly accurate and consistent, the search radius can be set to a larger value (e.g., a 50-kilometer radius centered on the target point) to fully utilize the wide-area data and improve fusion efficiency. On overcast or cloudy days, thick clouds may partially block infrared radiation, and microwave signals may be affected by scattering from water vapor and raindrops. For infrared data, surface temperatures in cloud-covered areas are difficult to directly observe, resulting in data gaps or increased noise. For microwave data (such as SAR radar), although microwaves can penetrate clouds, light rain or a humid atmosphere can increase signal scattering, making short-range data more reliable and long-range accuracy less accurate. To avoid the inclusion of low-quality data, the search radius during fusion needs to be reduced (for example, a 20-kilometer radius), focusing on the near-field area where the sensors can still effectively observe, ensuring accurate fusion results. During heavy rainfall or thunderstorms, dense clouds with heavy precipitation and strong convection create extremely strong atmospheric interference. For infrared data, it is virtually impossible to penetrate thick cloud cover, significantly reducing the observation range (only limited information such as cloud top temperature can be obtained). For microwave data, heavy precipitation can cause radar echo clutter (such as the "bright band" effect), potentially distorting short-range data due to clutter and making long-range data unreliable. The search radius during fusion needs to be further reduced to an extremely close range (for example, a 5-kilometer radius), or even using only localized high-quality data (such as the radar reflectivity core area) for fusion to avoid contamination of the results with large-scale low-quality data. Therefore, related techniques do not consider the weather conditions at the time of acquisition of multi-source satellite temperature data, resulting in inaccurate search radiuses.

[0072] In addition, since the collection of multi-source satellite sea surface temperature data has a time dimension, if the time dimension is not taken into account, the amount of calculation during data fusion will be very large. Therefore, the embodiment of the present disclosure introduces a spatiotemporal joint search radius (i.e., the time window for collecting SST data), which includes both a spatial search radius and a time search radius. The time search radius can be used to determine the time range in which observation data points can be taken into account in the time dimension. It defines the time range in which an event or observation point affects density calculation or other analysis in the time dimension. It helps to reduce the amount of calculation while avoiding the inclusion of observation points that are far away in time in the analysis, because these points may have a low temporal correlation with the current analysis point. The time search radius can also help to more accurately capture the local characteristics of events or observation points in time, and avoid being affected by noise data that are far away in time. Therefore, in the spatiotemporal analysis, the embodiment of the present disclosure combines the time search radius with the space search radius to provide a more comprehensive spatiotemporal pattern analysis.

[0073] Furthermore, the Fruit Fly Optimization Algorithm (FOA) itself is a swarm intelligence optimization algorithm for searching for optimal solutions, falling within the category of biomimetic optimization algorithms. By simulating the olfactory and visual behaviors of a fruit fly colony during foraging, it iteratively searches for the optimal solution within the solution space. In the disclosed embodiments, the FOA algorithm dynamically adjusts the optimal solution during iteration to determine the optimal spatiotemporal search radius, thereby improving search efficiency and accuracy.

[0074] In another embodiment provided by the present disclosure, the above step S102 of "determining the weather impact factor according to the weather information" can be implemented as follows:

[0075] Step 1: determining the physical quantities corresponding to the weather types included in the weather information, and the influence coefficient of each weather type on the search radius;

[0076] Step 2: Use the influence coefficient as the weight to determine the weighted average of the physical quantity of each weather type; and

[0077] Step 3: Determine the weather impact factor based on the weighted average value.

[0078] In the disclosed embodiment, the search radius is dynamically adjusted by introducing a weather impact factor W. Physical quantities of weather types may include precipitation, humidity, wind speed, etc. A corresponding weight can be determined based on the degree of impact of each weather type on the accuracy and precision of the spatiotemporal joint search radius.

[0079] Taking the physical quantities of weather types including precipitation P, humidity H, and wind speed V as an example, the weather impact factor W can be expressed as follows:

[0080]

[0081] Among them, k1, k2 and k3 represent the influence coefficients of precipitation P, humidity H and wind speed V on the search radius respectively.

