A method, system, device and storage medium for wind energy resource prediction

Abstract

This invention discloses a wind energy resource prediction method, system, device, and storage medium, relating to the field of renewable energy. The method includes: acquiring three-dimensional wind field observation data of a target area; performing spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence; fusing multi-source data from the spatiotemporal sequence to construct a multi-dimensional wind field matrix; establishing a wind energy potential prediction model based on the multi-dimensional wind field matrix; and predicting the wind energy data to be predicted based on the wind energy potential prediction model to obtain wind energy resource prediction results. By constructing a machine learning model, it provides efficient, accurate, and reliable resource assessment for wind energy development in various terrain areas.

Description

Technical Field

[0001] This invention relates to the field of renewable energy, and specifically to a method, system, device, and storage medium for wind energy resource forecasting. Background Technology

[0002] As an important clean and renewable energy source, the development of wind energy relies on accurate wind resource assessment. Currently, traditional assessment methods primarily depend on ground-based meteorological stations and mesoscale meteorological data, collecting parameters such as wind speed, wind direction, temperature, and humidity through fixed wind measurement towers or meteorological stations. This data provides the foundation for wind farm site selection and design, directly impacting the power generation efficiency and economic benefits of wind farms.

[0003] Traditional methods suffer from limitations such as a limited number of data collection points, limited coverage, and high costs, making them unsuitable for wind energy assessment in complex terrains or large areas. In recent years, drone technology has provided a new tool for wind energy assessment due to its flexible deployment, rapid response, and multi-point data collection capabilities. However, efficiently integrating the massive amounts of data collected by drones to improve the accuracy and efficiency of the assessment remains a pressing technical challenge. Summary of the Invention

[0004] The main objective of this invention is to provide a wind energy resource prediction method, system, device, and storage medium, which provides efficient, accurate, and reliable resource assessment for wind energy development in various terrain regions by constructing a machine learning model.

[0005] To achieve the above objectives, the embodiments of this application provide the following technical solutions:

[0006] According to a first aspect of the embodiments of this application, a wind energy resource prediction method is provided, the method comprising:

[0007] Acquire three-dimensional wind field observation data for the target area;

[0008] The three-dimensional wind field observation data are subjected to spatiotemporal alignment processing to generate a spatiotemporal sequence;

[0009] Multi-source data fusion is performed on the spatiotemporal sequence to construct a multi-dimensional wind field matrix;

[0010] A wind energy potential prediction model is established based on the aforementioned multidimensional wind field matrix;

[0011] Based on the wind energy potential prediction model, the wind energy data to be predicted is predicted to obtain the wind energy resource prediction results.

[0012] Optionally, three-dimensional wind field observation data of the target area are acquired, including:

[0013] A layered flight path planning instruction is generated based on the terrain data of the target area and the vertical gradient data of the wind measuring tower; the vertical gradient data of the wind measuring tower includes wind energy parameters at several preset heights within a preset height range;

[0014] The drone formation is invoked to perform cruises at several vertically spaced altitude layers within the preset altitude range according to the layered flight path planning instructions, while maintaining the cruise speed within a preset range; the drone formation includes several drones equipped with weather sensors;

[0015] Real-time three-dimensional wind field observation data are collected by various drones, and the collection time, spatial coordinates and attitude data are recorded; the sampling density is dynamically adjusted according to the vertical distribution law of wind speed, with the sampling frequency of high altitude layer set to be greater than that of low altitude layer.

[0016] Optionally, the three-dimensional wind field observation data undergoes spatiotemporal alignment processing to generate a spatiotemporal sequence, including:

[0017] Based on a preset physical threshold range and the setting conditions of the wind measurement tower data, abnormal data points in the three-dimensional wind field observation data are removed; the physical thresholds include wind speed, wind direction and temperature; the setting conditions of the wind measurement tower data include setting the data missing rate to be lower than a set missing rate threshold within a continuous time period.

[0018] A satellite timing system is used to unify the time reference of each UAV, and the local coordinate system data of each UAV is converted to a unified geographic coordinate system;

[0019] The spatiotemporal sequence is generated by combining real-time atmospheric parameter correction observation data.

[0020] Optionally, the spatiotemporal sequence is fused from multiple sources to construct a multidimensional wind field matrix, including:

[0021] A spatial interpolation algorithm is used to grid the wind field data of discrete observation points to generate continuous spatial distribution data. The spatial interpolation algorithm dynamically adjusts the interpolation weights in the vertical direction based on the wind shear index, which is obtained by fitting historical wind measurement data.

[0022] Wind field data missing time periods are filled in by a time-series prediction model to generate a continuous time series; the inputs of the time-series prediction model include the distribution characteristics of the dominant wind direction and the historical wind field change trend.

[0023] By integrating the continuous spatial distribution data with the continuous temporal sequence through a probabilistic inference algorithm, a multidimensional wind field matrix with a confidence level that meets a preset condition is generated; the multidimensional wind field matrix includes multidimensional data such as spatial coordinates, timestamps, wind field parameters, and confidence level ratings.

[0024] Optionally, wind energy potential modeling operations are performed based on the multidimensional wind field matrix to generate a wind energy potential prediction model, including:

[0025] Multidimensional wind field features are extracted from the multidimensional wind field matrix; the multidimensional wind field features include mean wind speed, wind direction frequency, turbulence intensity, and wind shear index; the feature extraction is performed based on preset safety boundary conditions, which are set according to historical maximum wind speed statistics.

[0026] A machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field characteristics to obtain the wind energy potential prediction model. The parameter settings of the machine learning algorithm meet the following optimization conditions: the number of nodes in the input layer matches the dimension of the multidimensional wind field characteristics; the output layer includes wind energy density and the corresponding probability distribution parameters; the training process is constrained by the turbulence intensity index to ensure that the model output conforms to the laws of fluid mechanics.

