Offshore wind resource forecasting method, apparatus, computer equipment, and storage medium

The offshore wind resource forecasting method addresses data inconsistencies in multimode systems by converting datasets and using a gradient boosting library to establish a prediction model, improving forecasting accuracy and reliability.

JP7949055B2Active Publication Date: 2026-11-11CHINA THREE GORGES INT CORP
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2024557873
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2023-07-18
Filing Date
2024-06-13
Publication Date
2026-11-11
Estimated Expiration
2044-06-13

Smart Images

  • Figure 0007949055000007
    Figure 0007949055000007
  • Figure 0007949055000008
    Figure 0007949055000008
  • Figure 0007949055000009
    Figure 0007949055000009
Patent Text Reader

Abstract

This application relates to the technical field of offshore wind power forecasting and discloses an offshore wind power resource forecasting method, apparatus, computer equipment, and storage medium. The application converts a first offshore meteorological field dataset into a second offshore meteorological field dataset that can identify weather forecast modes, and optionally uses a first wind tower wind speed dataset and a second wind tower wind speed dataset as the basis to build a target offshore wind power resource forecasting model using an advanced distributed gradient boosting library. This not only improves the accuracy of offshore wind power resource forecasting, but also greatly simplifies the data fusion process. The integrated use of multi-source forecast data based on a global combined forecasting system can make up for the deficiencies of different data, fully utilize the advantages of various data, and make the forecast results more reliable.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of offshore wind power forecasting, and specifically relates to an offshore wind resource forecasting method, apparatus, computer device and storage medium. BACKGROUND ART

[0002] The Weather Research and Forecast (WRF) model is a numerical model used for weather forecasting and research, which is a high-resolution, non-hydrostatic, non-hierarchical, non-uniform atmospheric numerical model. The WRF model divides the Earth's atmosphere into horizontal grids and vertical layers, and simulates atmospheric phenomena by solving atmospheric dynamics, thermodynamics and hydrological cycle equations through discrete equations. This model is widely applied in wind resource simulation and wind speed forecasting.

[0003] In order to improve the accuracy and reliability of the model, a global coupled forecasting system is generally adopted to provide accurate and comprehensive meteorological environment driving data for the WRF model. The fusion of multi-source forecasting data can maximize the advantages of each global coupled forecasting system, compensate for the defects between different systems, and improve the accuracy, completeness and reliability of forecasting.

[0004] Many studies have shown that multi-model forecasting is significantly improved compared with a single forecasting system, but there are still some problems in the fusion process. For example, there are differences between the data formats and temporal resolutions of different forecasting systems, which require data format conversion and temporal alignment; at the same time, the physical processes and parameter settings of the forecasting models of different systems are different, which require constant parameter matching and data calibration; furthermore, for conflicts or inconsistencies between different system forecasting data, if reasonable data processing and quality control are not carried out, the forecasting results will be adversely affected. SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION

[0005] [Means for solving the problem]

[0006] In view of this, the present invention provides an offshore wind resource forecasting method, apparatus, computer equipment, and storage medium to solve the problem of poor forecasting results due to problems in conventional multimode forecasting fusion processes.

[0007] In the first aspect, the present application provides a method for predicting offshore wind resources, and the offshore wind resource prediction method is The process includes the steps of: obtaining a first offshore meteorological field dataset and a first wind speed data set based on a global coupled forecasting system; driving a weather forecasting mode using the global coupled forecasting system and converting the first offshore meteorological field dataset into a second offshore meteorological field dataset based on the weather forecasting mode; performing interpolation on the second offshore meteorological field dataset to obtain a second wind speed data set; establishing a target offshore wind resource prediction model using a distributed gradient boosting library based on the first and second wind speed data sets; obtaining a third offshore meteorological field dataset and a third wind speed data set for the wind field in the area to be predicted; inputting the third offshore meteorological field dataset and the third wind speed data set into the target offshore wind resource prediction model and obtaining wind resource prediction results for the wind field in the area to be predicted.

[0008] The offshore wind resource forecasting method according to this application converts a first offshore meteorological field dataset into a second offshore meteorological field dataset that can be identified by a weather forecasting mode, and selectively constructs a target offshore wind resource forecasting model using an advanced distributed gradient boosting library based on the first and second wind tower wind speed datasets. This not only improves the accuracy of offshore wind resource forecasting but also significantly simplifies the data fusion process. Furthermore, the comprehensive use of multi-source forecasting data based on a global coupled forecasting system can compensate for deficiencies in different data, fully utilize the advantages of various data, and make forecasting results more reliable.

[0009] In one selectable embodiment, the step of using the global coupled forecasting system to drive a weather forecasting mode and converting a first offshore meteorological field dataset to a second offshore meteorological field dataset based on the weather forecasting mode includes the steps of using the global coupled forecasting system to drive a weather forecasting mode and generating a target simulation region based on the weather forecasting mode, and interpolating the first offshore meteorological field dataset to the target simulation region using a preset interpolation method to obtain a second offshore meteorological field dataset.

[0010] This invention enables the driving of a weather forecast mode using the global coupled forecasting system, and converts a first oceanic meteorological field dataset into a second oceanic meteorological field dataset that the weather forecasting mode can identify, thereby improving data accuracy. Furthermore, it enables the acquisition of the first oceanic meteorological field dataset based on the global coupled forecasting system, compensating for deficiencies in different data, fully utilizing the advantages of various data, and making forecast results more reliable.

[0011] In one selectable embodiment, the steps of using the global coupled forecasting system to drive a weather forecasting mode and generating a target simulation region based on the weather forecasting mode are: The process includes: driving a weather forecast mode using the global coupled forecasting system and generating an initial simulation area; obtaining the length and width of the area of ​​the initial simulation area; dividing the length of the area into multiple longitude segments and dividing the width of the area into multiple latitude segments; determining the grid point longitude and latitude of each grid point based on the multiple longitude segments and the multiple latitude segments, wherein each grid point is composed of longitude segments and latitude segments; and generating a target simulation area based on the weather forecast mode based on the initial simulation area and the grid point longitude and latitude of each grid point.

[0012] This invention allows the generated target simulation region to be converted from input data into latitude and longitude data that can be identified by the weather forecast mode, by determining the grid point latitude and longitude of each grid point.