[0082] Assuming P = 50 mm / h, H = 90%, V = 10 m / s, k1 = 0.05, k2 = 0.01, and k3 = 0.1, W is approximately 0.185. The search radius, which is 50 kilometers on a clear day, is only 9.25 kilometers under the above weather conditions. This shows that the search radius shrinks significantly under extreme weather conditions. On clear days, W can be set to 1.

[0083] In another embodiment provided by the present disclosure, the above-mentioned step S105 of "providing the boundary range of the spatiotemporal joint search radius and the weather impact factor to a preset fruit fly algorithm, optimizing the spatiotemporal joint search radius, and determining a target spatiotemporal joint search radius" can be implemented as follows:

[0084] Step 1: Determine the initialization parameters of the preset fruit fly algorithm; the initialization parameters include: fruit fly population size, maximum number of iterations and olfactory search step length; and

[0085] Step 2: Construct an objective function with the goal of minimizing the influence of the spatiotemporal joint search radius on multiple preset factors;

[0086] Step 3: determining an initial spatiotemporal joint search radius for each fruit fly according to the boundary range of the spatiotemporal joint search radius and a uniformly distributed random number formula;

[0087] Step 4: Determine an initial spatiotemporal joint search radius constrained by weather according to the weather influencing factors;

[0088] Step 5: Input the initial spatiotemporal joint search radius after weather constraints into the objective function, determine the objective function value, and determine the optimal solution;

[0089] Step 6. Perform olfactory search and visual search in multiple iterative operations. In each olfactory search iteration, randomly adjust the current spatiotemporal joint search radius of each fruit fly according to the olfactory search step size, and determine whether to update the current spatiotemporal joint search radius according to the objective function; in each visual search iteration, input the current spatiotemporal joint search radius of each fruit fly into the objective function to determine the function value, and determine the optimal solution based on the function value; according to the preset visual search step size attenuation formula, make the spatiotemporal joint search radius corresponding to other fruit flies close to the current optimal solution, and obtain a new current spatiotemporal joint search radius based on the weather influencing factor; until the iteration exit condition is met, the target spatiotemporal joint search radius is obtained.

[0090] In the embodiment of the present disclosure, a spatiotemporal joint search radius is set, that is, a spatial search radius R s (Units are usually degrees °, which can also be converted to length units) and the time search radius R t . Spatial search radius R s Used to filter pixels in multi-source satellite data that are close to the target analysis point in space, with a time search radius of R t (Units can be hours, etc.) It is used to filter observations that are close to the target time (for example, the time when a certain infrared observation data is collected). During implementation, it is necessary to first initialize the preset parameters of the fruit fly algorithm: the size of the fruit fly population, that is, the number of fruit flies N, which can be set between 30 and 50 by balancing the computational efficiency and search coverage; the maximum number of iterations (including olfactory search iterations and visual search iterations). The maximum number of iterations can be one of the termination conditions of the algorithm. In addition, the iteration can be terminated in advance based on the accuracy achieved during the iteration process; the olfactory search step size is used to determine the amplitude of the random variable in the global exploration phase. In addition, it is also necessary to determine the value range R of the initial spatial search radius. s ∈[R s,min , R s,max ], and the value range of the initial time search radius R t ∈[0, R t,max The spatial search radius can be set to match the resolution of the data. For example, if the resolution of the data is 0.25°, the spatial search radius can be set to 0.25° or higher. s,min , can be set to be no less than the spatial resolution of the microwave data, ensuring that at least one microwave pixel is included. s,max , can be set to 3-5 times the actual resolution to avoid including too many noise pixels. t,max It can be set according to the time interval of infrared data, but in order to improve calculation efficiency, it can be limited to plus or minus 2 days (48 hours).

[0091] In addition, when initially generating the spatiotemporal search radius of the fruit fly population, the entire parameter space can be covered as much as possible.