[0027] The effectiveness of the wind energy potential prediction model is evaluated through physical statistical verification and accuracy verification. The physical statistical verification includes fitting the wind energy distribution output by the model with the Weibull distribution to verify the rationality of the parameters. The accuracy verification includes ensuring that the model prediction error is below a set threshold through cross-validation.

[0028] Optionally, a machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field characteristics to obtain the wind energy potential prediction model, according to the following formula:

[0029]

[0030] The following constraints must be met:

[0031] y i -w T φ(x i -b≤∈+ξ i

[0032]

[0033] Where w is the weight vector of the multidimensional wind field features, b is the bias term; φ(xi) is the kernel function that maps the multidimensional wind field features xi to a high-dimensional space; yi is the target value of the wind field parameters; ∈ is the threshold of the insensitive loss function; C is the regularization parameter that controls the model complexity and error tolerance; ξi and ξi* are relaxation variables that allow samples to break through the ∈ band.

[0034] Optionally, based on the wind energy potential prediction model, predictions are made on the wind energy data to be predicted to obtain wind energy resource prediction results, including:

[0035] The wind energy data to be predicted is input into the wind energy potential prediction model to extract the target wind field characteristics; and the wind speed of the return period is fitted based on the extreme value distribution model to generate the wind speed extreme value prediction curve; the wind speed extreme value prediction curve includes the maximum wind speed value corresponding to several return periods and is marked with risk thresholds;

[0036] A wind power density map is generated based on the wind energy density distribution characteristics; the wind energy density distribution characteristics include wind energy density variation curves at vertical height levels and real-time atmospheric parameter correction observation data.

[0037] The wind turbine deployment restriction area is determined based on the set turbulence intensity threshold and the wind speed extreme value prediction results; the restriction area determination conditions include turbulence intensity exceeding the preset safety threshold and maximum wind speed exceeding the wind resistance limit of the wind turbine.

[0038] The wind energy resource prediction results are output, including the wind speed extreme value prediction curve, the wind power density map, and the wind turbine deployment restriction area.

[0039] According to a second aspect of the embodiments of this application, a wind energy resource prediction system is provided, the system comprising:

[0040] The data acquisition module is used to acquire three-dimensional wind field observation data of the target area;

[0041] The spatiotemporal sequence module is used to perform spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence.

[0042] The wind field matrix module is used to perform multi-source data fusion on the spatiotemporal sequence to construct a multi-dimensional wind field matrix.

[0043] The model building module is used to build a wind energy potential prediction model based on the multidimensional wind field matrix.

[0044] The prediction module is used to predict the wind energy data to be predicted based on the wind energy potential prediction model, and obtain the wind energy resource prediction result.

[0045] According to a third aspect of the present application, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first aspect above.

[0046] According to a fourth aspect of the embodiments of this application, a computer-readable storage medium is provided having computer-readable instructions stored thereon, the computer-readable instructions being executable by a processor to implement the method described in the first aspect above.

[0047] In summary, this application provides a wind energy resource prediction method, system, device, and storage medium. It involves acquiring three-dimensional wind field observation data of a target area; performing spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence; fusing multi-source data into the spatiotemporal sequence to construct a multi-dimensional wind field matrix; establishing a wind energy potential prediction model based on the multi-dimensional wind field matrix; and predicting the wind energy data to be predicted based on the wind energy potential prediction model to obtain wind energy resource prediction results. By constructing a machine learning model, it provides efficient, accurate, and reliable resource assessment for wind energy development in various terrain regions. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0049] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0050] Figure 1 This is a schematic diagram of the wind energy resource prediction method provided in the embodiments of this application;

[0051] Figure 2 A comparison chart of average wind speeds at different altitude levels provided for embodiments of this application;

[0052] Figure 3 This is a schematic diagram of wind measurement tower data provided in an embodiment of this application;

[0053] Figure 4 A schematic diagram of wind rose provided for an embodiment of this application;

[0054] Figure 5 This is a schematic diagram of a wind energy resource prediction system provided in an embodiment of this application;

[0055] Figure 6 This paper shows a structural diagram of an electronic device provided in an embodiment of this application;

[0056] Figure 7 A diagram of a computer-readable storage medium provided in an embodiment of this application is shown.

[0057] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0059] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0060] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0061] In this invention, unless otherwise explicitly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0062] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0063] Figure 1 This application illustrates a wind energy resource prediction method provided by an embodiment of the present application. The method includes:

[0064] Step 101: Obtain three-dimensional wind field observation data for the target area;

[0065] Step 102: Perform spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence;

[0066] Step 103: Perform multi-source data fusion on the spatiotemporal sequence to construct a multi-dimensional wind field matrix;

[0067] Step 104: Establish a wind energy potential prediction model based on the multidimensional wind field matrix;

[0068] Step 105: Based on the wind energy potential prediction model, predict the wind energy data to be predicted to obtain the wind energy resource prediction results.

[0069] In one possible implementation, step 101 involves acquiring three-dimensional wind field observation data of the target area, including:

[0070] A layered flight path planning instruction is generated based on the terrain data of the target area and the vertical gradient data of the wind measurement tower. The vertical gradient data of the wind measurement tower includes wind energy parameters at several preset heights within a preset height range. A UAV formation is invoked to perform cruise operations at several vertically spaced height layers within the preset height range according to the layered flight path planning instruction, and the cruise speed is maintained within a preset range. The UAV formation includes several UAVs equipped with meteorological sensors. Three-dimensional wind field observation data is collected in real time by each UAV, and the collection time, spatial coordinates, and attitude data are recorded. The sampling density is dynamically adjusted according to the vertical distribution law of wind speed, wherein the sampling frequency of the high-altitude layer is set to be greater than that of the low-altitude layer.

