Wind power prediction method, device, equipment, storage medium and product

Abstract

The application discloses a wind power prediction method, device, equipment, storage medium and product, relates to the technical field of wind power prediction, and the method comprises the steps of converting numerical weather forecast data in a preset area into a preset grid superposition format to obtain formatted regional numerical weather forecast data; combining preset deep learning network components to construct a deep learning wind power prediction model according to the best component combination; inputting the formatted regional numerical weather forecast data into the deep learning wind power prediction model to perform wind power prediction and obtaining a wind power prediction result. The application can comprehensively reflect the meteorological characteristics in the region by grid superposition formatting of regional meteorological data; and the prediction model can be constructed by combining deep learning network components, the total power generation of all wind farms in the region can be predicted by a single model, and the consumption of human resources and computing resources is low.

Description

Technical Field

[0001] This application relates to the field of wind power prediction technology, and in particular to a wind power prediction method, device, equipment, storage medium and product. Background Technology

[0002] Existing short-term wind power forecasting techniques typically rely solely on numerical weather forecast data from a single latitude and longitude point. Such data cannot fully reflect the meteorological factors influencing wind farm power generation. For short-term wind power forecasting of a region, the power generation of each wind farm is usually predicted separately and then summed to obtain the total predicted power for all wind farms in the region. However, this summation method for regional wind power forecasting requires building a model for each wind farm, resulting in multiple models and significant human and computational resources. Summary of the Invention

[0003] The main purpose of this application is to provide a wind power prediction method, device, equipment, storage medium and product, which aims to solve the technical problem that the existing method of regional wind power prediction requires a lot of human and computing resources.

[0004] To achieve the above objectives, this application proposes a wind power prediction method, which includes:

[0005] The acquired numerical weather forecast data within the preset area is converted into a preset grid overlay format to obtain formatted regional numerical weather forecast data.

[0006] The preset deep learning network components are combined, and a deep learning wind power prediction model is constructed based on the optimal component combination.

[0007] The formatted regional numerical weather forecast data is input into the deep learning wind power prediction model to predict wind power and obtain the wind power prediction results.

[0008] In one embodiment, the step of converting the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data includes:

[0009] The numerical weather forecast data is gridded based on the longitude, latitude and time information corresponding to the numerical weather forecast data within the preset area to obtain the gridded regional numerical weather forecast data.

[0010] When the gridded regional numerical weather forecast data includes multi-dimensional parameters, multi-parameter overlay processing is performed on the multi-dimensional parameters in the gridded regional numerical weather forecast data to obtain formatted regional numerical weather forecast data.

[0011] In one embodiment, the preset deep learning network components include convolutional layers, pooling layers, activation layers, and fully connected layers;

[0012] The step of combining preset deep learning network components and constructing a deep learning wind power prediction model based on the optimal component combination includes:

[0013] By combining convolutional layers, pooling layers, activation layers, and fully connected layers in various ways, multiple different component combinations can be obtained;

[0014] The performance of the multiple different component combinations is evaluated using cross-validation to determine the optimal component combination.

[0015] A deep learning wind power prediction model is constructed based on the optimal component combination.

[0016] In one embodiment, the multiple combination methods include sequential connection, superimposed connection and / or short-circuit connection.

[0017] In one embodiment, the step of evaluating the performance of the plurality of different component combinations according to the cross-validation method and determining the optimal component combination includes:

[0018] The performance of the multiple different component combinations is evaluated using cross-validation to obtain performance evaluation results.

[0019] The optimal component combination is determined from the multiple different component combinations based on the preset performance indicators of the performance evaluation results. The preset performance indicators include the correlation between the predicted value and the actual value, root mean square error, normalized root mean square error, mean absolute error, and normalized mean absolute error.

[0020] In one embodiment, the step of inputting the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain wind power prediction results includes:

[0021] The formatted regional numerical weather forecast data is input into the deep learning wind power prediction model;

[0022] Feature data is obtained by performing convolution operations on the formatted regional numerical weather forecast data through the convolutional layer of the deep learning wind power prediction model.

[0023] The feature data is sampled by the pooling layer of the deep learning wind power prediction model to obtain the sampled feature data.

[0024] The activated feature data is obtained by activating the sampled feature data through the activation layer of the deep learning wind power prediction model.

