A low-level wind speed integrated prediction system and method
By using an integrated low-level wind field forecasting system, which utilizes multi-source data and advanced network model processing, the problems of large errors and low accuracy in low-level wind speed forecasting have been solved, resulting in more accurate and reliable wind speed forecasts.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-28
- Publication Date
- 2026-03-27
AI Technical Summary
Existing low-altitude wind speed forecasts suffer from large errors and low accuracy, mainly due to differences in observational data, different parameter settings and algorithms in numerical models, and the nonlinear distribution caused by the complexity of atmospheric circulation systems, making it impossible to accurately obtain wind speed forecasts at specific vertical heights.
A low-altitude wind field integrated forecasting system is adopted, including a data acquisition module, a data integration module, a vertical downscaling module, and a horizontal downscaling module. It utilizes intelligent grid forecasting, EC fine grid forecasting, and CMA-MESO prediction data, combined with topographic, humidity, and temperature data. Through gated cyclic units and super-resolution generative adversarial network models, data integration and downscaling are performed to improve the accuracy of wind speed data.
Through systematic processing, the influence of terrain, humidity, temperature, and numerical models on wind speed forecasts is avoided, improving the accuracy and reliability of low-level wind speed forecasts and ensuring the accuracy and reliability of wind speed forecasts.
Smart Images
Figure CN120315070B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of low-altitude wind speed prediction, and in particular to a low-altitude wind speed integrated prediction system and method. BACKGROUND
[0002] Low-altitude wind speed prediction is a process of predicting the distribution of future wind speed in vertical and horizontal ranges. People's life and production are concentrated in the low-altitude range below 3000m. Fine low-altitude wind speed prediction is of great significance to disaster prevention and reduction, aviation services, new energy security and other fields.
[0003] However, there are some challenges and difficulties in existing low-altitude wind speed prediction. Since wind speed prediction relies on a large amount of observation data and numerical model, the acquisition and quality of observation data may differ, the acquisition and quality of low-altitude wind field prediction data may differ, and different parameter settings and algorithms of numerical models will produce different results of low-altitude wind speed prediction, resulting in increased error of low-altitude wind speed prediction results. In addition, due to the complexity of atmospheric circulation system such as terrain, ocean, solar radiation, etc., low-altitude wind speed appears nonlinear distribution, which further affects the accuracy of wind speed prediction, and different height layers are set in different numerical models, which cannot accurately obtain wind speed prediction at a specific vertical height. SUMMARY
[0004] The purpose of the present application is to provide an integrated prediction system and method for low-altitude wind field, which improves the accuracy of low-altitude wind field prediction measurement.
[0005] According to one aspect of the present application, the embodiment of the present application provides an integrated prediction system for low-altitude wind field, characterized in that it comprises:
[0006] a data acquisition module, a data integration module, a vertical downscaling module, a horizontal downscaling module and a data determination module;
[0007] The data acquisition module is connected with the data integration module, the data integration module is connected with the vertical downscaling module, the vertical downscaling module is connected with the horizontal downscaling module, and the horizontal downscaling module is connected with the data determination module.
[0008] The data acquisition module is configured to acquire first wind speed data of a plurality of first preset layers predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO prediction of a target area to be measured, and to acquire terrain, humidity and temperature data of the target area to be measured. The data integration module is configured to correspondingly integrate second wind speed data, first humidity data and first temperature data from the first wind speed data, humidity data and temperature data of the same first preset layer predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO prediction of the target area to be measured, respectively. The vertical downscaling module is configured to divide the target area to be measured into a plurality of second preset layers along a first direction at a preset interval according to the second wind speed data and the terrain, first humidity and first temperature data of the target area to be measured. The horizontal downscaling module is configured to predict third wind speed data in a preset area along a second direction for each second preset layer.
[0009] The data determination module is configured to compare the third wind speed data in the preset area of each second preset layer with preset wind speed data in the preset area of each second preset layer, determine a difference between the third wind speed data in the preset area of each second preset layer and the preset wind speed data in the preset area of each second preset layer, and compare the obtained difference with a preset difference threshold to determine whether the third wind speed data in the preset area of each second preset layer is accurate.
[0010] The first direction is a vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data is wind speed data in the preset area of each second preset layer measured by a ground meteorological station and wind speed data in the preset area of each second preset layer obtained by using a mesoscale numerical model to calculate ECMWF ERA5.