[0071] By conducting multi-level cruises within a preset altitude range (e.g., 50m-200m), vertical wind speed and direction gradient data are acquired, solving the problem that traditional single-point anemometer towers cannot capture wind shear index. Combined with terrain data, the flight path is dynamically adjusted to ensure complete coverage of areas such as ridges and canyons by the drone formation, avoiding data blind spots caused by terrain obstruction. This provides high-resolution, high-reliability input data for wind energy resource assessment, supporting wind turbine selection, layout optimization, and risk warning, ensuring maximum return on investment for wind farms.

[0072] Multiple drones collect data in parallel, covering a larger spatial area within the same timeframe (e.g., a 15km × 15km area in just 36 hours), increasing efficiency by 5 times compared to traditional mobile wind farms. Based on the vertical distribution of wind speed (where wind speed changes drastically near the ground), the sampling density is dynamically adjusted: low altitude layer (0-100m): sampling interval 50m, frequency 1Hz, capturing turbulence intensity I. TTransient changes of σv / μv; High-altitude layer (100-200m): Sampling interval 20m, frequency 10Hz, accurately quantifying the vertical gradient of upper-altitude wind energy density P = 0.5ρv³. Using UAV formations equipped with meteorological sensors, simultaneous acquisition of three-dimensional wind speed, wind direction, temperature, and humidity parameters at multiple altitudes (e.g., 50m / 100m / 150m) is achieved, capturing key indicators such as vertical wind shear and turbulence intensity.

[0073] Sampling resources are dynamically allocated based on wind field characteristics (e.g., high-altitude wind speed increases logarithmically with altitude), prioritizing the densification of high-altitude data acquisition (e.g., sampling point density at 200m altitude is 3 times higher than at 50m altitude) to ensure that the vertical resolution of the wind energy assessment model input data meets the IEC 61400-12-1 standard (vertical layer interval ≤ 50m). Based on real-time wind speed data feedback, trajectory replanning is automatically triggered (e.g., if a sudden wind speed change exceeding 3σ is detected, turbulence intensity I is temporarily increased). T (Sampling frequency of >15% of the area). For terrain-undulating areas such as mountains and coastlines, layered trajectory planning ensures data coverage without blind spots, overcoming the limitations of traditional single-point observation by wind towers.

[0074] In one possible implementation, step 102 involves performing a spatiotemporal alignment process on the three-dimensional wind field observation data to generate a spatiotemporal sequence, including:

[0075] Based on preset physical threshold ranges and anemometer data setting conditions, abnormal data points in the three-dimensional wind field observation data are removed; the physical thresholds include wind speed, wind direction, and temperature; the anemometer data setting conditions include setting the data missing rate to be lower than a set missing rate threshold within a continuous time period; a satellite timing system is used to unify the time reference of each UAV, and the local coordinate system data of each UAV is converted to a unified geographic coordinate system; the spatiotemporal sequence is generated by combining real-time atmospheric parameters to correct the observation data.

[0076] By setting physical thresholds such as wind speed (0-60m / s) and temperature (-30℃-60℃), invalid data caused by sensor noise and communication packet loss is filtered out to ensure the physical rationality of the input data. A missing rate threshold (e.g., <5%) is set for continuous time periods based on wind tower data to avoid data gaps caused by equipment failure or weather interference affecting model training. The time reference error of multiple UAVs is controlled to ≤1ms using a GPS / BeiDou satellite timing system to eliminate timing misalignments caused by asynchronous sampling (e.g., phase deviation during gust events). The local coordinate system (Body Frame) data of each UAV is converted to the WGS84 or UTM unified geographic coordinate system to compensate for the influence of UAV attitude angles (pitch / roll) on wind speed vector measurement (correction ≥3% when tilt angle > 5°). Multi-source heterogeneous data is transformed into a unified spatiotemporal sequence (timestamp-spatial coordinates-wind field parameters) to provide standardized input for subsequent data fusion and modeling. It provides high-quality, highly consistent input data for wind energy potential prediction models, ensuring that wind speed prediction errors are ≤0.5m / s (IEC L2 standard), supporting accurate wind farm planning and risk assessment.

[0077] In one possible implementation, in step 103, multi-source data fusion is performed on the spatiotemporal sequence to construct a multidimensional wind field matrix, including: using a spatial interpolation algorithm to perform gridding processing on the wind field data of discrete observation points to generate continuous spatial distribution data; the spatial interpolation algorithm dynamically adjusts the interpolation weights in the vertical direction based on the wind shear index, which is obtained by fitting historical wind measurement data; filling in the missing time period wind field data through a time series prediction model to generate a continuous time series; the input of the time series prediction model includes the dominant wind direction distribution characteristics and historical wind field change trends; integrating the continuous spatial distribution data and the continuous time series through a probabilistic inference algorithm to generate a multidimensional wind field matrix with a confidence level reaching a preset condition; the multidimensional wind field matrix includes multidimensional data such as spatial coordinates, timestamps, wind field parameters, and confidence level ratings.

[0078] Discrete UAV observation point data (e.g., 50m / 100m / 150m altitude layers) are converted into a continuous spatial distribution (100m×100m grid resolution) using Kriging interpolation or Inverse Distance Weighted (IDW) algorithms to address data sparsity issues in complex terrain areas. The interpolation weights are dynamically adjusted based on the wind shear index fitted from historical data to accurately recreate wind speed surge characteristics. Using LSTM (Long Short-Term Memory) or ARIMA models, data missing due to equipment failure or severe weather is predicted and filled in based on the dominant wind direction distribution (e.g., northwest winds accounting for 60%) and historical wind speed variation trends (e.g., daily periodic fluctuations of ±2m / s).