[0092] In another embodiment provided by the present disclosure, the objective function F(R s , R t ):

[0093]

[0094] Among them, R s Characterizes the spatial search radius in the spatiotemporal joint search radius; R t represents the time search radius in the spatiotemporal joint search radius; ω1, ω2, and ω3 represent the weight coefficients of the corresponding items; and W represents the weather influence factor.

[0095] In the embodiments of the present disclosure, the target function is related to the optimization direction of the search radius, and the advantages and disadvantages of the current search radius can be evaluated by using the intermediate iteration data. The target function provided by the embodiments of the present disclosure comprehensively considers the multi-objective optimization of the spatial search radius, the time search radius, and the external constraints (such as weather conditions). The spatial search radius R s , the time search radius R t , and the constraint factor (such as the weather influence degree W) are converted into normalized indexes, and the target function is the weighted sum of the indexes, which balances the search efficiency and the constraint compliance.

[0096] Further, ω1, ω2, and ω3 can be set as dynamic weights in order to adapt to the external environment (such as the reliability of the weather influence factor) in real time. In the case of poor weather conditions, the data timeliness is automatically increased, and ω2 can be increased. In the data sparse area, the spatial search radius weight ω1 can be reduced to allow a larger spatial error to obtain more data.

[0097] In still another embodiment of the present disclosure, the above step 3 “determining the initial spatio-temporal joint search radius for each fruit fly according to the boundary range of the spatio-temporal joint search radius and the uniformly distributed random number formula” can include the following steps:

[0098] The initial spatio-temporal joint search radius is determined by the following formula:

[0099] R s = R s，min + (R s，max -R s，min ) × rand()

[0100] R t = R t，min + (R t，max -R t，min ) × rand()

[0101] wherein R s represents the spatial search radius in the spatio-temporal joint search radius; R t represents the time search radius in the spatio-temporal joint search radius; R s，max represents the maximum value of the spatial search radius; R s，min represents the maximum value of the spatial search radius; R t，max represents the maximum value of the time search radius; R t，min represents the maximum value of the time search radius; and rand() represents a random function.

[0102] In another embodiment provided by the present disclosure, the above step 6 of "in each olfactory search iteration, randomly adjusting the current spatiotemporal joint search radius of each fruit fly according to the olfactory search step size, and determining whether to update the current spatiotemporal joint search radius according to the objective function" can be implemented as follows:

[0103] Step 1: In each olfactory search iteration, for each fruit fly, the current spatiotemporal joint search radius is randomly adjusted according to the olfactory search step size using the following formula:

[0104] R i,snew =(R si +rand s ()×step s )×W

[0105] R i,tnew =(R ti +rand t ()×step t )×W

[0106] Among them, R i,snew Represents the spatial search radius of the i-th fruit fly after this iteration adjustment; R si Represents the spatial search radius of the i-th fruit fly before this iterative adjustment; rand s Characterizes the random perturbation direction and size corresponding to this iteration of the spatial search radius; step s Represents the olfactory search step length for space; R i,tnew Represents the time search radius of the i-th fruit fly after this iteration adjustment; R ti Represents the time search radius of the i-th fruit fly before this iteration adjustment; rand t Characterizes the random perturbation direction and size corresponding to this iteration of the time search radius; step t Characterizes the olfactory search step length for time; W represents the weather impact factor

[0107] Step 2: Input the spatiotemporal joint search radius before and after adjustment in the current iteration into the objective function to obtain the objective function value;

[0108] Step 3: Determine whether to use the adjusted spatiotemporal joint search radius to update the spatiotemporal joint search radius before adjustment according to the objective function value.

[0109] In the embodiment of the present disclosure, multiple iterations may include olfactory search iterations and visual search iterations. These two iterations may be performed sequentially, for example, multiple olfactory searches are performed in a concentrated manner, and then multiple visual searches are performed in a concentrated manner. They may also be performed alternately, for example, a visual search is performed after each olfactory search. There is no limitation here.