[0013] In one selectable embodiment, the step of establishing a target offshore wind resource prediction model using a distributed gradient boosting library based on a first wind tower wind speed dataset and a second wind tower wind speed dataset is: The method includes the steps of: processing a first wind tower wind speed dataset and a second wind tower wind speed dataset using a data processing method to obtain a first target wind tower wind speed dataset and a second target wind tower wind speed dataset; generating a target feature model based on the first target wind tower wind speed dataset and the second target wind tower wind speed dataset by feature engineering processing and cross-validation processing; establishing an initial offshore wind resource prediction model using a distributed gradient boosting library based on the second target wind tower wind speed dataset and the target feature model; and adjusting the initial offshore wind resource prediction model using the first wind tower wind speed dataset until a target offshore wind resource prediction model is obtained.

[0014] This invention achieves the objective of eliminating the influence of randomness in annual wind speed fluctuations by processing the first and second wind speed data sets using a data processing method. Furthermore, it improves the prediction accuracy of the target offshore wind resource forecasting model by constructing a model using a distributed gradient boosting library and adjusting the model using the second wind speed data set.

[0015] In one selectable embodiment, the step of establishing an initial offshore wind resource prediction model using a distributed gradient boosting library based on a first target wind tower wind speed dataset and a target feature model is: The process includes the steps of obtaining a model feature map of a target feature model, inputting the model feature map into a distributed gradient boosting library to obtain multiple mark points corresponding to the model feature map, obtaining a construction tree based on the multiple mark points, and establishing an initial offshore wind resource forecasting model based on the construction tree.

[0016] This invention obtains a construction tree at multiple marker points corresponding to a model feature map, uses the construction tree to establish an initial offshore wind resource prediction model, and improves the model's prediction accuracy to a certain extent.

[0017] In one selectable embodiment, the method is: The method further includes the steps of calculating the mean absolute error, mean square error, and root mean square error of a target offshore wind resource forecasting model, and evaluating the target offshore wind resource forecasting model based on the mean absolute error, mean square error, and root mean square error, and obtaining evaluation results.

[0018] This invention enables the evaluation of a target offshore wind resource prediction model using mean absolute error, mean square error, and root mean square error, thereby obtaining the prediction performance of the target offshore wind resource prediction model, and further provides data support for offshore wind resource prediction.

[0019] In one selectable embodiment, after the step of using the global coupled forecasting system to drive a weather forecasting mode and generating a target simulation region based on the weather forecasting mode, the method: The method further includes the steps of dividing the target simulation domain into multiple subdomains, and setting the subdomains as nested domains when the subdomains configure computing resources.

[0020] This invention enables the use of a high-resolution grid only within a nested region, where sub-regions requiring computing resource configurations are designated as nested regions. Furthermore, it plays a role in balancing computing efficiency and the accuracy of weather forecast mode simulations.

[0021] In a second aspect, the present invention provides an offshore wind resource forecasting device, the offshore wind resource forecasting device comprising: a data acquisition module for acquiring a first offshore meteorological field dataset and a first wind tower wind speed dataset based on a global coupled forecasting system; a conversion module for driving a weather forecasting mode using the global coupled forecasting system and converting the first offshore meteorological field dataset into a second offshore meteorological field dataset based on the weather forecasting mode; a processing module for performing interpolation processing on the second offshore meteorological field dataset to obtain a second wind tower wind speed dataset; an establishment module for establishing a target offshore wind resource forecasting model using a distributed gradient boosting library based on the first wind tower wind speed dataset and the second wind tower wind speed dataset; and an acquisition / input module for acquiring a third offshore meteorological field dataset and a third wind tower wind speed dataset of a wind field in a forecasting area, inputting the second offshore meteorological field dataset and the third wind tower wind speed dataset into a target offshore wind resource forecasting model, and obtaining wind resource forecasting results for a wind field in a forecasting area.

[0022] In a third aspect, the present application provides a computer device comprising a memory and a processor communicatively connected to each other, wherein the memory stores computer instructions, and the processor executes the computer instructions to perform the offshore wind resource prediction method according to either the first aspect or the corresponding embodiments described above.

[0023] In a fourth aspect, the present application provides a computer-readable storage medium storing computer instructions for causing a computer to execute the offshore wind resource prediction method according to either the first aspect or the corresponding embodiments described above.

[0024] To more clearly describe the specific embodiments of the present application or the technical solutions of the prior art, the drawings required for the description of the specific embodiments or the prior art are briefly described below. As is apparent, the drawings described below illustrate several embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative efforts. [Brief Description of the Drawings]

[0025] [Figure 1] This is a flowchart of an offshore wind resource prediction method according to an embodiment of the present application. [Figure 2] This is a flowchart of another offshore wind resource prediction method according to an embodiment of the present application. [Figure 3] This is a flowchart of yet another offshore wind resource prediction method according to an embodiment of the present application. [Figure 4] This is a flowchart of an offshore wind resource prediction method using multi-source forecast data according to an embodiment of the present application. [Figure 5] This is a structural block diagram of an offshore wind resource prediction apparatus according to an embodiment of the present application. [Figure 6] This is a schematic hardware structure diagram of a computer device according to an embodiment of the present application. [Modes for Carrying Out the Invention]

[0026] To further clarify the purpose, technical solutions and advantages of the embodiments of this application, the embodiments of this application will be described below clearly and completely with reference to the drawings and technical solutions of the embodiments of this application, and as will be obvious, the embodiments described are some, but not all, embodiments of this application. All other embodiments that a person skilled in the art can obtain without creative work based on the embodiments of this application are within the scope of protection of this application.

[0027] The offshore wind resource forecasting method according to the embodiment of the present invention establishes a target offshore wind resource forecasting model by comprehensively using multi-source forecasting data and a distributed gradient boosting library, thereby improving the accuracy of offshore wind resource forecasting, compensating for deficiencies in different data, fully utilizing the advantages of various data, and achieving greater reliability of forecast results.

[0028] According to the embodiments of the present application, an embodiment of an offshore wind resource forecasting method is provided. The steps shown in the flowchart of the drawings may be executed in a computer system such as a set of computer-executable instructions. Although the flowchart shows a logical order, the steps shown or described may be executed in an order different from the order shown herein.