[0079] This step follows a technical chain of spatial interpolation → temporal prediction → probabilistic inference → matrix construction. Spatial interpolation algorithms (such as Kriging interpolation) are used to grid the discrete observation data, generating a continuous spatial distribution and addressing the insufficient coverage problem caused by sparse sampling. Temporal prediction models (such as LSTM) are used to fill in missing data periods caused by weather or equipment failure, constructing a continuous time series. Bayesian inference integrates multi-source data to generate a multi-dimensional matrix containing confidence ratings, providing reliability indicators for subsequent modeling.

[0080] In one possible implementation, step 104 involves performing a wind energy potential modeling operation based on the multidimensional wind field matrix to generate a wind energy potential prediction model, including:

[0081] Multidimensional wind field features are extracted from the multidimensional wind field matrix. These features include mean wind speed, wind direction frequency, turbulence intensity, and wind shear index. The feature extraction is performed based on preset safety boundary conditions, which are set according to historical maximum wind speed statistics. A machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field features to obtain the wind energy potential prediction model. The parameter settings of the machine learning algorithm meet the following optimization conditions: the number of nodes in the input layer matches the dimension of the multidimensional wind field features; the output layer includes wind energy density and its corresponding probability distribution parameters; the training process is constrained by the turbulence intensity index to ensure that the model output conforms to the laws of fluid mechanics; the effectiveness of the wind energy potential prediction model is evaluated through physical statistical verification and accuracy verification. The physical statistical verification includes fitting the wind energy distribution output by the model with a Weibull distribution to verify the rationality of the parameters; the accuracy verification includes using cross-validation to ensure that the model prediction error is below a set threshold.

[0082] The mean wind speed, wind direction frequency, turbulence intensity (IT = σv / μv), and wind shear index (α = ln(v2 / v1) / ln(z2 / z1)) are extracted from the multidimensional matrix to construct the model input feature set. A safety boundary (e.g., vmax_hist = 50 m / s) is set based on historical maximum wind speed statistics to exclude abnormal areas exceeding the safety threshold (e.g., data with wind speeds > 50 m / s are not included in training). The SVR algorithm is used to establish a mapping between wind energy density P = 0.5ρv3 and the multidimensional features to capture the wind field variation patterns under complex terrain.

[0083] Through a technical chain of feature engineering → SVR modeling → physical verification → accuracy control, the model output conforms to the Weibull distribution law and is consistent with the statistical characteristics of historical data. Safety boundaries and confidence ratings help avoid high-risk areas, supporting wind turbine layout optimization and investment decisions. It provides a high-precision, high-reliability forecasting tool for wind energy development in complex terrain.

[0084] In one possible implementation, in step 104, a machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field characteristics to obtain the wind energy potential prediction model, according to the following formula:

[0085]

[0086] The following constraints must be met:

[0087] y i -w T φ(x i -b≤∈+ξ i

[0088]

[0089] Where w is the weight vector of the multidimensional wind field features, b is the bias term; φ(xi) is the kernel function that maps the multidimensional wind field features xi to a high-dimensional space; yi is the target value of the wind field parameters; ∈ is the threshold of the insensitive loss function; C is the regularization parameter that controls the model complexity and error tolerance; ξi and ξi* are relaxation variables that allow samples to break through the ∈ band.

[0090] In one possible implementation, step 105 involves predicting the wind energy data to be predicted based on the wind energy potential prediction model to obtain wind energy resource prediction results, including:

[0091] The wind energy data to be predicted is input into the wind energy potential prediction model to extract the target wind field characteristics; and the wind speed is fitted based on the extreme value distribution model to generate a wind speed extreme value prediction curve; the wind speed extreme value prediction curve includes the maximum wind speed value corresponding to several return periods and is marked with risk thresholds; a wind power density map is generated according to the wind energy density distribution characteristics; the wind energy density distribution characteristics include the wind energy density change curve of the vertical height layer and real-time atmospheric parameter correction observation data; the wind turbine deployment restriction area is determined according to the set turbulence intensity threshold and the wind speed extreme value prediction results; the restriction area determination conditions include turbulence intensity exceeding the preset safety threshold and maximum wind speed exceeding the wind resistance limit of the wind turbine; the wind energy resource prediction results are output, the wind energy resource prediction results include the wind speed extreme value prediction curve, the wind power density map and the wind turbine deployment restriction area.

[0092] The maximum wind speed vmax at different return periods (50 years / 100 years) is predicted using an extreme value distribution model, marking high-risk areas exceeding the wind turbine's wind resistance limit. The mean wind speed μv and turbulence intensity IT output from the SVR model in step 104 are used as inputs to the extreme value model. The extreme value distribution parameters μ, σ, and ξ are dynamically calibrated using historical data and real-time predictions. Based on the wind speed v predicted by the SVR model and the real-time atmospheric parameter corrections from step 103, the corrected wind power density P = 0.5ρv³ is calculated, generating a spatially continuous visualization map. Vertical gradient optimization: Wind energy density curves at key heights such as 70m / 100m / 120m are extracted to guide tower height design and maximize energy capture efficiency. Composite risk areas are marked based on turbulence intensity thresholds (IT > 15%) and wind speed extreme value exceedances (vmax > vlimit) to avoid operational risks caused by the superposition of high turbulence and high wind speeds. Cost-benefit analysis data is provided when wind energy density enrichment areas overlap with high-risk areas. The GIS platform overlays wind power density heatmaps, extreme wind speed contour lines, and restricted area markers to create an intuitive site selection decision view. Dynamic report generation: Outputs PDF reports including theoretical power generation, equivalent full-load hours, and investment sensitivity analysis (such as IRR fluctuation curves with electricity prices). Through a technical chain of extreme prediction → resource quantification → risk marking → visualization output, areas exceeding IEC safety thresholds are avoided, and high-efficiency development zones with wind energy density exceeding set thresholds are accurately located. Visualization tools transform complex data into actionable site selection solutions, providing high-precision and high-reliability decision support tools for wind farm planning, ensuring that development plans achieve an optimal balance between safety and economy.