[0025] The wind power prediction result is obtained by mapping the activated feature data through the fully connected layer of the deep learning wind power prediction model.

[0026] Furthermore, to achieve the above objectives, this application also proposes a wind power prediction device, which includes:

[0027] The data formatting module is used to convert the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data.

[0028] The prediction model building module is used to combine preset deep learning network components and build a deep learning wind power prediction model based on the optimal component combination.

[0029] The wind power prediction module is used to input the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain wind power prediction results.

[0030] In addition, to achieve the above objectives, this application also proposes a wind power prediction device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the wind power prediction method as described above.

[0031] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the wind power prediction method described above.

[0032] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the wind power prediction method described above.

[0033] This application provides a wind power prediction method. It involves converting acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data. Preset deep learning network components are combined, and a deep learning wind power prediction model is constructed based on the optimal component combination. The formatted regional numerical weather forecast data is then input into the deep learning wind power prediction model to predict wind power, yielding the prediction results. This application, by formatting regional meteorological data through grid overlay, can comprehensively reflect the meteorological characteristics of the region. By selecting the optimal component combination through the combination of preset deep learning network components, a deep learning wind power prediction model can be constructed. This model can deeply extract features from large-scale regional meteorological data, accurately reflect regional meteorological characteristics, avoid model overfitting, and enable a single model to predict the total power generation of all wind farms within the region, achieving low consumption of human and computational resources. Attached Figure Description

[0034] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a flowchart illustrating an embodiment of the wind power prediction method of this application.

[0037] Figure 2 This is a flowchart illustrating Embodiment 2 of the wind power prediction method of this application;

[0038] Figure 3 This is a schematic diagram of the single-parameter numerical weather forecast data format at a certain moment in the wind power prediction method of this application;

[0039] Figure 4 This is a schematic diagram of the multi-parameter numerical weather forecast data format at a certain moment in the wind power prediction method of this application;

[0040] Figure 5 This is a flowchart illustrating Embodiment 3 of the wind power prediction method of this application;

[0041] Figure 6 This is a schematic diagram of the module structure of the wind power prediction device according to an embodiment of this application;

[0042] Figure 7This is a schematic diagram of the equipment structure of the hardware operating environment involved in the wind power prediction method in the embodiments of this application.

[0043] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0044] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0045] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0046] The main solution of this application embodiment is: to convert the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data; to combine preset deep learning network components and construct a deep learning wind power prediction model based on the optimal component combination; to input the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain wind power prediction results.

[0047] Existing short-term wind power forecasting techniques typically rely solely on numerical weather forecast data from a single latitude and longitude point. This data cannot fully reflect the meteorological factors influencing wind farm power generation. For short-term wind power forecasting of a region, the power generation of each wind farm is usually predicted separately and then summed to obtain the total predicted power for all wind farms in the region. However, this summation method for regional wind power forecasting requires building a model for each wind farm, resulting in multiple models and significant human and computational resources.

[0048] This application provides a solution that converts acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data; combines preset deep learning network components to construct a deep learning wind power prediction model based on the optimal component combination; and inputs the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power, thereby obtaining the wind power prediction result. This application, by formatting regional meteorological data through grid overlay, can comprehensively reflect the meteorological characteristics of the region; by selecting the optimal component combination through the combination of preset deep learning network components, a deep learning wind power prediction model can be constructed, which can deeply extract the characteristics of large-scale regional meteorological data, accurately reflect regional meteorological characteristics, avoid model overfitting, and enable a single model to predict the total power generation of all wind farms within the region, achieving low consumption of human and computational resources.

[0049] It should be noted that the executing entity of the method in this embodiment can be a computing service device with wind power prediction, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone; or it can be a wind power prediction device with the same or similar functions. This embodiment and the following embodiments will be described using a wind power prediction device as an example.

[0050] Based on this, embodiments of this application provide a wind power prediction method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the wind power prediction method of this application.

[0051] In this embodiment, the wind power prediction method includes steps S10 to S30:

[0052] Step S10: Convert the acquired numerical weather forecast data within the preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data.