[0011] Further, when the difference is less than or equal to the preset threshold, the data determination module is configured to determine that the third wind speed data in the preset area along the second direction for each second preset layer is accurate. When the difference is greater than the preset threshold, the data determination module is configured to send a calculation signal to the horizontal downscaling module, and the horizontal downscaling module is configured to increase the number of calculations of the third wind speed data in the preset area along the second direction for each second preset layer according to the calculation signal sent by the data determination module.
[0012] Further, the data acquisition module comprises a wind speed data acquisition unit, a terrain acquisition unit, a humidity acquisition unit and a temperature time acquisition unit; the wind speed data acquisition unit, the terrain acquisition unit, the humidity acquisition unit and the temperature time acquisition unit are connected with the vertical scale reduction module respectively; the wind speed data acquisition unit is used for acquiring first wind speed data of a plurality of first preset layers of intelligent grid prediction, EC fine grid prediction and CMA-MESO prediction of a target area to be measured; the terrain acquisition unit is used for acquiring terrain data of the target area to be measured; the humidity acquisition unit is used for acquiring humidity data of each first preset layer of the target area to be measured; and the temperature acquisition unit is used for acquiring temperature data of each first preset layer of the target area to be measured.
[0013] Further, the data integration module is used for integrating first wind speed data, humidity data and temperature data of the same first preset layer respectively predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO by using a gating cycle unit to correspond to second wind speed data, first humidity data and first temperature data;
[0014] The vertical scale reduction module is used for dividing the target area to be measured into a plurality of second preset layers according to the second wind speed data and the terrain, the first humidity and the first temperature data of the target area to be measured, and by using a conditional adversarial generative network model.
[0015] Further, the horizontal scale reduction module is used for determining third wind speed data of a preset area along a second direction of each second preset layer by using a super-resolution generative adversarial network model.
[0016] Further, the low-altitude wind speed integrated prediction system further comprises a data display module and a query module.
[0017] The data display module is connected with the query module and the horizontal scale reduction module respectively; the data display module is used for displaying third wind speed data in a preset area of each second preset layer; and the query module is used for querying a prediction range, a start time, a prediction validity period and a prediction element in the preset area of each second preset layer.
[0018] Further, the preset interval is greater than or equal to 100 meters and less than or equal to 500 meters; and the preset area is less than or equal to 40,000 square meters.
[0019] Further, the horizontal scale reduction module increases the number of calculations by greater than or equal to 1 and less than or equal to 3.
[0020] According to another aspect of the present application, embodiments of the present application provide a low-altitude wind field integrated prediction method, comprising:
[0021] The data acquisition module acquires first wind speed data of a plurality of first preset layers predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO prediction of the target area to be measured, and acquires air temperature, terrain and air temperature data of the target area to be measured;
[0022] The data integration module integrates the first wind speed data, humidity data and air temperature data of the same first preset layer predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO prediction of the target area to be measured into second wind speed data, first humidity data and first air temperature data respectively;
[0023] The vertical downscaling module divides the target area to be measured into a plurality of second preset layers along the first direction at a preset interval according to the second wind speed data and the terrain, first humidity and first air temperature data of the target area to be measured;
[0024] The horizontal downscaling module predicts third wind speed data in a preset area along the second direction of each second preset layer;
[0025] The data determination module compares the third wind speed data in the preset area of each second preset layer with preset wind speed data in the preset area of each second preset layer, determines a difference between the third wind speed data in the preset area of each second preset layer and the preset wind speed data in the preset area of each second preset layer, and compares the obtained difference with a preset difference threshold to determine whether the third wind speed data in the preset area of each second preset layer is accurate;
[0026] The first direction is a vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data is wind speed data in the preset area of each second preset layer measured by a ground meteorological station and an ECMEFERA5.
[0027] Further, the data determination module compares the third wind speed data in the preset area of each second preset layer with preset wind speed data in the preset area of each second preset layer, determines a difference between the third wind speed data in the preset area of each second preset layer and the preset wind speed data in the preset area of each second preset layer, and compares the obtained difference with a preset difference threshold to determine whether the third wind speed data in the preset area of each second preset layer is accurate, which includes:
[0028] When the difference is less than or equal to the preset difference threshold, the data determination module determines that the third wind speed data in the preset area along the second direction of each second preset layer is accurate; when the difference is greater than the preset difference threshold, the data determination module is configured to send a calculation signal to the horizontal downscaling module, and the horizontal downscaling module is configured to increase the number of calculations of the third wind speed data in the preset area along the second direction of each second preset layer according to the calculation signal sent by the data determination module.
