Wind speed prediction method and system, and electronic device and medium

Abstract

Provided in the present application are a wind speed prediction method and system, and an electronic device and a medium. The wind speed prediction method comprises: constructing a correction model on the basis of a deep learning model, so as to acquire a mesoscale driving field model; on the basis of the mesoscale driving field model and data information, screening the mesoscale driving field model, so as to acquire a screened mesoscale driving field model; on the basis of the screened mesoscale driving field model, acquiring a mesoscale prediction result; and performing interpolation processing on the mesoscale prediction result, so as to acquire a downscaled prediction result. The wind speed prediction method can effectively improve the efficiency of wind speed prediction, thereby acquiring a more accurate downscaled wind speed prediction result.

Description

Wind speed prediction method, system, electronic equipment and medium Technical Field

[0001] The present application belongs to the technical field of wind speed prediction, and relates to a wind speed prediction method, and in particular to a wind speed prediction method, system, electronic equipment, and medium. Background Art

[0002] Wind power and the power system are closely connected. The volatility and intermittency of wind energy, which can cause fluctuations in wind power output and adversely affect power quality, safe and stable operation, and economic efficiency, must be considered. The importance of wind resource forecasting is self-evident, playing a crucial role throughout the entire lifecycle of a wind farm. Wind resource forecasting is a crucial decision-making tool during the development and construction phase of a wind farm. Accurate wind resource forecasts can reduce damage to wind turbines and components from extreme weather events, enabling wind farm operators to formulate appropriate maintenance plans and prepare for weather disasters. Compared to onshore wind power, offshore wind resources offer superior power generation efficiency, making the development of offshore wind power a crucial area of ​​focus for renewable energy. However, due to limited offshore observation capabilities and significant spatial and temporal limitations, the lack of measured data presents a significant challenge in assessing and predicting offshore wind speeds. Consequently, a highly accurate and efficient offshore wind speed forecasting method is currently lacking.

[0003] Summary of the Invention

[0004] In view of the above-mentioned shortcomings of the prior art, the purpose of this application is to provide a wind speed prediction method, system, electronic equipment and medium to solve the problems of low accuracy and low efficiency of offshore wind speed prediction in the prior art.

[0005] In a first aspect, the present application provides a wind speed prediction method, which includes: constructing a correction model based on a deep learning model to obtain a mesoscale driving field model; screening the mesoscale driving field model according to the mesoscale driving field model and data information to obtain a screened mesoscale driving field model; obtaining a mesoscale prediction result according to the screened mesoscale driving field model; and interpolating the mesoscale prediction result to obtain a downscaled prediction result.

[0006] In this application, a deep learning model correction model is used to obtain a mesoscale driving field model. The mesoscale driving field model is then filtered using data information to obtain a mesoscale driving field model with higher prediction accuracy. The mesoscale prediction results are then obtained based on the filtered mesoscale driving field model. The mesoscale prediction results are then processed to obtain a downscaled prediction result. This method can obtain more accurate downscaled prediction results and improve the efficiency of wind speed prediction.

[0007] In an implementation of the first aspect, the parameters of the loss function of the correction model include the difference in wind speed of the input data at adjacent moments, and the difference in wind speed of the input data at adjacent moments is obtained using a momentum equation.

[0008] In an implementation of the first aspect, the parameters of the loss function further include: input data, model prediction results, and differences in wind speeds at adjacent moments of the model prediction results.

[0009] In an implementation of the first aspect, screening the mesoscale driven field model according to the mesoscale driven field model and the data information includes: obtaining a parameter scheme of the mesoscale driven field model according to the data information; obtaining a simulated wind speed of the mesoscale driven field model using the parameter scheme; and screening according to each of the simulated wind speeds and the observed wind speed to obtain the screened mesoscale driven field model.

[0010] In an implementation of the first aspect, screening according to each of the simulated wind speeds and the observed wind speeds includes: obtaining the root mean square error between the simulated wind speed and the observed wind speed, and obtaining the screened mesoscale driving field model according to the root mean square error with the smallest value.