[0029] This embodiment provides an offshore wind resource forecasting method, and Figure 1 is a flowchart of the offshore wind resource forecasting method according to this embodiment, and as shown in Figure 1, the process includes the following steps: Step S101: Obtain the first oceanic meteorological field dataset and the first wind speed data set from the global coupled forecasting system.

[0030] Here, the global coupled forecasting system may include the U.S. National Centers for Environmental Prediction (NCEP), the European Center for Medium-Range Weather Forecasts (ECMWF), the Japan Meteorological Agency (JMA), and the Meteorological Service of Canada (MSC), among others.

[0031] Each of the above global combined forecasting systems is configured with a corresponding data library, and meteorological data is stored in the data library.

[0032] The first offshore meteorological field dataset may include information such as the potential height of the offshore wind field, horizontal wind speed, temperature, atmospheric pressure, humidity, soil water content, and soil temperature.

[0033] The first wind speed data set consists of actual measurement data from wind towers installed in offshore wind fields, and may include measurements corresponding to different heights, wind speeds, wind direction values, surface temperature, humidity, and atmospheric pressure.

[0034] The steps to obtain the wind speed dataset from the first wind tower are as follows: A measurement period is set and marked as T. Actual measurement data is acquired within one measurement period T. The amount of abnormal data in the actual measurement data is acquired and marked as R. If the amount of abnormal data R exceeds the set period change value, the measurement period T is extended, and the actual measurement data within the extended measurement period is continuously monitored in real time until the corresponding first wind speed data set of the wind tower is obtained. In this embodiment, the objective of eliminating the influence of randomness in annual wind speed fluctuations is achieved by setting a measurement period and deciding whether or not to extend the measurement period based on the amount of abnormal data.

[0035] By selectively combining meteorological data from data libraries corresponding to each global coupled forecasting system, it is possible to generate multi-source meteorological data, i.e., a first oceanic meteorological field dataset. This compensates for deficiencies in different data sources, fully leverages the advantages of each data type, and makes forecast results more reliable.

[0036] Step S102: The weather forecast mode is activated using the global coupled forecasting system and the first oceanic meteorological field dataset is converted into a second oceanic meteorological field dataset based on the weather forecast mode.

[0037] Here, the weather forecast mode (WRF mode) is configured with a preprocessing process. This preprocessing process converts the data formed based on the global coupled forecasting system into a data format that the WRF mode can directly identify and read.

[0038] Specifically, when the WRF mode is driven using a global coupled forecasting system, the preprocessing process converts the first oceanic meteorological field dataset into a second oceanic meteorological field dataset that the WRF mode can directly identify and read.

[0039] Step S103: Interpolation is performed on the second oceanic meteorological field dataset to obtain the second wind speed data set from a wind sounding tower.

[0040] First, data that does not meet the pre-defined data criteria in the second oceanic meteorological field dataset is filtered and removed. Then, the filtered and removed second oceanic meteorological field dataset is interpolated to the corresponding wind station data to generate a refined dataset, i.e., the second wind station wind speed dataset, which includes Data-NCEP, Data-ECMWF, Data-JMA, and Data-MSC.

[0041] In step S104, a target offshore wind resource prediction model is established using a distributed gradient boosting library based on the first and second wind speed datasets.

[0042] Here, the distributed gradient boosting library (eXtreme Gradient Boosting, XGBoost) refers to a library of machine learning functions focused on gradient improvement algorithms that can be used for classification, regression, and sorting.

[0043] In this embodiment, the first wind speed dataset is used as the test set, and the second wind speed dataset is used as the training set. A regression model, i.e., a target offshore wind resource prediction model, is constructed based on the XGBoost library to predict offshore wind resources, improve the accuracy of offshore wind resource predictions, and significantly simplify the data fusion process.

[0044] In step S105, the third offshore meteorological field dataset and the third wind speed data set of the wind field in the area to be predicted are obtained, and the third offshore meteorological field dataset and the third wind speed data set of the wind field are input into the target offshore wind resource prediction model to obtain wind resource prediction results for the wind field in the area to be predicted.

[0045] Specifically, the acquired third offshore meteorological field dataset and third wind speed data set for the wind field in the target area can be input into the constructed target offshore wind resource prediction model, and wind resource prediction results for the wind field in the target area can be output.

[0046] The offshore wind resource forecasting method according to this embodiment converts a first offshore meteorological field dataset into a second offshore meteorological field dataset that can be identified by a weather forecasting mode. Furthermore, it constructs a target offshore wind resource forecasting model using an advanced distributed gradient boosting library based on the first and second wind tower wind speed datasets. This not only improves the accuracy of offshore wind resource forecasting but also significantly simplifies the data fusion process. Moreover, the comprehensive use of multi-source forecasting data based on a global coupled forecasting system can compensate for deficiencies in different data, fully utilize the advantages of various data, and make forecasting results more reliable.

[0047] This embodiment provides an offshore wind resource forecasting method, and Figure 2 is a flowchart of the offshore wind resource forecasting method according to the embodiment of this application, and as shown in Figure 2, the process includes steps S201 to S202, In step S201, the first oceanic meteorological field dataset and the first wind speed dataset are obtained based on the global coupled forecasting system. For details, please refer to step S101 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here. Step S202: Drive the weather forecast mode using the global coupled forecasting system and convert the first oceanic meteorological field dataset into a second oceanic meteorological field dataset based on the weather forecast mode.

[0048] Specifically, step S202 above includes steps S2021 to S2022, Step S2021: Drive the weather forecast mode using the global coupled forecasting system and generate a target simulation area based on the weather forecast mode. Here, the target simulation domain converts the input data into a data format that WRF mode can directly identify and read. Specifically, the corresponding target simulation domain can be obtained depending on the preprocessing process in WRF mode. Step S2022: The first oceanic meteorological field dataset is interpolated to the target simulation region using a pre-configured interpolation method to obtain the second oceanic meteorological field dataset. Here, different interpolation dimensions are set for the target simulation domain, and the first oceanic meteorological field dataset has corresponding interpolation feature coefficients. Specifically, after inputting the first oceanic meteorological field dataset into the target simulation domain, the first oceanic meteorological field dataset corresponding to the interpolated feature coefficients can be matched one-to-one with the interpolation dimensions in the target simulation domain based on the correspondence relationship of a pre-set reference interpolation correspondence table. After successful matching, the data in the first oceanic meteorological field dataset can be converted to corresponding longitude and latitude, and then converted into a second oceanic meteorological field dataset in WRF format. The WRF mode can then directly identify this second oceanic meteorological field dataset. Here, the reference interpolation correspondence table records the correspondence between interpolated feature coefficients and interpolation dimensions.