[0029] Advantages of the present application:
[0030] In the embodiment of the present application, the data acquisition module is used to acquire the predicted wind speed data of multiple pressure layers of the target area to be measured and the terrain, humidity and temperature data of the target area to be measured, the vertical scale reduction module divides the target area to be measured into multiple preset layers along a first direction at a preset interval according to the predicted wind speed data of multiple pressure layers and the terrain, humidity and temperature data of the target area to be measured, so that the influence of terrain, humidity and temperature data on the prediction data of the target area to be measured is avoided in the process of forecasting low-altitude wind field, and the accuracy and reliability of the prediction data of the target area to be measured are improved, and the horizontal scale reduction module is used to determine the predicted wind speed data of a preset area of each preset layer along a second direction, so that the inaccuracy of the prediction data of the target area to be measured due to the acquisition and quality of observation data is avoided through the data acquisition module, and the influence of parameter setting and algorithm selection of the numerical mode on the prediction data of the target area to be measured is avoided through the vertical scale reduction module and the horizontal scale reduction module, and the accuracy and reliability of the wind field prediction of the target area to be measured are improved. BRIEF DESCRIPTION OF DRAWINGS
[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.
[0032] Figure 1 is a structural schematic diagram of a low-altitude wind speed integrated prediction system provided by the embodiment of the present application;
[0033] Figure 2 is a structural schematic diagram of a data acquisition module provided by the embodiment of the present application;
[0034] Figure 3 is a structural schematic diagram of a gating cycle unit provided by the embodiment of the present application;
[0035] Figure 4 is a flow chart of a low-altitude wind speed integrated prediction method provided by the embodiment of the present application;
[0036] Figure 5 is a flow chart of another low-altitude wind speed integrated prediction method provided by the embodiment of the present application. DETAILED DESCRIPTION
[0037] Clearly, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the protection scope of the present application.
[0038] Figure 1 is a structural schematic diagram of a low-altitude wind field prediction measurement system provided by an embodiment of the present application, referring to Figure 1 The low-altitude wind field prediction measurement system comprises a data acquisition module 10, a data integration module 11, a vertical downscaling module 12, a horizontal downscaling module 13 and a data determination module 14; the data acquisition module 10 is connected with the data integration module 11, the data integration module 11 is connected with the vertical downscaling module 12, the vertical downscaling module 12 is connected with the horizontal downscaling module 13, and the horizontal downscaling module 13 is connected with the data determination module 14; the data acquisition module 10 is used to acquire first wind speed data of a plurality of first preset layers predicted by an intelligent grid prediction, an EC fine grid prediction and a CMA-MESO prediction of a target region to be measured, and to acquire air temperature, terrain and temperature data of the target region to be measured; the data integration module 11 is used to generate second wind speed data of the same first preset layer predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO prediction of the target region to be measured; the vertical downscaling module 12 is used to divide the target region to be measured into a plurality of second preset layers along a first direction at a preset interval according to the second wind speed data and the terrain, humidity and temperature data of the target region to be measured; the horizontal downscaling module 13 is used to predict third wind speed data in a preset region along a second direction of each second preset layer; the data determination module 14 is used to compare the third wind speed data in the preset region of each second preset layer with a wind speed data threshold value in the preset region of each second preset layer, to determine a difference value between the third wind speed data in the preset region of each second preset layer and the preset wind speed data in the preset region of each second preset layer, and to compare the obtained difference value with a preset difference value threshold value to determine whether the third wind speed data in the preset region of each second preset layer is accurate.
[0039] The first direction is a vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data is wind speed data in the preset region of each second preset layer measured by a ground meteorological station and an ECMEFERA5.
[0040] The target area to be measured is an area below 3000 meters above the ground. The first preset layer is the pressure layer within the target area, for example, the pressure layer can be the ground (the pressure on the ground is not fixed), 1000hPa, 975hPa, 950hPa, 925hPa, 900hPa, 875hPa, and 850hPa. The second preset layer consists of multiple height layers dividing the target area at preset intervals. The first wind speed data is the wind speed data on the first preset layer, the second wind speed data is the second wind speed data generated by averaging the first wind speed data from multiple first preset layers, and the third wind speed data is the... The second preset layer contains wind speed data within a preset area; the intelligent grid forecast is a refined numerical forecasting device produced by the China Meteorological Administration using high-resolution numerical forecasting models and numerical calculation techniques; the EC fine grid forecast is a device that uses the fine grid model of the European Centre for Medium-Range Weather Forecasts to make high-precision forecasts of wind speed in the target area; CMA-MESO is a device for forecasting meteorological elements using a high-resolution numerical model provided by the China Meteorological Administration; the preset spacing is greater than or equal to 100 meters and less than or equal to 500 meters, and the preset area refers to the area within 40,000 square meters around a certain point on the second preset layer.