[0011] In an implementation of the first aspect, interpolating the mesoscale prediction result includes: obtaining multiple interpolation schemes; obtaining the downscaled prediction result using the interpolation schemes; screening according to each of the downscaled prediction results to obtain a screened interpolation scheme; and interpolating the mesoscale prediction result using the screened interpolation scheme to obtain a downscaled prediction result.

[0012] In an implementation of the first aspect, the interpolation scheme includes an inverse distance weighting scheme, a bilinear interpolation scheme, a Kriging interpolation scheme, and a radial basis function scheme.

[0013] In the second aspect, the present application provides a wind speed prediction system, including: a model acquisition module, used to construct a correction model based on a deep learning model to obtain a mesoscale driving field model; a model processing module, used to filter the mesoscale driving field model according to the mesoscale driving field model and data information, and obtain the filtered mesoscale driving field model; a prediction result acquisition module, used to obtain a mesoscale prediction result based on the filtered mesoscale driving field model; a downscaling result acquisition module, used to interpolate the mesoscale prediction result to obtain a downscaled prediction result.

[0014] In a third aspect, the present application provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program stored in the memory, so that the electronic device performs a wind speed prediction method as described in any one of the first aspects.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the wind speed prediction method described in any one of the first aspects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] FIG1 shows a schematic diagram of the structure of the end-cloud interaction scenario in the implementation method described in an embodiment of the present application.

[0017] FIG2 is a schematic flow chart of a wind speed prediction method according to an embodiment of the present application.

[0018] FIG3 is a schematic diagram showing an implementation method of an embodiment of the present application.

[0019] FIG4 is a schematic flow chart showing a wind speed prediction method according to an embodiment of the present application.

[0020] FIG5 is a schematic flow chart showing a wind speed prediction method according to an embodiment of the present application.

[0021] FIG6 is a schematic diagram showing the structure of a wind speed prediction system according to an embodiment of the present application.

[0022] FIG7 is a schematic diagram showing the structure of an electronic device according to an embodiment of the present application.

[0023] Component Reference Numerals 1 End-Cloud Interaction System 10 Terminal 11 Cloud Server 600 Wind Speed ​​Prediction System 610 Model Acquisition Module 620 Model Processing Module 630 Prediction Result Acquisition Module 640 Downscaling Result Acquisition Module 700 Electronic Device 710 Memory 720 Processor 730 Display S11-S14 Steps S121-S123 Steps S141-S143 Steps DETAILED DESCRIPTION

[0024] The following describes the embodiments of the present application through specific examples. Those skilled in the art can easily understand the other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0025] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. Therefore, the illustrations only show components related to the present application and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0026] Numerical weather forecasting is the most common method for wind energy forecasting. For example, the Global Forecast System (GFS) can provide forecast results for major meteorological elements including wind fields around the world in the next few days. However, due to the imperfection of parameterization schemes, resolution limitations and insufficient representation of high-resolution features at the regional scale, this type of global numerical forecast system has relatively poor forecast capabilities for regional small-scale wind fields. The forecast error is particularly significant in offshore areas, which cannot meet the needs of actual development and construction of offshore wind power. Currently, regional high-resolution offshore wind speed forecasts are mostly based on mesoscale numerical models (Weather Research and Forecast, WRF for short). WRF is a numerical model used for meteorological forecasting and research. It is a high-resolution, non-static, non-hierarchical, and non-uniform atmospheric numerical model. The WRF model simulates atmospheric phenomena by dividing the Earth's atmosphere into horizontal grids and vertical layers, and solving atmospheric dynamics, thermodynamics, and hydrological cycle equations through discrete equations. This model is widely used in wind resource simulation and wind speed forecasting. The advantage of the WRF numerical model is that the physical process is clear. Even if there is less observation data, high-resolution regional data can be constructed by characterizing local thermodynamic processes. However, there are also problems such as large errors, difficulty in verifying prediction results, and high time cost.