[0093] The wind energy resource prediction method provided in this application embodiment is described in detail below with reference to the accompanying drawings, and specifically includes the following stages:

[0094] The first phase involves the establishment of a drone formation system and data collection.

[0095] Step 1: System Construction and Sensor Configuration; Build a formation system consisting of multiple drones, each drone equipped with the following sensors: anemometer (measures three-dimensional wind speed); wind vane (records horizontal / vertical wind direction); temperature sensor; humidity sensor; Customize the drones according to mission requirements, supporting rapid installation and calibration of sensor modules.

[0096] Step 2: Flight strategy design: trajectory planning and flight parameter setting.

[0097] 1. Trajectory Planning: Design formation flight paths based on the terrain features (shape, area) of the target area, prioritizing coverage of areas sensitive to wind field changes. Employ a linear or triangular spatial layout to ensure complete boundaries of the measurement area without blind spots.

[0098] 2. Flight parameter setting: Altitude control: Prioritize improving the sampling density in the vertical direction and adopt a layered flight mode perpendicular to the horizontal plane (such as gradient climbing). Airspeed adjustment: Dynamically adjust the flight speed according to measurement accuracy and endurance requirements to balance data resolution and mission duration.

[0099] Obtain vertical profile data such as wind shear index and turbulence intensity through layered flight (such as at 50m intervals). Adopt an alternating hover-climb mode to improve data continuity in the vertical direction.

[0100] Step 3: Cooperative operation and data collection.

[0101] Divide the formation into multiple cooperative operation groups (for example, 3-5 drones in each group) to achieve the following functions: Spatial coverage: Each group synchronously collects data in sub-regions according to a preset strategy. Redundant backup: Deploy overlapping observation groups in key areas to ensure data reliability. Record the following data in real-time: Vector data of wind speed / wind direction (including timestamp); Scalar parameters of temperature and humidity; UAV position (GPS coordinates) and attitude angle (IMU data). Support dynamic grouping and reorganization to adapt to the mission requirements of complex terrains (such as mountains and coastlines). The flight layout can be switched to modes such as "snake" and "figure-eight" to cope with different measurement scenarios.

[0102] The second stage is the design and implementation of the data collection strategy.

[0103] Step 1: Track planning and cooperative formation deployment; Develop a multi-dimensional track planning scheme according to the terrain characteristics and wind field distribution law of the target area. For plain areas, adopt a grid scanning path and arrange equally spaced observation lines along the latitude and longitude directions; For complex terrains such as mountains or coastlines, adopt an adaptive track generation algorithm to enable the UAV formation to automatically adjust the flight path along the contour line or the dominant wind direction. The formation adopts a master-slave cooperative architecture, with the leading UAV dynamically planning the track, and 3-5 subordinate UAVs forming a "pin" formation to collect data synchronously at a fixed interval. Dynamically increase the track density in areas with sudden wind speed changes to ensure capturing turbulence details.

[0104] Step 2: Altitude layer control and vertical sampling; Divide the measurement airspace into the near-surface layer (0-200m), the transition layer (200-500m), and the high-altitude layer (500-900m), and implement a differential altitude control strategy. The near-surface layer adopts a dense climbing mode with a 20m vertical interval, and obtains high-resolution wind shear data through alternating hover-slow climb actions; The middle and high-altitude layers fly at a gradient with an interval of 50-100m, and cooperate with the dynamic compensation technology of the UAV attitude angle to ensure that the sensor is always facing the windward direction. During the flight, monitor the data of the barometric altimeter and the radar altimeter in real-time to correct the measurement deviation caused by terrain undulation.

[0105] Step 3: Spatiotemporal Synchronization and Data Acquisition; Establish an atomic clock timing system, controlling the time reference error of the entire formation to within 1 millisecond. After the sensors are activated, the anemometer collects three-dimensional vector wind speed at a frequency of 10Hz, and the temperature and humidity sensors simultaneously record environmental parameters. All data are appended with millisecond-level timestamps. Dynamically adjust the sampling strategy: use 1Hz as the basic sampling frequency during calm periods, and automatically switch to 10Hz high-frequency mode when turbulence characteristics are detected. Implement triple redundancy acquisition in key areas, with different UAVs observing the same spatial point in three different time periods. Data is transmitted back to the ground server in real time via 5G and satellite dual channels.

[0106] Step Four: Integrated Navigation and Dynamic Positioning; employing GPS and Inertial Navigation Unit (IMU) fusion positioning technology. The GPS module receives multi-band satellite signals and eliminates ionospheric errors through real-time dynamic differential (RTK) technology to obtain centimeter-level latitude and longitude coordinates. The IMU has a built-in six-axis sensor; the gyroscope continuously monitors the UAV's pitch, roll, and yaw angle changes, and the accelerometer records three-axis acceleration data. The navigation computer fuses the GPS absolute position and IMU relative displacement through an extended Kalman filter, outputting three-dimensional spatial coordinates with 0.1-meter accuracy and attitude angle data with 0.5° attitude angle, providing a precise spatial reference for wind field measurement.