[0041] The preset wind speed data within each second preset layer preset area consists of wind speed data provided by ECMWFERA5 and wind field data of the target area measured by ground meteorological stations. ECMWFERA5 is an atmospheric reanalysis dataset provided by the European Centre for Medium-Range Weather Forecasts. ECMWFERA5 includes multiple low-level wind field data. The wind speed data of the target area measured by ground meteorological stations refers to the wind field data collected by wind field observation stations set up in the target area. These data include temperature, humidity, air pressure, wind direction, and wind speed. Ground meteorological station wind speed data and ECMWFERA5 are important data sources for wind field forecasting and wind field research, and can be used to verify and correct the third wind speed data within each second preset layer along the second direction preset area.
[0042] The data integration module 11 includes a gated recurrent unit (GRU), the vertical downscaling module 12 includes a conditional generative adversarial network (CCAN) model, and the horizontal downscaling module 13 includes a super-resolution generative adversarial network (SRGAN) model.
[0043] In the embodiment of the present application, the data acquisition module 10 is configured to acquire first wind speed data of a plurality of first preset layers of the intelligent grid forecast, the EC fine grid forecast and the CMA-MESO forecast of the target area to be measured, and acquire terrain, humidity and temperature data of the target area to be measured; the vertical scale reduction module 12 is configured to divide the target area to be measured into a plurality of second preset layers along a first direction at a preset interval according to the second wind speed data and the terrain, humidity and temperature data of the target area to be measured; and the horizontal scale reduction module 13 is configured to predict third wind speed data in a preset area along a second direction of each second preset layer. In this way, the influence of terrain, humidity and temperature data on the third wind speed data in the preset area along the second direction of each second preset layer is avoided in the process of forecasting the low-altitude wind field, the accuracy and reliability of the third wind speed data in the preset area of each second preset layer of the target area to be measured are improved, and the influence of the algorithm of the numerical model on the forecast data of the target area to be measured is avoided by setting the data integration module 11, the vertical scale reduction module 12, the horizontal scale reduction module 13 and the data determination module 14, thereby improving the accuracy and reliability of the wind field prediction of the target area to be measured.
[0044] Referring to Figure 1 Further, on the basis of the above-mentioned embodiment, the data determination module 14 is configured to determine that the third wind speed data in the preset area along the second direction of each second preset layer is accurate when the difference is less than or equal to a preset difference threshold value, and the data determination module 14 is configured to send a calculation signal to the horizontal scale reduction module 13 when the difference is greater than the preset difference threshold value, and the horizontal scale reduction module 13 is configured to increase the number of calculations of the third wind speed data in the preset area along the second direction of each second preset layer according to the calculation signal sent by the data determination module 14.
[0045] wherein the preset difference threshold value is greater than or equal to 0.8 and less than or equal to 0.9, and the number of third wind speed data calculations is greater than or equal to 1 and less than or equal to 3.
[0046] Specifically, when the difference is less than or equal to the preset difference threshold value, the data determination module 14 determines that the third wind speed data in the preset area along the second direction of each second preset layer is accurate; and when the difference is greater than the preset difference threshold value, the data determination module 14 sends a calculation signal to the horizontal scale reduction module 13, and the horizontal scale reduction module 13 increases the number of calculations of the third wind speed data in the preset area along the second direction of each second preset layer according to the calculation signal sent by the data determination module 14.
[0047] In the embodiment of the present application, the data determination module 14 is configured to determine the third wind speed data in each second preset layer preset area, and when the difference between the third wind speed data in each second preset layer preset area and the preset wind speed data in the second preset layer preset area is greater than the preset difference threshold, the horizontal downscaling module 13 increases the calculation times of the third wind speed data in each second preset layer preset area along the second direction according to the calculation signal sent by the data determination module 14, thereby realizing multiple operations of the third wind speed data in each preset layer preset area along the second direction and improving the accuracy of the predicted wind speed data.