[0027] At least to address the above-mentioned problems, an embodiment of the present application provides a wind speed prediction method, which includes: constructing a correction model based on a deep learning model to obtain a mesoscale driving field model; screening the mesoscale driving field model according to the mesoscale driving field model and data information to obtain a screened mesoscale driving field model; obtaining a mesoscale prediction result based on the screened mesoscale driving field model; and interpolating the mesoscale prediction result to obtain a downscaled prediction result.

[0028] In this embodiment, a deep learning model calibration model is used to obtain a mesoscale driving field model. This model is then filtered using data information to obtain a mesoscale driving field model with higher prediction accuracy. A mesoscale prediction result is obtained based on the filtered mesoscale driving layer, and the mesoscale prediction result is then processed to obtain a downscaled prediction result. This approach can obtain more accurate downscaled prediction results and improve the efficiency of wind speed prediction.

[0029] The wind speed prediction method described in this application can be applied to end-to-cloud interaction scenarios. Figure 1 shows a schematic diagram of the structure of an end-to-cloud interaction scenario in the implementation method described in an embodiment of this application. As shown in Figure 1, the end-to-cloud interaction system 1 includes a terminal 10 and a cloud server 11. The terminal 10 and the cloud server 11 can communicate with each other, and the communication method is not limited to wired or wireless.

[0030] The terminal 10 may be mobile or fixed. For example, the terminal 10 may be a wireless terminal or a wired terminal. A wireless terminal may refer to a device with wireless transceiver capabilities and may be deployed indoors, outdoors, or in offshore wind farms. The terminal 10 may be a mobile phone, a tablet computer, a laptop computer, or the like, without limitation.

[0031] The cloud server 11 may include one or more servers, or one or more processing nodes, or one or more virtual machines running on the server. The cloud server 11 may also be called a server cluster, a management platform, a data processing center, etc., which is not limited in the embodiments of the present application.

[0032] The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings in the embodiments of the present application.

[0033] The following embodiments of the present application provide a wind speed prediction method, which can be implemented, for example, by the cloud server 11 shown in Figure 1. Figure 2 shows a flow chart of the wind speed prediction method according to an embodiment of the present application. As shown in Figure 2, the wind speed prediction method includes steps S11 to S14.

[0034] Step S11: constructing a calibration model based on the deep learning model to obtain a mesoscale driving field model. The calibration model is obtained by adjusting the parameters of the deep learning model. The mesoscale driving field model (Weather Research and Forecast, WRF) is a numerical model used for weather forecasting and research.

[0035] Step S12: Filter the mesoscale driving field model based on the mesoscale driving field model and the data information to obtain a filtered mesoscale driving field model. The data information includes multi-source data such as satellite remote sensing data, site observation data, and reanalysis data. Optionally, the reanalysis data is a dataset obtained by quality control and assimilation of observation data (including ground observations, satellites, radar, soundings, buoys, aircraft, ships, etc.).

[0036] Step S13: obtaining a mesoscale prediction result based on the filtered mesoscale driving field model, wherein the mesoscale prediction result is a mesoscale wind speed prediction result.

[0037] In some possible implementations, Figure 3 shows a schematic diagram of an implementation of an embodiment of the present application. The screened mesoscale driving field model performs two-layer bidirectional nesting on the target area to obtain the mesoscale prediction results. The parent area in the bidirectional nesting provides boundary conditions for the child area, and the calculation results of the child area will be fed back to the parent area. As shown in Figure 3, the large square is the parent area, the small square is the child area, and the central longitude and latitude of the parent area and the child area are consistent, for example, 11.10° north latitude and -73.52° east longitude. The grid point distance of the parent area is set to 9 kilometers, and the total number of grid points is 75,922, so that the two-layer nesting resolution of the screened mesoscale driving field model is 9 kilometers and 3 kilometers, respectively, thereby obtaining high temporal and spatial resolution (hourly, 3-9 kilometers) prediction results of the target area.

[0038] It should be noted that the above is only one possible implementation method of the embodiment of the present application, and the present application is not limited thereto.

[0039] Step S14: performing interpolation processing on the mesoscale prediction result to obtain a downscaled prediction result.