[0110] Taking the two iteration sequences as an example, we first perform M olfactory searches in M ​​iterations. Each olfactory search process is carried out according to the following steps:

[0111] Assume that the population size is 3 fruit flies (A, B, C), and the space-time radius pairs are randomly initialized:

[0112] Drosophila A: (R sA =8 meters, R tA =4 seconds)

[0113] Drosophila B: (R sB =12 meters, R tB =6 seconds)

[0114] Drosophila B: (R sC =5 meters, R tC =3 seconds)

[0115] Smell search parameter: step s =2 meters, step t =1 second, step s and step t The average value is [-1, 1].

[0116] In one iteration the radius is first randomly adjusted:

[0117] For each fruit fly i, with the current radius R si Based on, randomly increase or decrease the random number step rand s ()×step s , and after weather constraints, the new spatial search radius R is obtained i,snew , but make sure R i,snew In R s，min and R s，max In the same way, we can get R i,tnew , I will not go into details here.

[0118] Then Drosophila A:R A,snew =(8+1×2)×0.5=5 meters; R A,tnew =(4+1×1)×0.5=2.5 seconds, where step s and step t The value is 1. Similarly, the R of fruit fly B is obtained B,snew =5 meters, R B,tnew = 2.5 seconds, here step s and step t The value is -1, and the R C,snew =3.5m, R C,tnew =2 seconds, here step s and step t The average value is 1.

[0119] Taking Drosophila A as an example, R A,snew 、R A,tnew Substitute the objective function to get the first objective function value, and then R sA 、R tA Substitute into the objective function and get the second objective function value. If the first objective function value is better than the second objective function value, then R tA Update to R A,tnew , as the R of the next iteration tA , R sA Update to R A,snew , as the R of the next iteration sA After multiple iterations of olfactory search, each fruit fly individual will obtain a better R si and R ti .

[0120] In another embodiment provided by the present disclosure, in step 6 above, "making the spatiotemporal joint search radius corresponding to other fruit flies approach the current optimal solution according to a preset visual search step size attenuation formula, and obtaining a new current spatiotemporal joint search radius based on the weather influencing factor" may include the following steps:

[0121] Step 1: For each fruit fly among the other fruit flies, the adjustment step size of the spatiotemporal joint search radius of this iteration is determined according to the following formula;

[0122]

[0123] Among them, step i，s Represents the adjustment step of the search radius of the i-th fruit fly in this iteration space; R si Represents the initial spatial search radius of the i-th fruit fly in this iteration; step i，t Represents the adjustment step of the search radius of the i-th fruit fly in this iteration; R ti represents the initial time search radius of the i-th fruit fly in this iteration; t represents the number of local iterations, and T represents the total number of visual search iterations;

[0124] Step 2: Use the following formula to determine the spatiotemporal joint search radius for this iteration:

[0125] R simove =(R si +step i，s ×rand s ())×W

[0126] R timove =(R ti +step i，t ×rand t ())×W

[0127] Among them, R simove Represents the spatial search radius of the i-th fruit fly after it approaches the optimal solution of the current spatial search radius; rand s Characterizes the approach direction and random size of the spatial search radius corresponding to this iteration; R timove Represents the time search radius after the i-th fruit fly approaches the optimal solution of the current time search radius; rand t Represents the approach direction and random size corresponding to the time search radius of this iteration; W represents the weather impact factor.

[0128] In the disclosed embodiment, after the olfactory search iterations have yielded an optimal spatiotemporal search radius for each individual fruit fly, the optimal solution that optimizes the target parameters for each individual fruit fly can be determined. In subsequent visual search iterations, other fruit flies can be directed toward the optimal solution, and new optimal solutions are continuously updated as they approach, until the iteration termination condition is triggered or the algorithm converges prematurely.

[0129] Assume that in the first visual search iteration (assuming the total number of visual search iterations is 3), the current optimal solution comes from fruit fly A, and the corresponding spatiotemporal joint search radius of the optimal solution is (8 meters, 6 seconds). Fruit fly B's current joint search radius is (10 meters, 5 seconds), and fruit fly C's current joint search radius is (7 meters, 4 seconds). Fruit flies B and C will adjust closer to fruit fly A.