[0048] Figure 2 is a structural schematic diagram of a low-altitude wind field prediction measurement system provided by the embodiment of the present application, referring to Figure 2 Further, on the basis of the above-mentioned embodiment, the data acquisition module 10 comprises a wind speed data acquisition unit 101, a terrain acquisition unit 102, a humidity acquisition unit 103, and a temperature time acquisition unit; the wind speed data acquisition unit 101, the terrain acquisition unit 102, the humidity acquisition unit 103, and the temperature acquisition unit are respectively connected with the vertical downscaling module 12; the wind speed data acquisition unit 101 is configured to acquire the first wind speed data of a plurality of first preset layers in the intelligent grid prediction, the EC fine grid prediction, and the CMA-MESO prediction of the target area to be measured; the terrain acquisition unit 102 is configured to acquire the terrain data of the target area to be measured; the humidity acquisition unit 103 is configured to acquire the humidity data of each first preset layer of the target area to be measured; and the temperature acquisition unit is configured to acquire the temperature data of each first preset layer of the target area to be measured.
[0049] The wind speed data acquisition unit 101 comprises a wind speed collection sensor, the terrain acquisition unit 102 comprises a terrain measurer, the humidity acquisition unit 103 comprises a humidity collection sensor, and the temperature time acquisition unit 104 comprises a temperature collection sensor, which is configured to collect the temperature of the target area to be measured in real time. The wind speed data acquisition unit 101 is configured to acquire the first wind speed data of a plurality of first preset layers in the intelligent grid prediction, the EC fine grid prediction, and the CMA-MESO prediction of the target area to be measured; the terrain acquisition unit 102 is configured to acquire the terrain data of the target area to be measured, such as altitude, slope, and slope direction data.
[0050] In the embodiment of the present application, the terrain acquisition unit 102 is configured to acquire the terrain data of the target area to be measured, thereby avoiding the influence of the terrain on the wind speed data and improving the accuracy of the third wind speed data prediction of the target area to be measured. The humidity acquisition unit 103 is configured to acquire the humidity data of each first preset layer of the target area to be measured, and the temperature time acquisition unit 104 is configured to acquire the temperature data of each first preset layer of the target area to be measured, thereby avoiding the influence of the temperature and humidity on the third wind speed data and improving the accuracy of the third wind speed data prediction of the target area to be measured.
[0051] Figure 3 is a structural schematic diagram of a gated recurrent unit provided by an embodiment of the present application; see Figure 1 and Figure 3 Further, on the basis of the above-mentioned embodiments, the data integration module 11 is configured to use the gated recurrent unit to integrate the first wind speed data, humidity data and temperature data of the same first preset layer predicted by the intelligent grid prediction of the target area to be measured, the EC fine grid prediction and the CMA-MESO prediction respectively to obtain second wind speed data, first humidity data and first temperature data; the vertical scale reduction module 12 is configured to divide the target area to be measured into a plurality of second preset layers according to the second wind speed data and the terrain, the first humidity and the first temperature data of the target area to be measured, and using a conditional generative adversarial network model.
[0052] As shown in Figure 3 , the gated recurrent unit (GRU) is a variant of a recurrent neural network for processing and modeling sequence data; the GRU mainly consists of an update gate (UG) and a reset gate (RG), the update gate vector zt controls the influence of the state information ht-1 of the previous moment on the state of the current moment, the greater the value of the update gate, the more the state information ht-1 of the previous moment is brought in; the reset gate vector rt is responsible for controlling the degree of ignoring the state information ht-1 of the previous moment, the smaller the value of the reset gate, the more it is ignored. Note the expressions of the first two, update gate and reset gate, * represents two matrix convolutions, and σ represents the sigmoid function. As shown in the data flow, the reset gate vector rt after “reset” is convolved with the previous moment state ht-1, and then the result is spliced with the input xt, and then the data is scaled to the range of -1~1 through the activation function tanh; here the input data xt is included, and the convolution result of the previous moment state is added to the current hidden state, which remembers the state of the current moment through this method; as shown in the update memory stage, the forgetting and memory steps are performed simultaneously, and the same gate zt is used to forget and select memory, (1-zt) * ht-1 represents the original state of the hidden, selectively forgotten, and zt * ht represents the current node information, selectively remembered.
[0053] The GRU is calculated along the arrow flow direction in the figure, the cell at the previous moment outputs information as the input of the cell at the next moment, the activation function used in the block, and the operation (such as vector addition, subtraction and multiplication) performed in each step is in the circle. The black line represents the flow of information. It can be understood that the prediction integration module uses the GRU model for training, which can fuse different types of wind field data to generate a multi-mode integrated model. By evaluating the technical indicators of the model, its performance on the test set can be evaluated and compared with the horizontal downscaling model to verify the technical effect of the module.
[0054] In the embodiment of the application, by setting the generative adversarial network model and the gated recurrent unit, the influence of the parameter setting and the algorithm of the numerical mode on the prediction of the third wind speed data in the preset region along the second direction of each second preset layer of the target region to be measured is avoided.