[0049] JPEG0007949055000001.jpg132159

[0050] JPEG0007949055000002.jpg35159

[0051] Retrieve the longitude and latitude of the grid points, making them selectable.

[0052] Specifically, each of two adjacent longitude segments and each of two adjacent latitude segments forms one square grid, and the intersections formed by the points where longitude and latitude segments meet are called grid points. In other words, the initial simulation region consists of multiple square grids, and each square grid contains multiple grid points.

[0053] To make them selectable, mark the grid points as X, X=<longitude, latitude>, and the corresponding grid point longitude and latitude for each grid point will be X=<w、h> Therefore, the range of the value of w is w1 to wN Therefore, the range of h values ​​is h1 to h M That is the case. Finally, based on the initial simulation area and taking into account the grid point longitude and latitude of each grid point, the target simulation area based on the corresponding weather forecast mode can be determined.

[0054] In some selectable embodiments, after step a1, step S2021 further includes steps a6 and a7. Step a6: Divide the target simulation domain into multiple subdomains. Step a7, when a sub-region configures computing resources, set the sub-region as a nested region.

[0055] Specifically, multiple longitude segments and multiple latitude segments are combined, and the longitude and latitude segments intersect to form an isolongitude-latitude grid, with the simulation area covering the entire isolongitude-latitude grid.

[0056] While equilatitude grids offer some options, they lack local encryption capabilities, and although they are high-resolution grids, high-resolution grids consume a large amount of computing resources.

[0057] Therefore, the target simulation domain can be divided into multiple subdomains, and each subdomain can configure computing resources. When a subdomain configures computing resources, it obtains its positioning number and sets the subdomain with that positioning number as a nested domain.

[0058] In this embodiment, sub-regions requiring computing resource configuration are designated as nested regions, and a high-resolution grid is employed only within these nested regions. The portion belonging to the target simulation region but not to the sub-regions where computing resources are configured is divided into a parent region, which provides a coarser resolution and plays a role in balancing computing efficiency and simulation accuracy.

[0059] Optionally, the second oceanic meteorological field dataset obtained through the above steps is called downscaled data due to its low spatial resolution.

[0060] In step S203, interpolation is performed on the second offshore meteorological field dataset to obtain the second wind speed dataset from the wind sounding tower. For details, please refer to step S103 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here.

[0061] In step S204, a target offshore wind resource prediction model is established using a distributed gradient boosting library based on the first and second wind speed datasets. For details, please refer to step S104 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here.

[0062] In step S205, the third offshore meteorological field dataset and the third wind speed data set of the wind field in the area to be predicted are obtained. These datasets are then input into the target offshore wind resource prediction model to obtain the wind resource prediction results for the wind field in the area to be predicted. For details, please refer to step S105 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here.

[0063] Step S206: Calculate the mean absolute error, mean square error, and root mean square error of the target offshore wind resource forecasting model.

[0064] JPEG0007949055000003.jpg43160

[0065] JPEG0007949055000004.jpg35160

[0066] JPEG0007949055000005.jpg35160

[0067] In step S207, the target offshore wind resource forecasting model is evaluated based on the mean absolute error, mean squared error, and root mean squared error, and the evaluation results are obtained.

[0068] JPEG0007949055000006.jpg25160

[0069] The offshore wind resource forecasting method according to this embodiment determines the grid point longitude and latitude of each grid point, thereby enabling the generated target simulation region to convert the input first offshore meteorological field dataset into a second offshore meteorological field dataset that the weather forecasting mode can identify, improving data accuracy. Furthermore, the first offshore meteorological field dataset can be acquired based on a global coupled forecasting system, compensating for deficiencies in different data, fully utilizing the advantages of various data, and making forecast results more reliable. Selectively, the model forecasting performance of the target offshore wind resource forecasting model is evaluated using mean absolute error, mean squared error, and root mean square error, providing data support for offshore wind resource forecasting.

[0070] This embodiment provides an offshore wind resource forecasting method, and Figure 3 is a flowchart of the offshore wind resource forecasting method according to the embodiment of the present application, and as shown in Figure 3, the process includes steps S301 to S305, In step S301, the first oceanic meteorological field dataset and the first wind speed dataset are obtained based on the global coupled forecasting system. For details, please refer to step S101 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here. In step S302, the weather forecast mode is driven using the global coupled forecasting system, and the first oceanic meteorological field dataset is converted into a second oceanic meteorological field dataset based on the weather forecast mode. For details, please refer to step S202 of the embodiment shown in Figure 2, and a repeated explanation will be omitted here. In step S303, interpolation is performed on the second offshore meteorological field dataset to obtain the second wind speed dataset from the wind sounding tower. For details, please refer to step S103 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here. In step S304, a target offshore wind resource prediction model is established using a distributed gradient boosting library based on the first and second wind speed datasets.

[0071] Specifically, step S304 above includes steps S3041 to S3044, In step S3041, the first wind speed data set and the second wind speed data set are processed by a data processing method to obtain the first target wind speed data set and the second target wind speed data set.

[0072] Here, the data processing method may include data evaluation, handling of missing values, handling of repeated values, handling of outliers, and variable transformation. (1) Data evaluation: Completeness detection is performed on the test set and training set through data evaluation, and missing values, repeated values, and outliers are selected in the test set and training set. (2) Missing value handling: Missing value handling identifies the content of missing values ​​and the position of the matrix in which the missing values ​​exist, and performs a fill operation on the missing values. (3) Iterative value processing: Iterative value processing detects repeating data in the test set (second wind tower wind speed dataset) and the training set (first wind tower wind speed dataset). Traversal is performed sequentially on the data in the test set and the training set. The first traversed data is marked as A, and subsequent traversals of the same data are marked as A'. After the traversal is complete, the number of times the data has been acquired is obtained and marked as Y, and Y-1 such data are removed. (4) Anomaly handling: Anomaly handling detects abnormal data values ​​in the training set and test set, and the abnormal data values ​​are held, deleted, or corrected. (5) Variable transformation: The training set and dataset are transformed into a modeling format, and the transformed modeling format can be used to build the data model.