[0053] Understandably, existing forecasting algorithms, which rely solely on numerical weather forecast data from a single latitude and longitude point, often fail to fully reflect the meteorological factors influencing wind farm power generation. Therefore, this embodiment employs a deep learning algorithm to establish a short-term wind power forecasting model based on regional numerical weather forecasts. The aim is to leverage regional numerical weather forecast data to grasp meteorological trends within a specific spatial range and the correlation between meteorological conditions at different locations, eliminating the need to establish multiple models for multiple wind farms and reducing human and computational resources.

[0054] It should be understood that, firstly, numerical weather forecast data for a predefined area is acquired, which may include multiple wind farms. Then, the acquired regional numerical weather forecast data is organized into a gridded, multi-parameter overlay data format, which can comprehensively reflect the meteorological characteristics of the corresponding area.

[0055] Step S20: Combine the preset deep learning network components and construct a deep learning wind power prediction model based on the optimal component combination.

[0056] Understandably, by combining preset deep learning network components in different ways, a deep learning wind power prediction model is constructed based on the optimal component combination. Using this deep learning model, features of large-scale regional meteorological data can be deeply extracted, accurately reflecting regional meteorological characteristics while avoiding model overfitting. A single model can directly predict the total power generation of all wind farms within a region, with relatively low human and computational resource consumption.

[0057] Step S30: Input the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain the wind power prediction result.

[0058] Understandably, regional numerical weather forecast data for the time to be predicted can be organized into formatted regional numerical weather forecast data and used as input to a deep learning wind power prediction model to calculate the wind power prediction result for the corresponding time.

[0059] This embodiment provides a wind power prediction method. It involves converting acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data. Preset deep learning network components are combined, and a deep learning wind power prediction model is constructed based on the optimal component combination. The formatted regional numerical weather forecast data is then input into the deep learning wind power prediction model to predict wind power, yielding the prediction result. This application, by formatting regional meteorological data through grid overlay, can comprehensively reflect the meteorological characteristics of the region. By selecting the optimal component combination through the combination of preset deep learning network components, a deep learning wind power prediction model can be constructed. This model can deeply extract features from large-scale regional meteorological data, accurately reflect regional meteorological characteristics, avoid model overfitting, and enable a single model to predict the total power generation of all wind farms within the region, achieving low consumption of human and computational resources.

[0060] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 Step S10, the wind power prediction method further includes steps S101 to S102:

[0061] Step S101: The numerical weather forecast data is gridded according to the longitude, latitude and time information corresponding to the numerical weather forecast data in the preset area to obtain the gridded regional numerical weather forecast data.

[0062] Understandably, the numerical weather forecast data acquired within a preset area can be gridded according to longitude, latitude, and time information. The processed results can be used as a reference. Figure 3 Each grid represents a latitude and longitude. Figure 3This diagram illustrates the format of single-parameter numerical weather forecast data at a specific moment in the wind power prediction method of this application. First, the numerical weather forecast data within a preset area is sorted according to its time information. Then, the minimum latitude, maximum latitude, minimum longitude, and maximum longitude of the data at each time point are determined. The maximum latitude and minimum longitude are used as the starting and ending points of the grid, respectively. The numerical weather forecast data within the preset area is then placed into the grid according to the corresponding latitude and longitude, thus obtaining the gridded regional numerical weather forecast data corresponding to the single-parameter data at a specific moment.

[0063] Step S102: When the gridded regional numerical weather forecast data includes multi-dimensional parameters, multi-parameter overlay processing is performed on the multi-dimensional parameters in the gridded regional numerical weather forecast data to obtain formatted regional numerical weather forecast data.

[0064] It should be noted that regional numerical weather prediction data may include multi-dimensional parameters, such as latitude and longitude, wind speed and direction at various altitudes, temperature, humidity, and pressure. Therefore, when multi-dimensional parameters exist, they can be overlaid. The format of the processed data can be found by referring to [reference needed]. Figure 4 , Figure 4 This is a schematic diagram of the multi-parameter numerical weather forecast data format at a certain moment in the wind power prediction method of this application. Figure 4 The superimposed data can include multiple layers. The parameters in the superimposed data can be selected from all parameters or a few parameters in the numerical weather prediction data, or it can be data after transforming a certain parameter or some parameters.