[0055] Referring to Figure 1 Further, on the basis of the above-mentioned embodiment, the horizontal downscaling module 13 is configured to determine the third wind speed data of each second preset layer along the preset region in the second direction by using the super-resolution generative adversarial network model.
[0056] The super-resolution generative adversarial network model includes a generator and a discriminator. The generator is configured to generate the third wind speed data in the preset region along the second direction of each second preset layer. The discriminator is configured to determine whether the third wind speed data in the preset region of each second preset layer is accurate according to the third wind speed data in the preset region of each second preset layer and the preset wind speed data in the preset region of each second preset layer.
[0057] In the embodiment of the application, the third wind speed data of the second preset layer along the preset region in the second direction is generated by the super-resolution generative adversarial network model, thereby improving the accuracy of the third wind speed data in the preset region of each second preset layer.
[0058] Referring to Figure 1 Further, on the basis of the above-mentioned embodiment, the low-altitude wind speed integrated prediction system further comprises a data display module 15 and a query module 16; the data display module 15 is connected with the query module 16 and the horizontal downscaling module 13; the data display module 15 is configured to display the third wind speed data in the preset region of each second preset layer, and the query module 16 is configured to query the prediction range, the reporting time, the prediction validity and the prediction elements in the preset region of each second preset layer.
[0059] The prediction range is greater than or equal to 100 meters and less than or equal to 500 meters, the prediction time is the time of measuring the wind speed of the second preset layer, the prediction validity is the length of time of the prediction of the second preset layer, and the prediction elements are the wind speed data of the second preset layer.
[0060] In the embodiment of the present application, the data display module 15 is used to display the data, and the query module 16 is used to query the data.
[0061] Further, in the above embodiment, the preset interval is greater than or equal to 100 meters and less than or equal to 500 meters, and the preset area is less than or equal to 40000 square meters.
[0062] If the preset interval is too large, the second preset layer of the target area to be measured is less, and the third wind speed data of each second preset layer of the target area to be measured cannot be accurately predicted. If the preset interval is too small, the calculation resource is occupied too much, which leads to the failure to accurately complete the calculation. Therefore, the preset interval is greater than or equal to 100 meters and less than or equal to 500 meters, which can accurately predict the third wind speed data of each second preset layer of the target area to be measured, and can also avoid the failure to complete the calculation due to the excessive occupation of the calculation resource.
[0063] If the preset area is too large, the fine prediction requirement cannot be met. Therefore, the preset area is set to be less than or equal to 40000 square meters, which can meet the requirement of predicting the third wind speed data of each second preset layer of the preset area.
[0064] Referring to Figure 1 Further, in the above embodiment, the calculation number of the horizontal downscaling module 13 is greater than or equal to 1 and less than or equal to 3.
[0065] Specifically, if the calculation number is too large, the structure of the horizontal downscaling module 13 may be increased, which leads to the overweight of the second downscaling module. Therefore, the calculation number of the horizontal downscaling module 13 is set to be greater than or equal to 1 and less than or equal to 3, which can avoid the overweight of the horizontal downscaling module 13 and improve the accuracy of the model.
[0066] Figure 4 is a flowchart of a low-altitude wind speed integrated prediction method provided by the embodiment of the present application, as Figure 4 shown, the embodiment of the present application provides a low-altitude wind speed integrated prediction method, which comprises the following steps:
[0067] S110, a data acquisition module acquires first wind speed data of a plurality of first preset layers of intelligent grid prediction, EC fine grid prediction and CMA-MESO prediction of a target area to be measured, and acquires temperature, terrain and temperature data of the target area to be measured.
[0068] S120, a data integration module generates second wind speed data of the same first preset layer of the first wind speed data predicted by the intelligent grid prediction, the EC fine grid prediction and the CMA-MESO of the target area to be measured, respectively.
[0069] S130, the vertical downscaling module divides the target area to be measured into a plurality of second preset layers according to the second wind speed data and the terrain, humidity and temperature data of the target area to be measured along the first direction at a preset interval.
[0070] S140, the horizontal downscaling module predicts third wind speed data in a preset area along the second direction of each second preset layer.
[0071] S150, the data determining module compares the third wind speed data in the preset area of each second preset layer with preset wind speed data in the preset area of each second preset layer, determines the difference between the third wind speed data in the preset area of each second preset layer and the preset wind speed data in the preset area of each second preset layer, and compares the obtained difference with a preset difference threshold to determine whether the third wind speed data in the preset area of each second preset layer is accurate.
[0072] The first direction is a vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data is the wind speed data in the preset area of each second preset layer measured by a ground meteorological station and ECMEFERA5.