[0073] In step S3042, a target feature model is generated based on the first target wind speed dataset and the second target wind speed dataset through feature engineering and cross-validation processing.

[0074] Here, the feature engineering process may include feature extraction, feature transformation, and feature construction. Specifically, after obtaining the modeling format, feature extraction, feature transformation, and feature construction are performed on the first target wind speed dataset and the second target wind speed dataset corresponding to the modeling format. First, multiple valid feature points and invalid noise points are obtained from the first target wind speed tower dataset and the second target wind speed tower dataset through feature extraction, and the invalid noise points are removed. Next, the effective feature points obtained through feature transformation are converted into mark points in the feature map information, and multiple mark points are combined to generate a model feature map. Finally, the model feature map is modeled into a corresponding feature model through feature construction.

[0075] In the process of constructing the above feature model, it is necessary to perform cross-validation against the second target wind speed dataset (training set). First, the training set is divided into K equal subsets using K-set cross-validation. These K subsets are then traversed, with each subset encountered during the traverse being used as the validation set, and the remaining K-1 subsets being used as subtraining sets. Next, the feature model is trained K-1 times using the subtraining set. During each of the K-1 training iterations, the performance of the feature model is evaluated using the validation set. This training of the feature model on the subtraining set and the evaluation of the feature model's performance on the test set are repeated until each subset functions as a validation set at least once and the optimal model parameters are obtained. Finally, based on the optimal model parameters and the corresponding feature model constructed through feature engineering, the corresponding target feature model can be determined.

[0076] In step S3043, an initial offshore wind resource prediction model is established using a distributed gradient boosting library, based on the second target wind speed dataset and the target feature model. Specifically, by using the XGBoost library and the second target wind speed dataset, a target feature model can be trained to obtain a corresponding initial offshore wind resource prediction model.

[0077] Step S3044: The initial offshore wind resource forecasting model is adjusted using the first wind speed data set until a target offshore wind resource forecasting model is obtained. Specifically, the model parameters and hyperparameters of the initial offshore wind resource prediction model are adjusted using the first wind speed data set from the wind sounding tower, and the corresponding target offshore wind resource prediction model is output until the model parameters and hyperparameters are optimized.

[0078] Here, model parameters can be selected based on a set of initial parameters. Hyperparameters can include the learning rate, the number of build trees, the maximum depth of the build trees, and the column sampling rate.

[0079] Hyperparameter optimization can be performed using either grid search or random search, allowing users to select and combine any two different hyperparameters. The performance of the initial offshore wind resource prediction model with these different hyperparameters is then evaluated. The performance of the initial offshore wind resource prediction model is determined by a performance coefficient, and the hyperparameter combination with the highest performance coefficient is selected as the optimal hyperparameter combination.

[0080] The model parameters can be optionally adjusted and optimized without changing hyperparameters to continuously improve the performance of the initial offshore wind resource prediction model. Model parameter optimization may be performed by adjusting the learning rate to different set learning coefficient levels, building the maximum tree depth to different set tree depth levels, or increasing the column sampling rate.

[0081] In some selectable embodiments, step S3043 includes steps b1 to b3, Step b1: Obtain the model feature map of the target feature model. Step b2: Input the model feature map into the distributed gradient boosting library to obtain multiple marker points corresponding to the model feature map. Step b3: Obtain a construction tree based on multiple marker points, and establish an initial offshore wind resource forecasting model based on the construction tree.

[0082] Specifically, the model feature map of the acquired target feature model is input into the XGBoost library, and the XGBoost library can retrieve multiple mark points included in the model feature information deconstructed from the model feature map.

[0083] After that, the construction tree is obtained at the marked point.

[0084] Selectively, the constructed tree includes multiple tree nodes, each with a corresponding partitioning contribution, each with a corresponding feature importance, and each feature importance has a different importance score. In this embodiment, the feature with the highest importance score is selected as the final feature subset, and this feature subset is obtained to construct and generate the initial offshore wind resource prediction model.

[0085] Here, the importance score can be obtained using the `feature_import` attribute of the XGBoost library.

[0086] In step S305, the third offshore meteorological field dataset and the third wind speed data set of the wind field in the area to be predicted are obtained. These datasets are then input into the target offshore wind resource prediction model to obtain the wind resource prediction results for the wind field in the area to be predicted. For details, please refer to step S105 of the embodiment shown in Figure 1, and a repeated explanation will be omitted here.

[0087] The offshore wind resource prediction method according to this embodiment processes the first wind tower wind speed dataset and the second wind tower wind speed dataset using a data processing method, thereby achieving the objective of eliminating the influence of randomness in annual wind speed fluctuations. Furthermore, a construction tree is obtained at multiple marker points corresponding to the model feature map obtained using a distributed gradient boosting library, and a target offshore wind resource prediction model is established using the first wind tower wind speed dataset and the construction tree, thereby improving the prediction accuracy of the target offshore wind resource prediction model.

[0088] One embodiment provides an offshore wind resource forecasting method based on multi-source forecast data, which includes steps S1 to S3 as shown in Figure 4. Step S1: Access the global coupled forecasting system and use data-driven WRF mode to define the simulation domain and nested domain, calculate grid point longitude and latitude, acquire ground data and meteorological field data, and interpolate them within the simulation domain. Step S2: Acquire meteorological field data of the target area wind field, interpolate the downscaled data generated in WRF mode onto the wind sounding tower to obtain refined wind speed data, acquire actual measured data from the wind sounding tower, use the refined wind speed data and the measured data as the training set and test set respectively, perform data preprocessing and feature engineering operations, and perform cross-validation against the training set to select the optimal model parameters. Step S3: Construct a regression model using the XGBoost library, train the regression model using the training set, adjust the model parameters and hyperparameters of the regression model according to the test set, calculate the mean absolute error, mean squared error and root mean squared error of the regression model to evaluate the prediction performance of the regression model, and predict the wind speed at the wind field location using the trained regression model.

[0089] The process for defining the simulation domain and nested domain, as selectable, includes the following: The system acquires an offshore measurement area after the data has been driven into WRF mode, marks the offshore measurement area as a simulation area, obtains the length and width of the area of ​​the offshore measurement area, divides the length and width of the area into multiple longitude and latitude segments, the longitude and latitude segments intersect to form an equilongitude-latitude grid, divides the simulation area into multiple sub-areas, allows each sub-area to configure computing resources, and sets the sub-area as a nested area when it configures computing resources.