[0107] Step 5: Formation Coordination and Task Optimization; Constructing a Hierarchical Communication Network: Within the formation, a Time Division Multiple Access (TDMA) protocol is used, allocating communication time slots every 5 milliseconds to achieve millisecond-level command synchronization; between formations, a Mesh self-organizing network is built through relay drones, dynamically selecting the optimal transmission path. The task allocation system is based on reinforcement learning algorithms, analyzing the battery level, location, and data quality of each drone in real time, and dynamically adjusting the data acquisition tasks. When equipment abnormalities or communication interruptions are detected, a backup drone is automatically activated to take over the task, and local solid-state storage is used to cache data to ensure continuous data acquisition.

[0108] Step Six: Quality Control and Real-Time Verification;

[0109] A three-tiered data verification mechanism is implemented during flight: the first tier filters outomas using sensor physical thresholds (e.g., data with wind speeds > 40 m / s is automatically marked); the second tier cross-verifies data from synchronous observations by adjacent UAVs to eliminate systemic errors between devices; the third tier compares the data in real time with ground-based wind towers, triggering a go-around procedure when the data deviation at 120 meters altitude exceeds 0.7%. All valid data is appended with a SHA-256 hash checksum and distributedly stored using blockchain technology to ensure data integrity and traceability.

[0110] The third stage is data synchronization and fusion processing.

[0111] Step 1: Unify the spatiotemporal reference and synchronize the clock; establish a multi-level time synchronization system, adopting a joint time synchronization mechanism of GPS second pulse signal and atomic clock. Each UAV has a built-in high-precision clock module, achieving microsecond-level time alignment through the NTP protocol, ensuring that the sampling time deviation of the entire formation's sensors is less than 1 millisecond. For areas with satellite signal obstruction, the timekeeping function of the inertial navigation unit (IMU) is enabled, maintaining the cumulative clock error of no more than 10 milliseconds during a 30-minute disconnection.

[0112] Step 2: Multi-source data spatial calibration; achieving spatial benchmark unification through coordinate system transformation:

[0113] 1. Sensor coordinate system alignment: The anemometer measurement direction is dynamically corrected based on the UAV's attitude angles (pitch / roll / yaw) to eliminate observation errors caused by the tilt of the aircraft.

[0114] 2. Geographic coordinate mapping: Convert the data collected by each UAV in its local coordinate system to the WGS84 geodetic coordinate system and use UTM projection to eliminate the influence of the Earth's curvature.

[0115] 3. Elevation Correction: Combined with the digital elevation model (DEM), the barometric altimeter data is corrected for terrain, improving the vertical positioning accuracy to ±0.5 meters.

[0116] Step 3: Multimodal data fusion processing; Implementing a three-level data fusion process:

[0117] 1. Raw-level fusion: For observation data of the same spatial point at multiple time periods, a time-weighted average algorithm is used to generate a continuous time series, and the weights are dynamically allocated according to the sensor accuracy and sampling time.

[0118] 2. Feature-level fusion: Extract parameters such as wind speed spectrum features and turbulence intensity index, and integrate multi-unit data through Bayesian inference algorithm to generate a fused feature set with confidence ≥95%.

[0119] 3. Decision-level fusion: Combining historical data from ground-based wind measurement towers, a support vector regression (SVR) model is used to establish regional wind field mapping relationships, outputting a four-dimensional wind energy distribution matrix with a grid precision of 100 meters.

[0120] Step 4: Real-time data transmission and verification; Constructing an integrated air-ground transmission network:

[0121] 1. In-formation transmission: The TDMA (Time Division Multiple Access) protocol is adopted, and each UAV is allocated a dedicated 5ms transmission time slot to transmit compressed data packets (H.265 encoding, compression rate ≥50%) in real time.

[0122] 2. Cross-swarm relay: Mesh network is built by relay drones, the optimal transmission path is dynamically selected, and the packet loss rate is controlled below 0.1%.

[0123] 3. Quality closed-loop verification: Real-time calculation of the Pearson correlation coefficient between UAV data and ground-based wind measurement tower. When the deviation of data at a height of 120 meters exceeds 0.7%, an abnormal area re-flight command is automatically triggered.

[0124] Step 5: Spatiotemporal database construction; create a four-dimensional spatiotemporal index matrix to achieve efficient management of massive amounts of data:

[0125] 1. Time dimension: Sliced ​​and stored according to UTC timestamp, with the smallest time unit being 1 second.

[0126] 2. Spatial dimension: Geohash encoding is used to convert geographic coordinates into 64-bit strings, and an R-tree spatial index is built.

[0127] 3. Data cleaning rules: Set physical thresholds such as wind speed (0-60m / s) and temperature (-30℃-60℃) to automatically filter outliers and generate data quality reports.

[0128] By employing data synchronization and fusion technology, all data collected by drones are processed and merged synchronously to ensure data accuracy and consistency.

[0129] The fourth stage involves high-precision data fitting and wind field modeling.

[0130] Step 1: Spatiotemporal data preprocessing; Based on the spatiotemporal fusion data generated in the third stage, implement multi-level data optimization:

[0131] 1. Outlier cleaning: A dual filtering mechanism using the 3σ criterion and physical thresholds (wind speed 0-60m / s, temperature -30℃-60℃) is adopted to remove sensor noise and outliers.

[0132] 2. Missing value imputation: The K-nearest neighbor algorithm (K=5) is used to interpolate within the spatiotemporal neighborhood, prioritizing observation data from the same height layer and adjacent grid points.

[0133] 3. Data standardization: Z-score standardization is applied to characteristic variables such as wind speed and turbulence intensity to eliminate the influence of dimensional differences on the model.