[0073] In the embodiment of the application, the data acquisition module is configured to acquire predicted wind speed data of a plurality of pressure layers of the target area to be measured and terrain, humidity and temperature data of the target area to be measured, and the vertical downscaling module is configured to divide the target area to be measured into a plurality of preset layers according to the predicted wind speed data of the plurality of pressure layers and the terrain, humidity and temperature data of the target area to be measured along the first direction at a preset interval, so that the influence of the terrain, humidity and temperature data on the predicted data of the target area to be measured is avoided in the process of forecasting the low-altitude wind field, and the accuracy and reliability of the predicted data of the target area to be measured are improved. The horizontal downscaling module is configured to determine wind speed data of each preset layer along a preset area in the second direction. The data acquisition module avoids inaccurate predicted data of the target area to be measured due to different acquisition and quality of observation data. The vertical downscaling module and the horizontal downscaling module avoid the influence of parameter setting and algorithm selection of the numerical mode on the predicted data of the target area to be measured, and improve the accuracy and reliability of the wind field prediction of the target area to be measured.
[0074] Figure 5 is a flowchart of another low-altitude wind speed integrated forecasting method provided by the embodiment of the application, as shown in Figure 5 Further, on the basis of the above-mentioned embodiment, the low-altitude wind field forecasting method comprises:
[0075] S110, the data acquisition module acquires first wind speed data of a plurality of first preset layers of the target area to be measured in intelligent grid forecasting, EC fine grid forecasting and CMA-MESO prediction, and acquires air temperature, terrain and temperature data of the target area to be measured.
[0076] S120, the data integration module generates second wind speed data from the first wind speed data of the same first preset level predicted by the target area to be measured intelligent grid prediction, EC fine grid prediction and CMA-MESO respectively.
[0077] S130, the vertical downscaling module divides the target area to be measured into a plurality of second preset layers according to the second wind speed data and the terrain, humidity and temperature data of the target area to be measured along the first direction at a preset interval.
[0078] S140, the horizontal downscaling module predicts third wind speed data in a preset area along the second direction of each second preset layer.
[0079] S1501, the data determination module determines that the third wind speed data in the preset area along the second direction of each second preset layer is accurate when the difference is less than or equal to the preset difference threshold.
[0080] S1502, when the difference is greater than the preset difference threshold, the data determination module is used to send a calculation signal to the horizontal downscaling module, and the horizontal downscaling module is used to increase the number of calculations of the third wind speed data in the preset area along the second direction of each second preset layer according to the calculation signal sent by the data determination module.
[0081] In the embodiments of the present application, the data determination module is provided to accurately judge the predicted wind speed data of each preset layer, and when the difference between the predicted wind speed data of each preset layer and the wind speed data of each preset layer is greater than the preset difference threshold, the horizontal downscaling module increases the number of calculations of the wind speed data of each preset layer along the second direction according to the calculation signal sent by the data determination module, which realizes multiple operations of the wind speed data of each preset layer along the second direction and improves the accuracy of the predicted wind speed data.
[0082] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.
Claims
1. A low-altitude wind speed integrated forecasting system, characterized in that, include: The module includes a data acquisition module, a data integration module, a vertical downscaling module, a horizontal downscaling module, and a data determination module. The data acquisition module is connected to the data integration module, the data integration module is connected to the vertical downscaling module, the vertical downscaling module is connected to the horizontal downscaling module, and the horizontal downscaling module is connected to the data determination module. The data acquisition module is used to acquire first wind speed data from multiple first preset layers predicted by the intelligent grid forecast, EC fine grid forecast, and CMA-MESO of the target area to be measured, as well as topographic, humidity, and temperature data of the target area to be measured; the data integration module is used to integrate the first wind speed data, humidity data, and temperature data of the same first preset layer predicted by the intelligent grid forecast, EC fine grid forecast, and CMA-MESO of the target area to be measured into second wind speed data, first humidity data, and first temperature data in a one-to-one correspondence; the vertical downscaling module is used to divide the target area to be measured into multiple second preset layers along a first direction at preset intervals based on the second wind speed data and the topographic, first humidity, and first temperature data of the target area to be measured; The horizontal downscaling module is used to predict the third wind speed data within a preset area along the second direction for each second preset layer. The data determination module is used to compare the third wind speed data in the preset area of each second preset layer with the preset wind speed data in the preset area of each second preset layer, and determine the difference between the third wind speed data in the preset area of each second preset layer and the preset wind speed data in the preset area of each second preset layer. The data determination module is also used to compare the obtained difference with a preset difference threshold to determine whether the third wind speed data in the preset area of each second preset layer is accurate. Wherein, the first direction is the vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data are the wind speed data of each second preset level preset area measured by the ground meteorological station and the wind speed data of each second preset level preset area obtained by using the mesoscale numerical model to calculate ECMWFERA5. The data determination module is used to determine the accuracy of the third wind speed data within a preset area along the second direction for each second preset layer when the difference is less than or equal to a preset difference threshold; when the difference is greater than the preset difference threshold, the data determination module is used to send a calculation signal to the horizontal downscaling module, and the horizontal downscaling module is used to increase the number of calculations of the third wind speed data within the preset area along the second direction for each second preset layer according to the calculation signal sent by the data determination module. The data integration module is used to integrate the first wind speed data, humidity data and temperature data of the same first preset layer predicted by the smart grid forecast, EC fine grid forecast and CMA-MESO respectively into the second wind speed data, first humidity data and first temperature data by using the gated loop unit to integrate the first wind speed data, first humidity data and first temperature data of the target area to be measured into the first wind speed data and first humidity data respectively. The vertical downscaling module is used to divide the target area into multiple second preset layers based on the second wind speed data and the terrain, first humidity and first temperature data of the target area to be measured, and using a conditional adversarial generative network model.