[0130] In this iteration, the optimal solution of fruit fly A's spatiotemporal joint search radius remains unchanged. B，s =8×1 / 3=2.67, rounded to 3, step B，t =5×1 / 3=1.67, rounded to 2, then the spatiotemporal joint search radius R of fruit fly B after this iteration adjustment sBmove =(10+3×-1)×0.5=3.5 meters, R tBmove =(5+2×-1)×0.5=1.5 seconds. It should be noted that rand s and rand t The value here can be close to the direction. If the current space-time joint search radius is larger than the optimal solution, rand s and rand t The value can be negative; otherwise, it can be positive to approach the optimal solution. The same applies to fruit fly C and will not be further explained here. After fruit flies B and C have completed the adjustment of the spatiotemporal joint search radius, the target values ​​of fruit flies A, B, and C for the objective function can be re-determined. The target values ​​are used to determine the optimal solution for this iteration, and the next iteration can be entered.

[0131] It should be noted that if the preset number of iterations finds that the optimal solution has no obvious change, it can be considered that the optimal solution has been obtained, and it can also be suspected that a local optimal solution is entered, at this time, the step size can be increased or the amplitude of the decrease can be increased, and if the optimal solution obtained after the step size is increased or the amplitude of the decrease is increased still has no obvious change, it can be considered that the optimal solution has been obtained.

[0132] In yet another embodiment of the present disclosure, the multi-element satellite sea surface temperature data includes infrared observation data and microwave observation data.

[0133] The step S103 "taking preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data" can include the following steps:

[0134] Step one, unify the infrared observation data and the microwave observation data to a global coordinate system;

[0135] Step two, for infrared observation data and microwave observation data of the same physical point, determine the confidence of the infrared observation data and the microwave observation data;

[0136] Step three, taking the confidence as the weight of the corresponding observation data, weighted average of the infrared observation data and the microwave observation data to obtain the mapped data of the same physical point;

[0137] Step four, taking the infrared observation data position as the reference, for the spatially adjacent physical points in the microwave data, determining the physical point with the minimum distance from the corresponding microwave observation data from the infrared observation data to obtain a matched physical point pair;

[0138] Step five, for each matched physical point pair, determine the confidence of the corresponding infrared observation data and microwave observation data;

[0139] Step six, taking the confidence as the weight of the corresponding observation data, weighted average of the infrared observation data and the microwave observation data of each matched physical point pair to obtain the mapped data of the matched physical point pair.

[0140] In the embodiment of the present disclosure, the microwave sensor (such as radar) usually takes itself as the origin to establish a polar coordinate system where R is the distance, and θ is the azimuth angle, is the pitch angle. Infrared sensors (such as thermal imagers) establish an image pixel coordinate system (u, v) with themselves as the origin, which must be converted to a three-dimensional world coordinate system (X, Y, Z) through camera calibration. The aforementioned polar coordinate system and image pixel coordinate system can be aligned to a global coordinate system. Using sensor extrinsics (such as the rotation matrix R and translation vector T), microwave and infrared observation data can be converted to coordinates in the same global coordinate system (such as the Cartesian coordinate system \(O-XYZ\)). This will not be discussed in detail here.

[0141] Furthermore, the confidence of microwave observation data and infrared observation data can be determined. For example, for microwave observation data, the confidence c1 is calculated based on the echo intensity and signal-to-noise ratio (SNR) (e.g., c1 = SNR / SNR max ); For infrared observation data, the confidence level c2 is calculated based on the temperature gradient stability and image contrast (e.g. c2 = ΔT\ΔT max (ΔT is the temperature difference between the target and the background).

[0142] Furthermore, infrared observation data can be used as reference data, and for microwave observation data and infrared observation data whose coordinates can coincide, they can be directly mapped by using confidence as a weighted value. For microwave observation data where there is no completely overlapping data in the infrared observation data, the minimum distance value can be determined from the infrared observation data, and then mapped by using confidence as a weighted value.