[0040] In this embodiment, a deep learning model calibration model is used to obtain a mesoscale driving field model. The mesoscale driving field model is then filtered using data information to obtain a mesoscale driving field model with higher prediction accuracy. A mesoscale prediction result is obtained based on the filtered mesoscale driving field model. The mesoscale prediction result is then processed to obtain a downscaled prediction result. This approach can obtain more accurate downscaled prediction results and improve the efficiency of wind speed prediction.

[0041] In one embodiment of the present application, the parameters of the loss function of the correction model include the difference in wind speed between adjacent moments of the input data, and the difference in wind speed between adjacent moments of the input data is obtained using a momentum equation.

[0042] In one embodiment of the present application, the parameters of the loss function also include: input data, model prediction results, and the difference in wind speeds at adjacent moments of the model prediction results.

[0043] In some possible implementations, the input data is reanalysis data. First, the data of the global ensemble forecast system is obtained as the driving data of the mesoscale driving field model. Then, according to the ERA5 (European Centre for Medium-Range Weather Forecasts Reanalysis v5) reanalysis data and the FNL (Finally Opera Global Analysis) reanalysis data, a correction model is constructed based on the deep learning model to obtain the mesoscale driving field model. In the process of constructing the correction model, the momentum equation is combined to perform physical constraints on the wind speed characteristics. Optionally, the input data of the deep learning model is a forecast product data set using two adjacent moments, and the output data of the deep learning model is the model prediction result. The calculation formula of the loss function of the deep learning model is:

[0044] Where g1(x) is the prediction result of the deep learning model, y1 is the prediction result of the input data, g2(x) is the difference in wind speed between two adjacent moments calculated based on the momentum equation, y2 is the difference in wind speed between two adjacent moments calculated based on the prediction result of the deep learning model, i is a grid point, and n is the total number of grid points. The calculation method for the difference in wind speed between two adjacent moments based on the momentum equation is:

[0045] in, is the advection term, is the pressure gradient term, is the Coriolis force term.

[0046] In an embodiment of the present application, the momentum equation is used to construct the loss function of the deep learning model, so that the correction model obtained can have a stronger constraint effect on the correction of low-level wind speed based on the momentum propagation principle of mid- and high-level wind speed and low-level wind speed.

[0047] Figure 4 is a schematic flow chart of the wind speed prediction method according to an embodiment of the present application. As shown in Figure 4 , step S12 includes steps S121 to S123.

[0048] Step S121: Acquire a parameter scheme for the mesoscale driving field model based on data information, wherein the data information includes reanalysis data, satellite remote sensing data, and a small amount of site observation data.

[0049] Step S122: using the parameter scheme to obtain the simulated wind speed of the mesoscale driving field model.

[0050] Step S123 , performing screening according to the simulated wind speeds and the observed wind speeds to obtain the screened mesoscale driving field model.

[0051] In the embodiment of the present application, a multi-source data acquisition parameter solution is adopted. The simulation accuracy of the mesoscale driving field model obtained in this way is higher, and the simulated wind speed obtained is more accurate.

[0052] In one embodiment of the present application, screening according to each of the simulated wind speeds and the observed wind speeds includes obtaining the root mean square error between the simulated wind speed and the observed wind speed, and obtaining the screened mesoscale driving field model according to the root mean square error with the smallest value.

[0053] In some possible implementations, a parameter scheme for the mesoscale driven field model is obtained based on the data information, a simulated wind speed of the mesoscale driven field model is obtained based on the parameter scheme, and the simulated wind speeds are screened based on the observed wind speeds to obtain the screened mesoscale driven field model. Optionally, the root mean square error (RMSE) between the simulated wind speed and the observed wind speed is calculated by comparing the simulated wind speed with the observed wind speed. The formula for calculating the RMS SE between the simulated wind speed and the observed wind speed is:

[0054] Where M is the output result of the mesoscale driving field model, Station is the site result, Remote is the remote sensing data, Reanalysis is the reanalysis data, i is the i-th grid point, and a, b, and c are the weights of the three root mean square errors, respectively.