[0065] In this embodiment, the numerical weather forecast data is gridded according to the longitude, latitude and time information corresponding to the numerical weather forecast data in the preset area to obtain the gridded regional numerical weather forecast data. When the gridded regional numerical weather forecast data includes multi-dimensional parameters, the multi-dimensional parameters in the gridded regional numerical weather forecast data are superimposed to obtain the formatted regional numerical weather forecast data. This data format can more comprehensively reflect the meteorological characteristics of the area.

[0066] Based on the first embodiment of this application, in the third embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 5 The preset deep learning network components include convolutional layers, pooling layers, activation layers, and fully connected layers; step S20, the wind power prediction method further includes steps S201 to S203:

[0067] Step S201: Combine convolutional layers, pooling layers, activation layers, and fully connected layers according to various combination methods to obtain multiple different component combinations.

[0068] It is understandable that deep learning mapping model structures include various basic components such as convolutional layers, pooling layers, activation layers, and fully connected layers. Different combinations of these basic components can be used to obtain multiple different deep learning network structures.

[0069] In one feasible implementation, the multiple combination methods include sequential connection, superimposed connection and / or short-circuit connection.

[0070] It should be understood that the various combination methods include sequential connection, superimposed connection, and / or short-circuit connection. That is, it can include multiple convolutional layers, pooling layers, activation layers directly connected in sequence, or multiple convolutional layers connected before being connected to other components, or combinations using short-circuit connections, etc.

[0071] Step S202: Evaluate the performance of the multiple different component combinations according to the cross-validation method to determine the optimal component combination.

[0072] Understandably, cross-validation can be used to evaluate the performance of multiple different component combinations and determine the optimal combination. Cross-validation is a statistical method for evaluating the generalization ability of a model. It involves splitting the dataset into multiple subsets, using one subset for testing, and the remaining subsets for training, repeating this process until each subset has been used as a test set. Cross-validation methods include K-fold cross-validation, leave-one-out cross-validation, and random subsampling validation.

[0073] In one possible implementation, step S202 may include steps S2021 to S2022:

[0074] Step S2021: Perform performance evaluation on the multiple different component combinations according to the cross-validation method to obtain performance evaluation results.

[0075] Step S2022: Determine the optimal component combination among the multiple different component combinations based on the preset performance indicators of the performance evaluation results. The preset performance indicators include the correlation between the predicted value and the actual value, root mean square error, normalized root mean square error, mean absolute error, and normalized mean absolute error.

[0076] It's important to note that performance evaluations can be performed on multiple different component combinations using cross-validation to select the optimal network structure as the deep learning mapping model. When selecting the optimal model, its performance in cross-validation is typically used for evaluation. Performance metrics include the correlation between predicted and actual values, root mean square error (RMSE), normalized RMSE, mean absolute error (MAE), and normalized mean absolute error (MAE). Lower values ​​for these metrics indicate better predictive performance. Furthermore, the correlation between the model's predictions and actual values ​​is also crucial; higher correlation indicates more reliable predictions.

[0077] In this embodiment, standard components of deep learning networks are combined in a combination manner to form different networks. The optimal component combination is determined by cross-validation and performance indicators as the preferred model, thereby selecting the best deep learning wind power prediction model.

[0078] Step S203: Construct a deep learning wind power prediction model based on the optimal component combination.

[0079] In one feasible implementation, step S30 may include steps S301 to S305:

[0080] Step S301: Input the formatted regional numerical weather forecast data into the deep learning wind power prediction model.

[0081] It is understandable that the data after grid overlay can be used as input to the deep learning mapping model, that is, the formatted regional numerical weather forecast data can be input into the deep learning wind power prediction model for wind power prediction.

[0082] It should be noted that after constructing the deep learning wind power prediction model based on the optimal component combination, the prediction model can be trained according to the pre-constructed training sample set. Each sample in the training sample set includes two data points: input data and output data. Each sample also needs to be formatted by grid overlay. The output data is the wind power at the corresponding time.

[0083] Step S302: Perform convolution operation on the formatted regional numerical weather forecast data through the convolutional layer of the deep learning wind power prediction model to obtain feature data.

[0084] It is understandable that different combinations result in different model structures. In the best combination, the convolutional layer may be the first layer, and there may be multiple convolutional layers. Therefore, each convolutional layer can perform convolution operations on the output of the previous layer to obtain a series of feature data.

[0085] Step S303: The feature data is sampled through the pooling layer of the deep learning wind power prediction model to obtain the sampled feature data.