[0090] The process for obtaining the longitude and latitude of grid points, which can be selected, includes the following: Multiple square grids are obtained, each consisting of two adjacent longitude segments and two adjacent latitude segments within the simulation domain. The intersections of the longitude and latitude segments constitute grid points, and each grid point has a corresponding longitude and latitude. The longitude and latitude of the grid points for the entire simulation domain are obtained based on the longitude and latitude of each grid point.

[0091] The interpolation process, as selectable, includes the following: After acquiring ground data and weather field data, the simulation domain is divided into a ground layer and a weather field layer. The ground data and weather field data have corresponding interpolation feature coefficients, and different interpolation dimensions are set for the ground layer and the weather field layer. After the ground data and weather field data are input into the simulation domain, the ground data and weather field data related to the interpolation feature coefficients are matched with the interpolation dimensions corresponding to the ground layer and the weather field layer.

[0092] The process for generating optional training and test sets includes the following: The wind field region of the wind speed to be detected is set as the target region, meteorological field data of the target region is acquired, downscaled data is interpolated into the installed wind sounding tower, refined data is generated after interpolating the downscaled data into the wind sounding tower, actual measured data is acquired from the wind sounding tower, the acquired refined data is used as the training set, and the acquired actual measured data is used as the test set.

[0093] The optional data preprocessing and feature engineering operations include the following: Data preprocessing includes data evaluation, missing value handling, repeated value handling, outlier handling, and variable transformation. Data evaluation identifies missing values, repeated values, and outliers in the test and training sets. Missing value handling identifies the content of missing values ​​and their matrix positions and fills in the missing values. Repeated value handling detects repeated data in the training and test sets and removes them through traversal. Outlier handling detects anomalous data values ​​in the training and test sets and performs retention, deletion, and modification on anomalous data values. Variable transformation is used to transform the training and datasets into a modeling format. Feature extraction is performed on the training and datasets corresponding to the modeling format to obtain multiple effective feature points. Feature transformation converts the obtained effective feature points into mark points for feature map information. Multiple mark points are combined to generate a model feature map, and feature construction models the model feature map into a corresponding feature model.

[0094] The process for cross-validating the training set, which can be selected, includes the following: After obtaining the training set, the training set is divided into K equal subsets, the K subsets are traversed, each subset in the traversal is sequentially used as the validation set, and the remaining K-1 subsets are used as subtraining sets. The feature model is trained K-1 times using the subtraining sets, and the performance of the feature model is evaluated using the validation set after each training session. The training of the feature model using the subtraining sets and the evaluation of the feature model's performance using the test sets are repeated until each subset functions as a validation set at least once, and after cross-validation is complete, the optimal model parameters are generated.

[0095] The process of constructing a regression model and adjusting the model parameters and hyperparameters, as selected, includes the following: The process involves obtaining a model feature map, inputting it into the XGBoost library, obtaining multiple mark points from the deconstructed feature map information, obtaining a constructed tree from the mark points, the constructed tree containing multiple tree nodes, different tree nodes having corresponding partitioning contributions, partitioning contributions having different feature importances, and feature importances having different importance scores, selecting the importance score with the highest value as the feature subset, constructing a regression model according to the feature subset, the regression model having corresponding model parameters and hyperparameters, transferring the model parameters and hyperparameters to the regression model, training the regression model using the training set, and optimizing the model parameters and hyperparameters according to the representation of the test set.

[0096] For the optional process of calculating the mean absolute error, mean squared error, and root mean squared error of the regression model, refer to the above relationships (1) to (3).

[0097] Compared to conventional technology, the beneficial effects of the offshore wind resource forecasting method based on multi-source forecast data according to this embodiment are as follows: 1. Based on the results of predicting target area wind speed in WRF mode and actual measurement data, fitting predictions are performed using the advanced XGBoost regression fitting algorithm. This not only improves the accuracy of offshore wind resource predictions but also significantly simplifies the data fusion process. Furthermore, the comprehensive use of multi-source forecast data can compensate for deficiencies in different data, fully utilizing the advantages of various data and making prediction results more reliable. 2. The equilongitude-latitude grid lacks local encryption capabilities. While the equilongitude-latitude grid is a high-resolution grid, high-resolution grids consume a large amount of computing resources. Sub-regions requiring computing resource configurations are designated as nested regions, and high-resolution grids are employed only within these nested regions. The portion belonging to the simulation region but not to a sub-region where computing resources are configured is divided into a parent region. The parent region provides a coarser resolution, balancing computing efficiency and simulation accuracy. By setting a measurement cycle and deciding whether to extend it based on the amount of anomalous data, the objective of eliminating the influence of randomness in annual wind speed fluctuations is achieved.

[0098] This embodiment further provides an offshore wind resource forecasting device, which is used to implement the above embodiment and optional embodiments, and parts that have been described are omitted for repetition. As used below, the term “module” can implement a combination of software and / or hardware with a predetermined function, and the devices described in the following embodiments are preferably implemented in software, but can also be implemented in hardware, or a combination of software and hardware, and are conceived.

[0099] This embodiment provides an offshore wind resource forecasting device, as shown in Figure 5, A data acquisition module 501 for acquiring a first oceanic meteorological field dataset and a first wind speed data set from a wind sounding tower based on a global coupled forecasting system, A conversion module 502 drives the weather forecast mode using a global coupled forecasting system and converts the first oceanic meteorological field dataset to a second oceanic meteorological field dataset based on the weather forecast mode. A processing module 503 performs interpolation on the second oceanic meteorological field dataset to obtain the second wind speed data set from a wind sounding tower, Based on the first and second wind speed datasets, establishment module 504 is used to establish a target offshore wind resource prediction model using a distributed gradient boosting library. The system includes an acquisition / input module 505 for acquiring a third offshore meteorological field dataset and a third wind speed data set for the wind field in the area to be predicted, inputting the third offshore meteorological field dataset and the third wind speed data set into a target offshore wind resource prediction model, and obtaining wind resource prediction results for the wind field in the area to be predicted.