[0143] In the disclosed embodiments, the target analysis point can be an observation point of interest in the mapped data. Multiple target analysis points can be selected, and a target spatiotemporal joint search radius can be determined for each target analysis point. The resulting target spatiotemporal joint search radius can be used to perform data fusion for the corresponding target analysis point.

[0144] Based on the same disclosed concept, the embodiments of the present disclosure also provide a device and electronic device for determining the search radius of meteorological data fusion. Since the principles of the problems solved by these devices and electronic devices are similar to the aforementioned method for determining the search radius of meteorological data fusion, the implementation of the device and equipment can refer to the implementation of the aforementioned method, and the repeated parts will not be repeated.

[0145] The embodiment of the present disclosure provides a device for determining a search radius for meteorological data fusion, such as Figure 2 Shown, including:

[0146] An acquisition module 201 is configured to acquire multi-source satellite sea surface temperature data of a target area and determine weather information when the multi-source satellite sea surface temperature data is collected;

[0147] A weather factor determination module 202 is configured to determine a weather influencing factor based on the weather information; and

[0148] The data fusion module 203 is configured to use the preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as the reference data, and map the other satellite sea surface temperature data to the reference data to obtain mapped data;

[0149] a boundary determination module 204 for determining a target analysis point from the mapped data and determining a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data;

[0150] The search radius determination module 205 is used to provide the target analysis point position, the boundary range of the spatiotemporal joint search radius and the weather impact factor to the preset fruit fly algorithm, optimize the spatiotemporal joint search radius, and determine the target spatiotemporal joint search radius.

[0151] Based on the same technical concept, the embodiment of the present application also provides an electronic device. Figure 3 30. The figure shows a schematic diagram of the structure of an electronic device 300 provided in an embodiment of the present application, including a processor 301, a memory 302, and a bus 303. The memory 302 is used to store execution instructions, including a memory 3021 and an external memory 3022. The memory 3021 is also called an internal memory, which is used to temporarily store the operation data in the processor 301 and the data exchanged with the external memory 3022 such as a hard disk. The processor 301 exchanges data with the external memory 3022 through the memory 3021. When the electronic device 300 is running, the processor 301 communicates with the memory 302 through the bus 303, so that the processor 301 executes the following instructions:

[0152] Acquiring multi-source satellite sea surface temperature data for a target area and determining weather information when the multi-source satellite sea surface temperature data was collected;

[0153] Determining weather influencing factors based on the weather information; and

[0154] Using preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data;

[0155] Determining a target analysis point from the mapped data, and determining a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data;

[0156] The boundary range of the spatiotemporal joint search radius and the weather influencing factor are provided to a preset fruit fly algorithm, the spatiotemporal joint search radius is optimized, and a target spatiotemporal joint search radius is determined.

[0157] The present disclosure also provides a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program executes the steps of the method for determining the search radius for meteorological data fusion described in the above method embodiment. The storage medium may be a volatile or non-volatile computer-readable storage medium.

[0158] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments of the present disclosure can be implemented through hardware or by means of software plus a necessary general hardware platform. Based on this understanding, the technical solutions of the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present disclosure.

[0159] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present disclosure.

[0160] Those skilled in the art will appreciate that the modules in the devices of the embodiments may be distributed in the devices of the embodiments as described in the embodiments, or may be located in one or more devices different from the embodiments with corresponding changes. The modules of the above embodiments may be combined into one module or further split into multiple submodules.

[0161] The serial numbers of the above-mentioned embodiments of the present disclosure are for description only and do not represent the advantages or disadvantages of the embodiments.

[0162] Obviously, those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include these modifications and variations.

Claims

1. A method for determining a search radius for meteorological data fusion, characterized in that: include: Acquiring multi-source satellite sea surface temperature data for a target area and determining weather information when the multi-source satellite sea surface temperature data was collected; determining a weather influencing factor based on the weather information; and Using preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data; Determining a target analysis point from the mapped data, and determining a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data; The boundary range of the spatiotemporal joint search radius and the weather influencing factor are provided to a preset fruit fly algorithm, the spatiotemporal joint search radius is optimized, and a target spatiotemporal joint search radius is determined.