[0055] Optionally, the weights a, b, and c in the formula for calculating the root mean square error are 0.6, 0.2, and 0.3, respectively. The filtered mesoscale driving field model is obtained based on the parameter solution with the smallest root mean square error value. A mesoscale prediction result is obtained based on the mesoscale driving field model.

[0056] Figure 5 is a flow chart of the wind speed prediction method according to an embodiment of the present application. As shown in Figure 5 , step S14 includes the following steps S141 to S144 .

[0057] Step S141: Acquire multiple interpolation schemes.

[0058] Step S142: Obtain the downscaled prediction result using the interpolation scheme.

[0059] Step S143 , performing screening according to the downscaled prediction results to obtain a screened interpolation scheme.

[0060] Step S144: Interpolate the mesoscale prediction result using the filtered interpolation scheme to obtain a downscaled prediction result. Optionally, the mesoscale prediction result is a wind speed prediction result with an accuracy of 3 to 9 kilometers, while the downscaled prediction result is a wind speed prediction result with an accuracy of 1 kilometer. The downscaled prediction result at the 1 kilometer scale has higher prediction accuracy than the mesoscale prediction result at the 3 to 9 kilometer scale.

[0061] In some possible implementations, multiple interpolation schemes are obtained, including an inverse distance weighting scheme, a bilinear interpolation scheme, a Kriging interpolation scheme, and a radial basis function scheme, but the present application is not limited thereto. Each of the interpolation schemes is used to perform spatial interpolation on the area where the downscaled prediction results need to be obtained. Specifically, the mesoscale prediction results of 3 to 9 kilometers are interpolated to a 1 kilometer scale, and the root mean square error between the results obtained by each interpolation scheme and the actual observed wind speed of the mesoscale driving field model is calculated. The interpolation scheme with the smallest root mean square error value is screened and obtained as the screened interpolation scheme. The calculation formula for the root mean square error is:

[0062] Where Intep is the interpolated result, WRF_1km is the simulated wind speed result of the mesoscale driving field model WRF at a scale of 1 km, i is the i-th grid point, and n is the total number of grid points.

[0063] In the embodiment of the present application, the actual observed wind speed at the kilometer level of the mesoscale driving field model is compared with the simulated wind speed obtained by multiple interpolation schemes to obtain a screened interpolation scheme, thereby achieving rapid encryption of the area where the downscaled prediction results need to be obtained.

[0064] FIG6 is a schematic diagram of the structure of a wind speed prediction system according to an embodiment of the present application. As shown in FIG6 , the wind speed prediction system 600 includes a model acquisition module 610 , a model processing module 620 , a prediction result acquisition module 630 , and a downscaling result acquisition module 640 .

[0065] The model acquisition module 610 is used to construct a correction model based on the deep learning model to obtain a mesoscale driving field model.

[0066] The model processing module 620 is used to filter the mesoscale driving field model according to the mesoscale driving field model and data information to obtain the filtered mesoscale driving field model.

[0067] The prediction result acquisition module 630 is used to obtain the mesoscale prediction result according to the filtered mesoscale driving field model.

[0068] The downscaling result acquisition module 640 is configured to perform interpolation processing on the mesoscale prediction result to obtain a downscaled prediction result.

[0069] It should be noted that the modules 610 to 640 included in the wind speed prediction system 600 correspond one-to-one to steps S11 to S14 in the wind speed prediction method shown in FIG. 2 , and are not described in detail here.

[0070] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices or methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of modules / units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules or units, which can be electrical, mechanical or other forms.

[0071] The modules / units described as separate components may or may not be physically separate, and the components displayed as modules / units may or may not be physical modules, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules / units may be selected according to actual needs to achieve the purpose of the embodiments of the present application. For example, the functional modules / units in the various embodiments of the present application may be integrated into a processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into a single module / unit.

[0072] Those skilled in the art should further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0073] The present invention also provides an electronic device. FIG7 shows a schematic diagram of the structure of an electronic device 700 according to an embodiment of the present invention. As shown in FIG7 , the electronic device 700 in this embodiment includes a memory 710 and a processor 720 .