2. The low-altitude wind speed integrated forecasting system according to claim 1, characterized in that, The data acquisition module includes a wind speed data acquisition unit, a terrain acquisition unit, a humidity acquisition unit, and a temperature-time acquisition unit; the wind speed data acquisition unit, terrain acquisition unit, humidity acquisition unit, and temperature acquisition unit are respectively connected to the vertical downscaling module; the wind speed data acquisition unit is used to acquire the first wind speed data of multiple first preset layers of the target area predicted by intelligent grid forecast, EC fine grid forecast, and CMA-MESO; the terrain acquisition unit is used to acquire the terrain data of the target area; the humidity acquisition unit is used to acquire the humidity data of each first preset layer of the target area; and the temperature acquisition unit is used to acquire the temperature data of each first preset layer of the target area.
3. The low-altitude wind speed integrated forecasting system according to claim 1, characterized in that, The horizontal downscaling module is used to determine the third wind speed data of each second preset layer along the second direction preset region using a super-resolution generative adversarial network model.
4. The low-altitude wind speed integrated forecasting system according to claim 1, characterized in that, Also includes: Data display module and query module; The data display module is connected to the query module and the horizontal downscaling module respectively; the data display module is used to display the third wind speed data in each preset area of the second preset layer, and the query module is used to query the forecast range, start time, forecast lead time and forecast elements in each preset area of the second preset layer.
5. The low-altitude wind speed integrated forecasting system according to claim 1, characterized in that, The preset spacing is greater than or equal to 100 meters and less than or equal to 500 meters; the preset area is less than or equal to 40,000 square meters.
6. The low-altitude wind speed integrated forecasting system according to claim 2, characterized in that, The horizontal downscaling module increases the number of calculations by more than or equal to 1 and less than or equal to 3.
7. A low-level wind speed integrated forecasting method, applied to the low-level wind speed integrated forecasting system according to any one of claims 1-6, characterized in that, include: The data acquisition module acquires the first wind speed data of multiple first preset layers of the intelligent grid forecast, EC fine grid forecast and CMA-MESO prediction of the target area to be measured, as well as the air temperature, topography and temperature data of the target area to be measured; The data integration module integrates the first wind speed data, humidity data and temperature data of the same first preset level predicted by the smart grid forecast, EC fine grid forecast and CMA-MESO of the target area to be measured into the second wind speed data, first humidity data and first temperature data in a one-to-one correspondence. The vertical downscaling module divides the target area into multiple second preset layers along the first direction at preset intervals based on the second wind speed data and the terrain, first humidity, and first temperature data of the target area to be measured. The horizontal downscaling module predicts the third wind speed data within a preset area along the second direction for each second preset layer; The data determination module compares the third wind speed data in each second preset layer preset area with the preset wind speed data in each second preset layer preset area to determine the difference between the third wind speed data in each second preset layer preset area and the preset wind speed data in each second preset layer preset area. The data determination module then compares the obtained difference with a preset difference threshold to determine whether the third wind speed data in each second preset layer preset area is accurate. Wherein, the first direction is the vertical direction, the second direction is perpendicular to the first direction, and the preset wind speed data are the wind speed data of each second preset level preset area measured by ground meteorological stations and ECMEFERA5.