[0100] In some selectable embodiments, the conversion module 502 is A generation unit for driving the weather forecast mode using a global coupled forecasting system and generating a target simulation area based on the weather forecast mode, The system includes an interpolation unit for interpolating a first oceanic meteorological field dataset to a target simulation region using a pre-defined interpolation method to obtain a second oceanic meteorological field dataset.

[0101] In some selectable embodiments, the generate unit is A first generation subunit for driving the weather forecast mode using a global coupled forecasting system and generating the initial simulation region, A first acquisition subunit for obtaining the length and width of the area of ​​the initial simulation region, A division subunit for dividing the length of the area into multiple longitude segments and the width of the area into multiple latitude segments, Based on multiple longitude segments and multiple latitude segments, the grid point longitude and latitude of each grid point are determined, and the grid point is a determination subunit composed of a longitude segment and a latitude segment. It includes a second generation subunit for generating a target simulation region based on a weather forecast mode, based on an initial simulation region and the grid point longitude and latitude of each grid point.

[0102] In some selectable embodiments, the establishment module 504 is A first processing unit for processing the first wind speed data set and the second wind speed data set using a data processing method to obtain the first target wind speed data set and the second target wind speed data set, A second processing unit generates a target feature model by performing feature engineering and cross-validation processing based on the first target wind speed dataset and the second target wind speed dataset. Based on the second target wind speed dataset and target feature model, an establishment unit is provided to establish an initial offshore wind resource prediction model using a distributed gradient boosting library, It includes an adjustment unit for adjusting an initial offshore wind resource forecasting model until a target offshore wind resource forecasting model is obtained, using the first wind speed data set from a wind sounding tower.

[0103] In some selectable embodiments, the establishment unit is A second acquisition subunit for obtaining the model feature map of the target feature model, An input subunit for inputting a model feature map into a distributed gradient boosting library and obtaining multiple mark points corresponding to the model feature map, This includes a construction subunit for obtaining a construction tree based on multiple mark points and for establishing an initial offshore wind resource forecasting model based on the construction tree.

[0104] In some selectable embodiments, the offshore wind resource forecasting device is A computing module for calculating the mean absolute error, mean square error, and root mean square error of a target offshore wind resource forecasting model, The system further includes an evaluation module for evaluating a target offshore wind resource forecasting model based on mean absolute error, mean squared error, and root mean squared error, and for obtaining evaluation results.

[0105] In some selectable embodiments, the conversion module 502 is A division unit for dividing the target simulation domain into multiple subdomains, When a subdomain configures computing resources, it further includes a configuration unit for setting the subdomain as a nested domain.

[0106] Further functional descriptions of each of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0107] The offshore wind resource forecasting device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0108] The embodiment of the present invention further provides computer equipment including an offshore wind resource forecasting device as shown in Figure 5 above.

[0109] As shown in Figure 6, which is a schematic diagram of the structure of a computer device according to an optional embodiment of the present invention, the computer device includes one or more processors 10, memory 20, and interfaces for connecting each component, including a high-speed interface and a low-speed interface. Each component communicates with one another via different buses and may be mounted on a common motherboard or otherwise mounted as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory for displaying GUI graphic information on an external input / output device (e.g., a display device coupled to the interface). In some optional embodiments, multiple processors and / or multiple buses may be used together with multiple memories as needed. Similarly, multiple computer devices may be connected, each providing a portion of the required operations (e.g., functioning as a server array, a set of blade servers, or a multiprocessor system). In Figure 6, one processor 10 is used as an example.

[0110] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include hardware chips. The hardware chips may be application-specific integrated circuits, programmable logic devices, or a combination thereof. The programmable logic devices may be complex programmable logic devices, field programmable logic gate arrays, generic array logic, or any combination thereof.

[0111] However, the memory 20 stores instructions that can be executed by at least one processor 10, thereby enabling at least one processor 10 to execute and implement the method shown in the above embodiment.

[0112] Memory 20 may include an operating system, a program storage area for storing application programs required for at least one function, and a data storage area for storing data created in accordance with the use of the computer equipment. Memory 20 may also include high-speed random-access memory and may further include non-temporary memory such as at least one disk storage device, flash memory device, or other non-temporary solid-state storage device. In some selectable embodiments, memory 20 may optionally include memory remotely installed relative to the processor 10, and these remote memories may be connected to the computer equipment via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0113] Memory 20 may include volatile memory such as random access memory, or non-volatile memory such as flash memory, hard disk, or solid-state drive, and memory 20 may include a combination of the above types of memory.

[0114] The computer equipment further includes a communication interface 30 for the computer equipment to communicate with other equipment or a communication network.

[0115] Embodiments of the present application further provide a computer-readable storage medium, and the methods according to the embodiments of the present application may be implemented in hardware, firmware, or in a recordable manner on a storage medium, or as computer code downloaded over a network, originally stored on a remote storage medium or a non-temporary machine-readable storage medium but stored on a local storage medium, thereby enabling the methods described herein to be processed by a general-purpose computer, a dedicated processor, or software stored on a storage medium using programmable or dedicated hardware. The storage medium may be a magnetic disk, an optical disk, read-only memory, random access memory, flash memory, a hard disk, or a solid-state drive, and optionally, the storage medium may include a combination of the above types of memory. As understood, the computer, processor, microprocessor controller, or programmable hardware includes a storage component capable of storing or receiving software or computer code, and when the software or computer code is accessed and executed by the computer, processor, or hardware, the methods shown in the embodiments are implemented.

[0116] While embodiments of the present application have been described with reference to the drawings, those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application, and such changes and modifications all fall within the scope defined in the attached claims.