2. The method according to claim 1, characterized in that Determining weather influencing factors based on the weather information includes: Determining physical quantities corresponding to weather types included in the weather information, and an influence coefficient of each weather type on the search radius; Use the influence coefficient as a weight to determine the weighted average of the physical quantities of each weather type; and A weather impact factor is determined based on the weighted average.

3. The method according to claim 1, characterized in that Providing the boundary range of the spatiotemporal joint search radius and the weather influencing factor to a preset fruit fly algorithm, optimizing the spatiotemporal joint search radius, and determining a target spatiotemporal joint search radius, including: Determine the initialization parameters of the preset fruit fly algorithm; the initialization parameters include: fruit fly population size, maximum number of iterations, and olfactory search step length; and The objective function is constructed with the goal of minimizing the influence of the spatiotemporal joint search radius on multiple preset factors; Determining an initial spatiotemporal joint search radius for each fruit fly according to the boundary range of the spatiotemporal joint search radius and a uniformly distributed random number formula; Determining an initial spatiotemporal joint search radius constrained by weather according to the weather influencing factors; Inputting the initial spatiotemporal joint search radius after weather constraints into the objective function, determining the objective function value, and determining the optimal solution; Olfactory search and visual search are performed in multiple iterative operations. In each olfactory search iteration, the current spatiotemporal joint search radius of each fruit fly is randomly adjusted according to the olfactory search step size, and whether the current spatiotemporal joint search radius is updated is determined according to the objective function; in each visual search iteration, the current spatiotemporal joint search radius of each fruit fly is input into the objective function to determine the function value, and the optimal solution is determined based on the function value; according to the preset visual search step size attenuation formula, the spatiotemporal joint search radii corresponding to other fruit flies are made close to the current optimal solution, and a new current spatiotemporal joint search radius is obtained based on the weather influencing factor; until the iteration exit condition is met, the target spatiotemporal joint search radius is obtained.

4. The method according to claim 3, characterized in that The objective function is constructed using the following formula: Among them, R s Characterizes the spatial search radius in the spatiotemporal joint search radius; R t represents the time search radius in the spatiotemporal joint search radius; ω1, ω2, and ω3 represent the weight coefficients of the corresponding items; and W represents the weather influence factor.

5. The method according to claim 3, characterized in that According to the boundary range of the spatiotemporal joint search radius and the uniformly distributed random number formula, determining the initial spatiotemporal joint search radius for each fruit fly includes: The initial space-time joint search radius is determined using the following formula: R s =R s，min +(R s，max -R s，min )×rand() R t =R t，min +(R t，max -R t，min )×rand() Among them, R s Characterizes the spatial search radius in the spatiotemporal joint search radius; R t Characterizes the time search radius in the space-time joint search radius; R s，max Characterizes the maximum value of the spatial search radius; R s，min Characterizes the maximum value of the spatial search radius; R t，max Characterizes the maximum value of the time search radius; R t，min Represents the maximum value of the time search radius; rand() represents the random function.

6. The method according to claim 3, characterized in that In each olfactory search iteration, the current spatiotemporal joint search radius of each fruit fly is randomly adjusted according to the olfactory search step size, and whether to update the current spatiotemporal joint search radius is determined according to the objective function, including: In each olfactory search iteration, for each fruit fly, the current spatiotemporal joint search radius is randomly adjusted according to the olfactory search step size using the following formula: R i,snew =(R si +rand s ()×step s )×W R t,snew =(R ti +rand t ()×step t )×W Among them, R i,snew Represents the spatial search radius of the i-th fruit fly after this iteration adjustment; R si Represents the spatial search radius of the i-th fruit fly before this iterative adjustment; rand s Characterizes the random perturbation direction and size corresponding to this iteration of the spatial search radius; step s Represents the olfactory search step length for space; R t,snew Represents the time search radius of the i-th fruit fly after this iteration adjustment; R ti Represents the time search radius of the i-th fruit fly before this iteration adjustment; rand t Characterizes the random perturbation direction and size corresponding to this iteration of the time search radius; step t Characterizes the olfactory search step length for time; W represents the weather impact factor Inputting the spatiotemporal joint search radius before and after adjustment in the current iteration into the objective function respectively to obtain the objective function value; Whether to use the adjusted spatiotemporal joint search radius to update the spatiotemporal joint search radius before adjustment is determined according to the objective function value.