[0086] It should be understood that the pooling layer receives the feature data output by the convolutional layer and performs sampling operations on the feature data to obtain the sampled feature data.

[0087] Step S304: The sampled feature data is activated by the activation layer of the deep learning wind power prediction model to obtain activated feature data.

[0088] It should be understood that the activation layer performs an activation operation on the sampled feature data to obtain activated feature data, thereby improving the nonlinearity of the network.

[0089] Step S305: The wind power prediction result is obtained by performing power mapping based on the activated feature data through the fully connected layer of the deep learning wind power prediction model.

[0090] It is understandable that the fully connected layer of the deep learning wind power prediction model performs power mapping based on the activated feature data to obtain the wind power prediction result.

[0091] It should be noted that, depending on the different selections of the power data in the training sample set, predictions can be made for single wind farm power, total regional power, or total power of several wind farms within a region. If the power data in the training sample set represents the power of a specific wind farm, the predicted result will be the power of that wind farm; if the power data in the training sample set represents the total power of all wind farms within the region, the predicted result will be the total regional power; and if the power data in the training sample set represents the total power of several wind farms within the region, the predicted result will be the total power of those wind farms.

[0092] It is worth noting that this application can improve the short-term power prediction accuracy of a single wind farm, and also improve the short-term prediction accuracy of the total added power of multiple or all wind farms in a certain region. Simultaneously, it can reduce the modeling time of regional wind power prediction models, thereby reducing the required computational and human resources. Table 1 shows the prediction errors of the existing scheme and this application for a certain wind farm from July to November 2019. Table 2 shows the prediction errors of the existing scheme and this application for 10 wind farms in a certain region from July to November 2019. Table 3 shows the modeling time required for the existing scheme and this application for 10 wind farms in a certain region.

[0093] Table 1 Comparison of prediction errors for a wind farm between existing solutions and this application.

[0094] July 2019 August 2019 September 2019 October 2019 November 2019 Existing solutions 0.1332 0.1431 0.1002 0.1803 0.1472 This application 0.1151 0.1278 0.0808 0.1783 0.1375

[0095] Table 2 Comparison of prediction errors of total power generation for 10 wind farms in a certain region between existing schemes and this application.

[0096] July 2019 August 2019 September 2019 October 2019 November 2019 Existing solutions 0.0850 0.0804 0.0798 0.0940 0.1394 This application 0.0602 0.0673 0.0756 0.0856 0.1294

[0097] Table 3 shows the modeling time required for existing solutions and this application for predicting 10 wind farms in a certain region.

[0098] Modeling time Existing solutions 153 minutes This application 48 minutes

[0099] In this embodiment, by combining various basic components such as convolutional layers, pooling layers, activation layers, and fully connected layers included in the deep learning mapping model structure, different deep learning network structures are obtained by combining the basic components. By using cross-validation combined with multiple performance indicators, the optimal network structure can be selected as the deep learning mapping model.

[0100] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the wind power prediction method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0101] This application also provides a wind power prediction device, please refer to... Figure 6 The wind power prediction device includes:

[0102] The data formatting module 10 is used to convert the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data.

[0103] The prediction model building module 20 is used to combine preset deep learning network components and build a deep learning wind power prediction model based on the optimal component combination.

[0104] The wind power prediction module 30 is used to input the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain wind power prediction results.

[0105] The wind power prediction device provided in this application, employing the wind power prediction method in the above embodiments, can solve the technical problem. Compared with the prior art, the beneficial effects of the wind power prediction device provided in this application are the same as those of the wind power prediction method provided in the above embodiments, and other technical features in the wind power prediction device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0106] This application provides a wind power prediction device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the wind power prediction method in the above embodiment 1.

[0107] The following is for reference. Figure 7 The diagram illustrates a structural schematic of a wind power prediction device suitable for implementing embodiments of this application. The wind power prediction device in these embodiments may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 7 The wind power prediction device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0108] like Figure 7As shown, the wind power forecasting device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the wind power forecasting device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the wind power forecasting device to communicate wirelessly or wiredly with other devices to exchange data. While the figure shows wind power forecasting devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0109] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0110] The wind power prediction device provided in this application, employing the wind power prediction method described in the above embodiments, can solve the technical problem of wind power prediction. Compared with the prior art, the beneficial effects of the wind power prediction device provided in this application are the same as those of the wind power prediction method provided in the above embodiments, and other technical features of the wind power prediction device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0111] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0112] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0113] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the wind power prediction method in the above embodiments.