Claims

1. A method for predicting offshore wind resources, The steps include obtaining a first oceanic meteorological field dataset and a first wind speed dataset based on a global coupled forecasting system, The steps include: driving a weather forecast mode using the global coupled forecasting system and converting the first offshore meteorological field dataset into a second offshore meteorological field dataset based on the weather forecasting mode; The steps include performing interpolation on the second oceanic meteorological field dataset to obtain a second wind speed dataset from a wind measuring tower, The steps include establishing a target offshore wind resource prediction model using a distributed gradient boosting library based on the first wind speed data set and the second wind speed data set, The process includes the steps of: obtaining a third offshore meteorological field dataset and a third wind speed data set for the wind field in the area to be predicted; inputting the third offshore meteorological field dataset and the third wind speed data set into the target offshore wind resource prediction model; and obtaining wind resource prediction results for the wind field in the area to be predicted. The steps of using the global coupled forecasting system to drive the weather forecasting mode and converting the first offshore meteorological field dataset to a second offshore meteorological field dataset based on the weather forecasting mode are: The steps include: driving a weather forecast mode using the global coupled forecasting system and generating a target simulation area based on the weather forecast mode; The process includes the step of interpolating the first oceanic meteorological field dataset to the target simulation region using a pre-set interpolation method to obtain the second oceanic meteorological field dataset, After the steps of driving a weather forecast mode using the global coupled forecasting system and generating a target simulation region based on the weather forecast mode, the method: The steps include dividing the target simulation domain into multiple subdomains, When a sub-region configures computing resources, the configured sub-region is set as a nested region, a high-resolution grid is adopted only within the nested region, and the portion belonging to the target simulation region but not belonging to the sub-region where the computing resources are configured is set as a parent region, and the parent region provides a coarse resolution. The step of establishing a target offshore wind resource prediction model using a distributed gradient boosting library based on the first wind speed data set and the second wind speed data set is as follows: The steps include processing the first wind speed data set and the second wind speed data set using a data processing method to obtain a first target wind speed data set and a second target wind speed data set, The steps include generating a target feature model based on the first target wind speed data set and the second target wind speed data set by performing feature engineering processing and cross-validation processing, The steps include establishing an initial offshore wind resource prediction model using the distributed gradient boosting library based on the second target wind speed dataset and the target feature model, An offshore wind resource prediction method characterized by comprising the step of adjusting the initial offshore wind resource prediction model using the first wind speed data set of the wind sounding tower until the target offshore wind resource prediction model is obtained.

2. The steps of driving a weather forecast mode using the global coupled forecasting system and generating a target simulation region based on the weather forecast mode are: The steps include: driving the weather forecast mode using the global coupled forecasting system and generating an initial simulation area; The steps include obtaining the length and width of the area of ​​the initial simulation region, The steps include dividing the length of the area into multiple longitude segments and dividing the width of the area into multiple latitude segments, A step of determining the grid point longitude and latitude of each grid point based on the plurality of longitude segments and the plurality of latitude segments, wherein the grid point is composed of a longitude segment and a latitude segment, The method according to claim 1, comprising the step of generating the target simulation region based on the weather forecast mode, based on the initial simulation region and the grid point longitude and latitude of each of the grid points.

3. The step of establishing an initial offshore wind resource prediction model using the distributed gradient boosting library based on the second target wind speed dataset and the target feature model is as follows: The steps include obtaining a model feature map of the aforementioned target feature model, The steps include inputting the aforementioned model feature map into the distributed gradient boosting library and obtaining a number of mark points corresponding to the aforementioned model feature map, The method according to claim 1, comprising the steps of obtaining a construction tree based on the plurality of mark points and establishing the initial offshore wind resource prediction model based on the construction tree.

4. The steps include calculating the mean absolute error, mean square error, and root mean square error of the target offshore wind resource prediction model, The method according to claim 1, further comprising the steps of evaluating the target offshore wind resource prediction model based on the mean absolute error, the mean square error, and the root mean square error, and obtaining an evaluation result.

5. An offshore wind resource forecasting device, A data acquisition module for obtaining the first oceanic meteorological field dataset and the first wind speed dataset from a wind sounding tower based on a global coupled forecasting system, A conversion module for driving a weather forecast mode using the global coupled forecasting system and converting the first offshore meteorological field dataset into a second offshore meteorological field dataset based on the weather forecast mode, A processing module for performing interpolation on the second oceanic meteorological field dataset to obtain a second wind speed dataset from a wind measuring tower, Based on the first wind speed data set and the second wind speed data set, an establishment module for establishing a target offshore wind resource prediction model using a distributed gradient boosting library, The system includes an acquisition / input module for acquiring a third offshore meteorological field dataset and a third wind speed data set for the wind field in the area to be predicted, inputting the third offshore meteorological field dataset and the third wind speed data set into the target offshore wind resource prediction model, and obtaining wind resource prediction results for the wind field in the area to be predicted, The aforementioned conversion module is A generation unit for driving the weather forecast mode using a global coupled forecasting system and generating a target simulation area based on the weather forecast mode, It includes an interpolation unit for interpolating a first oceanic meteorological field dataset to a target simulation region using a pre-defined interpolation method to obtain a second oceanic meteorological field dataset, The aforementioned conversion module is A division unit for dividing the target simulation domain into multiple subdomains, When a sub-domain configures computing resources, the configured sub-domain is set as a nested domain, a high-resolution grid is adopted only within the nested domain, and the portion belonging to the target simulation domain but not belonging to the sub-domain where the computing resources are configured is set as a parent domain, and the parent domain further includes a configuration unit to provide a coarser resolution. The aforementioned establishment module is A first processing unit for processing the first wind speed data set and the second wind speed data set using a data processing method to obtain the first target wind speed data set and the second target wind speed data set, A second processing unit generates a target feature model by performing feature engineering and cross-validation processing based on the first target wind speed dataset and the second target wind speed dataset. Based on the second target wind speed dataset and target feature model, an establishment unit is provided to establish an initial offshore wind resource prediction model using a distributed gradient boosting library, An offshore wind resource forecasting device characterized by including an adjustment unit for adjusting an initial offshore wind resource forecasting model until a target offshore wind resource forecasting model is obtained, using a first wind speed data set from a wind sounding tower.

6. Computer equipment, A computer device comprising a memory and a processor that are connected to each other in communication, wherein the memory stores computer instructions, and the processor executes the offshore wind resource forecasting method according to any one of claims 1 to 4 by executing the computer instructions.

7. A computer-readable storage medium characterized in that it stores computer instructions for causing a computer to execute the offshore wind resource forecasting method described in any one of claims 1 to 4.

Citation Information

Patent Citations

  • Wind resource prediction method based on WRF and CNN convolutional neural network

    CN112100922A

  • Wind power prediction method based on virtual meteorological technology

    CN113344252A

  • Method for computer-implemented forecasting of wind phenomena having influence on wind turbine

    CN114585950A

  • Offshore wind plant power prediction method, device, storage medium and system

    CN115034159A

  • Training method and device of wind speed prediction model, electronic equipment and storage medium

    CN115049024A