7. The method according to claim 3, characterized in that According to the preset visual search step attenuation formula, the spatiotemporal joint search radius corresponding to other fruit flies is made close to the current optimal solution, and a new current spatiotemporal joint search radius is obtained based on the weather influencing factors, including: For each fruit fly among the other fruit flies, the adjustment step size of the spatiotemporal joint search radius of this iteration is determined according to the following formula; Among them, step i，s Represents the adjustment step of the search radius of the i-th fruit fly in this iteration space; R si Represents the initial spatial search radius of the i-th fruit fly in this iteration; step i，t Represents the adjustment step of the search radius of the i-th fruit fly in this iteration; R ti represents the initial time search radius of the i-th fruit fly in this iteration; t represents the number of local iterations, and T represents the total number of visual search iterations; The following formula is used to determine the spatiotemporal joint search radius of this iteration: R simove =(R si +step i，s ×rand s ())×W R timove =(R ti +step i，t ×rand t ())×W Among them, R simove Represents the spatial search radius of the i-th fruit fly after it approaches the optimal solution of the current spatial search radius; rand s Characterizes the approach direction and random size of the spatial search radius corresponding to this iteration; R timove Represents the time search radius after the i-th fruit fly approaches the optimal solution of the current time search radius; rand t Represents the approach direction and random size of the time search radius corresponding to this iteration; W represents the weather impact factor.

8. The method according to claim 1, characterized in that The multi-element satellite sea surface temperature data includes: infrared observation data and microwave observation data; Using preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, mapping other satellite sea surface temperature data to the reference data to obtain mapped data, including: Unifying the infrared observation data and the microwave observation data into a global coordinate system; For infrared observation data and microwave observation data of the same physical point, determining the confidence level of the infrared observation data and microwave observation data; Using the confidence as the weight of the corresponding observation data, performing weighted averaging on the infrared observation data and the microwave observation data to obtain the mapped data of the same physical point; Taking the infrared observation data position as the reference, for the spatially adjacent physical points in the microwave data, the physical point with the smallest distance to the corresponding microwave observation data is determined from the infrared observation data to obtain the matching physical point pair; For each pair of matching physical points, determine the confidence level of the corresponding infrared observation data and microwave observation data; The confidence level is used as the weight of the corresponding observation data, and the infrared observation data and the microwave observation data of each pair of matching physical points are weighted averaged to obtain the mapped data of the matching physical point pair.

9. A device for determining a search radius for meteorological data fusion, characterized in that: include: an acquisition module, configured to acquire multi-source satellite sea surface temperature data of a target area and determine weather information when the multi-source satellite sea surface temperature data is collected; A weather factor determination module, configured to determine a weather influencing factor based on the weather information; and a data fusion module, configured to use the preset satellite sea surface temperature data in the multi-source satellite sea surface temperature data as reference data, map other satellite sea surface temperature data to the reference data, and obtain mapped data; a boundary determination module, configured to determine a target analysis point from the mapped data and determine a boundary range of a spatiotemporal joint search radius based on the spatiotemporal range of the multi-source satellite sea surface temperature data; The search radius determination module is used to provide the target analysis point position, the boundary range of the spatiotemporal joint search radius and the weather influencing factor to the preset fruit fly algorithm, optimize the spatiotemporal joint search radius, and determine the target spatiotemporal joint search radius.

10. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor and the memory communicate via the bus. When the machine-readable instructions are executed by the processor, the steps of the method for determining the search radius of meteorological data fusion as described in any one of claims 1 to 8 are performed.

Images