[0074] The memory 710 is used to store computer programs; preferably, the memory 710 includes: ROM, RAM, magnetic disk, USB flash drive, memory card or optical disk, etc., various media that can store program codes.

[0075] Specifically, the memory 710 may include a computer system readable medium in the form of a volatile memory, such as a random access memory (RAM) and / or a cache memory. The electronic device 700 may further include other removable / non-removable, volatile / non-volatile computer system storage media. The memory 710 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present application.

[0076] The processor 720 is connected to the memory 710 and is used to execute the computer program stored in the memory 710 so that the electronic device 700 executes the wind speed prediction method described in any embodiment of the present application.

[0077] Optionally, the processor 720 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0078] Optionally, the electronic device 700 in this embodiment may further include a display 730. The display 730 is communicatively connected to the memory 710 and the processor 720, and is used to display a graphical user interface (GUI) interactive interface related to the wind speed prediction method described in the embodiment of the present application.

[0079] The present application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the wind speed prediction method described in any embodiment of the present application.

[0080] The descriptions of the processes or structures corresponding to the above figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0081] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical concepts disclosed in this application shall be covered by the claims of this application.

Claims

1. A wind speed prediction method, characterized in that: include: Build a correction model based on the deep learning model to obtain the mesoscale driving field model; screening the mesoscale driving field model according to the mesoscale driving field model and the data information to obtain a screened mesoscale driving field model; Obtaining a mesoscale prediction result according to the screened mesoscale driving field model; Interpolation processing is performed on the mesoscale prediction result to obtain a downscaled prediction result.

2. The wind speed prediction method according to claim 1, wherein: The parameters of the loss function of the correction model include the difference in wind speed between adjacent moments of the input data, and the difference in wind speed between adjacent moments of the input data is obtained using a momentum equation.

3. The wind speed prediction method according to claim 2, characterized in that: The parameters of the loss function also include: input data, model prediction results, and the difference in wind speed between adjacent moments of the model prediction results.

4. The wind speed prediction method according to claim 1, wherein: Screening the mesoscale driving field model according to the mesoscale driving field model and the data information includes: Acquiring a parameter scheme of the mesoscale driving field model according to the data information; obtaining a simulated wind speed of the mesoscale driving field model using the parameter scheme; Screening is performed based on the simulated wind speeds and the observed wind speeds to obtain the screened mesoscale driving field model.

5. The wind speed prediction method according to claim 4, characterized in that: Screening according to the simulated wind speeds and the observed wind speeds includes: obtaining a root mean square error between the simulated wind speeds and the observed wind speeds, and obtaining the screened mesoscale driving field model according to the root mean square error with the smallest value.

6. The wind speed prediction method according to claim 1, characterized in that: Performing interpolation processing on the mesoscale prediction result includes: Get multiple interpolation schemes; Obtaining the downscaled prediction result using the interpolation scheme; Screening is performed according to the downscaled prediction results to obtain a screened interpolation scheme; The filtered interpolation scheme is used to perform interpolation processing on the mesoscale prediction result to obtain a downscaled prediction result.

7. The wind speed prediction method according to claim 6, characterized in that: The interpolation schemes include an inverse distance weighting scheme, a bilinear interpolation scheme, a Kriging interpolation scheme and a radial basis function scheme.

8. A wind speed prediction system, characterized in that: include: The model acquisition module is used to build a correction model based on the deep learning model to obtain the mesoscale driving field model; a model processing module, configured to filter the mesoscale driving field model according to the mesoscale driving field model and data information, and obtain a filtered mesoscale driving field model; A prediction result acquisition module, configured to acquire a mesoscale prediction result based on the filtered mesoscale driving field model; The downscaling result acquisition module is used to perform interpolation processing on the mesoscale prediction result to obtain a downscaled prediction result.

9. An electronic device, characterized in that: The electronic device comprises: Memory for storing computer programs; A processor, wherein the processor is configured to execute the computer program stored in the memory so as to enable the electronic device to perform the wind speed prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the wind speed prediction method according to any one of claims 1 to 7 is implemented.

Images