[0134] Step Two: Feature Engineering and Variable Selection; Establish a feature correlation analysis system and select core modeling variables:

[0135] 1. Basic characteristics: wind speed (horizontal / vertical components), wind direction frequency, air density (calculated in real time based on temperature and humidity).

[0136] 2. Derivative characteristics: Wind power density (P=0.5ρv) 3ρ is the air density; wind shear index (α = ln(v2 / v1) / ln(z2 / z1)); turbulence intensity (IT = σv / μv, where σv is the standard deviation of wind speed)

[0137] 3. Statistical characteristics: Weibull distribution parameters (k,c), effective power generation hours, and maximum wind speed with a 50-year return period.

[0138] Key variables were selected and a feature subset was constructed using Pearson correlation coefficient (|r|>0.7) and random forest feature importance score (Gini index ≥0.15).

[0139] Step 3: Support Vector Regression (SVR) Modeling; Based on the user-provided SVR algorithm framework, establish a wind energy resource distribution model:

[0140] In one possible implementation, the high-precision data fitting algorithm employs a Support Vector Regression (SVR) model and optimizes the parameter learning strategy and the use of the penalty function. By adjusting these parameters, the SVR model can be optimized to achieve a balance between prediction accuracy and generalization ability, thereby enhancing the model's generalization ability and optimization effect. The underlying concepts and algorithms are as follows: SVR finds a hyperplane in the input space such that the difference between the mapping values ​​of training samples on this hyperplane and the target values ​​is minimized, while simultaneously keeping the error within a certain range. The user pre-defines this boundary, typically denoted by ∈(epsilon). Therefore, the core of SVR lies in maximizing the boundary to minimize the model error. The optimization problem of SVR can be expressed as the following formula:

[0141]

[0142] The following constraints must be met:

[0143] y i -w T φ(x i -b≤∈+ξ i

[0144]

[0145] Where w is the weight vector, b is the bias term, and w is the weight vector, b is the bias term, (φ(x) i )) is a function that maps the input vector (xi) to a higher-dimensional space. (y i ) is the target value. ∈ is the threshold of the insensitive loss function, set to twice the standard deviation of the wind speed observation error (typically ε = 0.3-0.5 m / s). C is the regularization parameter, controlling model complexity and error tolerance (C ∈ [1, 10]). (ξi )and It is a slack variable that allows some samples to break through the ε band, in order to deal with sudden turbulence in the wind field data.

[0146] Model training uses the SVRG (Stochastic Variance Reduced Gradient) optimizer to accelerate convergence, with a batch size of 256 and ≥1000 iterations.

[0147] In one possible implementation, the Stochastic Variance Reduced Gradient (SVRG) algorithm is used for gradient descent to improve the efficiency and convergence of parameter estimation by reducing the variance of gradient estimates. In wind resource assessment feature vector parameter estimation, the SVRG algorithm can more efficiently optimize the parameters of the Support Vector Regression (SVR) model. The core is reducing the variance of gradient estimates. In each iteration, the algorithm calculates a full gradient (using all samples) and a stochastic gradient (using one or a small batch of samples), and then updates the parameters using the difference between these two gradients. This method reduces the variance of gradient estimates in each iteration, thus accelerating convergence; there is an internal iteration process in each iteration. Before the internal iteration begins, the algorithm calculates the full gradient for the current parameter values. Then, the internal iteration uses the difference between this full gradient and the stochastic gradient to update the parameters. This method ensures that the gradient estimate in each iteration has a continuously decreasing upper bound on variance, thus achieving linear convergence. The specific parameter update specifications are as follows:

[0148]

[0149] Where (θ_t) is the parameter of the (t)th iteration, (θ_(t+1)) is the parameter of the (t+1)th iteration, and (α) is the learning rate. It is the gradient under parameters (θ_t) and random samples (ξ_t). In the parameters The gradient of the random sample (ξ_t). In the parameters The full gradient under the given conditions. Parameters It is resampled in each outer iteration, usually randomly selected at the beginning of each epoch.

[0150] The fifth stage is wind energy potential assessment and decision support.

[0151] This phase achieves quantitative assessment of wind energy resources through multi-scale data fusion and machine learning optimization. First, it integrates macro, meso, and micro-level data: at the macro level, it integrates global reanalysis climate fields (such as ERA5 data) to construct a regional wind climate background; at the meso level, it spatially interpolates and aligns UAV-fitted wind field data (100m resolution) with the output of mesoscale meteorological models (WRF); at the micro level, it combines high-frequency observations from wind towers and vertical profile data from lidar, overlaying digital elevation models to calculate terrain roughness correction coefficients. A wind energy density mapping relationship is established based on a support vector regression (SVR) model, with the optimization objective function being to minimize the sum of the weight vector magnitude and the slack variable penalty term (formula: min1 / 2||w||). 2 +C∑(ξ_i+ξ_i*)), where the regularization parameter C∈[1,10] is determined through grid search, and the insensitive loss threshold ε is set to 0.3m / s to balance prediction accuracy and generalization ability. The stochastic variance reduced gradient (SVRG) algorithm is used to accelerate model training, and parameter updates follow... The rule is to calculate the full gradient after every 1000 mini-batch iterations, and the learning rate α decays from 0.05 to 0.01 using a cosine annealing strategy to achieve linear convergence.