[0114] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0115] The aforementioned computer-readable storage medium may be included in the wind power prediction device; or it may exist independently and not assembled into the wind power prediction device.

[0116] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by the wind power prediction device, the wind power prediction device causes the following: it converts the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data; it combines preset deep learning network components to construct a deep learning wind power prediction model based on the optimal component combination; and it inputs the formatted regional numerical weather forecast data into the deep learning wind power prediction model to perform wind power prediction and obtain wind power prediction results.

[0117] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0118] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0119] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0120] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described wind power prediction method, thereby solving the technical problem. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the wind power prediction method provided in the above embodiments, and will not be repeated here.

[0121] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the wind power prediction method described above.

[0122] The computer program product provided in this application can solve the technical problem. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the wind power prediction method provided in the above embodiments, and will not be repeated here.

[0123] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for predicting wind power output, characterized in that, The method includes: The acquired numerical weather forecast data within the preset area is converted into a preset grid overlay format to obtain formatted regional numerical weather forecast data. The preset deep learning network components are combined, and a deep learning wind power prediction model is constructed based on the optimal component combination. The formatted regional numerical weather forecast data is input into the deep learning wind power prediction model to predict wind power and obtain the wind power prediction results.

2. The method as described in claim 1, characterized in that, The step of converting the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data includes: The numerical weather forecast data is gridded based on the longitude, latitude and time information corresponding to the numerical weather forecast data within the preset area to obtain the gridded regional numerical weather forecast data. When the gridded regional numerical weather forecast data includes multi-dimensional parameters, multi-parameter overlay processing is performed on the multi-dimensional parameters in the gridded regional numerical weather forecast data to obtain formatted regional numerical weather forecast data.

3. The method as described in claim 1, characterized in that, The preset deep learning network components include convolutional layers, pooling layers, activation layers, and fully connected layers; The step of combining preset deep learning network components and constructing a deep learning wind power prediction model based on the optimal component combination includes: By combining convolutional layers, pooling layers, activation layers, and fully connected layers in various ways, multiple different component combinations can be obtained; The performance of the multiple different component combinations is evaluated using cross-validation to determine the optimal component combination. A deep learning wind power prediction model is constructed based on the optimal component combination.

4. The method as described in claim 3, characterized in that, The various combination methods include sequential connection, superimposed connection and / or short-circuit connection.

5. The method as described in claim 3, characterized in that, The step of evaluating the performance of the multiple different component combinations according to the cross-validation method and determining the optimal component combination includes: The performance of the multiple different component combinations is evaluated using cross-validation to obtain performance evaluation results. The optimal component combination is determined from the multiple different component combinations based on the preset performance indicators of the performance evaluation results. The preset performance indicators include the correlation between the predicted value and the actual value, root mean square error, normalized root mean square error, mean absolute error, and normalized mean absolute error.

6. The method according to any one of claims 3 to 5, characterized in that, The step of inputting the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain the wind power prediction result includes: The formatted regional numerical weather forecast data is input into the deep learning wind power prediction model; Feature data is obtained by performing convolution operations on the formatted regional numerical weather forecast data through the convolutional layer of the deep learning wind power prediction model. The feature data is sampled by the pooling layer of the deep learning wind power prediction model to obtain the sampled feature data. The activated feature data is obtained by activating the sampled feature data through the activation layer of the deep learning wind power prediction model. The wind power prediction result is obtained by mapping the activated feature data through the fully connected layer of the deep learning wind power prediction model.

7. A wind power prediction device, characterized in that, The wind power prediction device includes: The data formatting module is used to convert the acquired numerical weather forecast data within a preset area into a preset grid overlay format to obtain formatted regional numerical weather forecast data. The prediction model building module is used to combine preset deep learning network components and build a deep learning wind power prediction model based on the optimal component combination. The wind power prediction module is used to input the formatted regional numerical weather forecast data into the deep learning wind power prediction model to predict wind power and obtain wind power prediction results.

8. A wind power prediction device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the wind power prediction method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the wind power prediction method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the wind power prediction method as described in any one of claims 1 to 6.

Images