[0152] In the risk assessment phase, the maximum wind speed with a 50-year return period was calculated based on the Gumbel extreme value distribution, and high-risk areas exceeding the IEC III wind zone threshold (37.5 m / s) were marked. Simultaneously, a turbulence intensity heatmap (threshold 15%) was generated. The economic assessment introduced a development priority index (PI = 0.6P / P_max + 0.3T_eq / 4000 + 0.1 / I_T), dividing the region into two development zones: the first-level zone requires a wind power density > 400 W / m³. 2 Furthermore, the equivalent full-load hours must be >3000h, and for secondary zones, the requirement is >250W / m. 2 With a lifespan of >2500h, the wind turbine layout optimization employs the NSGA-II multi-objective genetic algorithm to maximize the total power generation after wake correction under a 5D×3D minimum spacing constraint, with the Pareto front convergence threshold controlled within 1%.

[0153] The final results are visualized through a 3D GIS platform, including wind power density heat maps at heights of 70 / 100 / 120 meters (color scale 0-800W / m). 2 The system includes isopleths of the maximum wind speed expected once every 50 years (intervals of 5 m / s) and development priority zoning layers. The decision support report automatically generates core indicators such as regional theoretical reserves and histograms of equivalent full-load hours distribution, along with electricity price sensitivity analysis curves (IRR trends with ±20% fluctuation in electricity prices) and discount rate-NPV matrix diagrams. This evaluation system has undergone 5-fold cross-validation; the model's predicted wind speed RMSE ≤ 0.5 m / s, and the correlation r of the meteorological tower data validates the system. 2With an accuracy of ≥0.95, it meets the L2 level accuracy requirement of the IEC 61400-12-1 standard and can provide a reliable basis for the site selection of GW-level wind farms.

[0154] Based on the obtained wind energy resource data, a wind energy resource assessment model is used to calculate and evaluate wind energy potential. Combining the fitted wind energy resource information and geographical location information, the amount of wind energy resources in the region is calculated. Based on the fitted wind energy resource meteorological data, a quantitative evaluation of the wind energy resources in the target area is completed. The evaluation results can be presented in charts or other forms to better understand and analyze the status of wind energy resources, assisting decision-makers in making informed decisions regarding wind energy investment and site selection. In one possible implementation, the wind energy resource assessment process combines macro-, meso-, and micro-scale data and models with ground meteorological station data for comprehensive evaluation. In another possible implementation, big data analytics and machine learning algorithms are used to process and analyze the data during the wind energy resource assessment process to evaluate the wind energy resource potential of the target area.

[0155] This invention provides a wind energy resource assessment method based on UAV formations. By using UAV formations to conduct measurements, it simultaneously collects wind energy resource data from multiple points and employs a high-precision data fitting algorithm to simulate wind energy resource data from a large-scale wind farm, thereby improving the accuracy and efficiency of wind energy resource assessment. This method has advantages such as high data acquisition efficiency, high data accuracy, high wind energy resource assessment accuracy, high wind farm design efficiency, and a good user experience, and has broad market prospects.

[0156] Specific Implementation Case: The site conditions and requirements for a planned mountain wind farm are as follows: The planned site is located in a low-altitude mountain area, with the elevation of the topographic map base between 700 and 900 meters, a relative elevation difference of approximately 150 meters; the planned site is situated in a V-shaped valley mouth, surrounded by mountains on three sides, exhibiting characteristics of forming a local high-pressure basin; the atmospheric environmental quality in the planned site area is good to slightly polluted; the annual prevailing wind direction is localized, with no significant external prevailing wind influence. The annual power generation utilization hours of the planned site wind farm must be no less than 2100 hours. The annual on-grid electricity generation of the planned site wind farm must be 6 million kWh. The total investment of the planned site wind farm must not exceed 5 million yuan. The annual operation and maintenance cost of the planned site wind farm must not exceed 0.8 yuan / kWh. The unit capacity construction cost of the planned site wind farm must not exceed 0.6 yuan / kW.

[0157] Based on the above conditions and requirements, the wind energy resource analysis of the planning site was carried out using the method of this invention as follows: (1) Using a UAV formation, 72 stations (one station every 1°*1°) within a 15km*15km range of the planning site were collected, with elevations of 100-900m and near-surface wind speed, wind direction, temperature, humidity, and other parameters at a height of 50m; (2) The collected data were processed and synchronized to eliminate missing and outlier values; (3) A custom high-precision fitting algorithm was used to numerically simulate the wind energy resources within a 15km*15km range to obtain a wind energy resource distribution map of the planning site area; (4) Based on the wind energy resource distribution map and other geographical information of the planning site, the wind energy resource quantity was calculated. After calculation and analysis, it was concluded that the total wind energy resource quantity of the planning site is approximately 7.779 million kW. If general measures are taken, the wind energy resource may be further increased by about 40%, reaching approximately 10.862 million kW.

[0158] This project involves the construction of one wind measurement tower, with parameters as shown in Table 1 below:

[0159] Table 1

[0160]

[0161] Anemometer tower configuration: Vertical gradient coverage of 7 height layers from 10m to 120m, with wind speed / direction sensors deployed across the entire gradient, and temperature, humidity and pressure sensors only installed at 10m and 120m.

[0162] This analysis covers data records from the date of the tower's erection to September 30, 2024. To ensure unbiased statistical representation within daily timeframes, only dates with complete 24-hour data were used. The final selected data period for the evaluation is shown in Table 2 below:

[0163] Table 2

[0164] Start time 2024 / 4 / 1 00:00:00 End time 2024 / 9 / 30 23:59:59

[0165] During 183 days of continuous monitoring (April 1, 2024 - September 30, 2024), 15,811,200 seconds of data were collected with zero missing data (100% completeness), providing a reliable benchmark for modeling.

[0166] The data records for this period are summarized in Tables 3 and 4 below: Table 3 can be used as a calibration anchor point for UAV data (the error between the data from the UAV at a height of 120m and the data from the wind measurement tower should be <0.7%).

[0167] Table 3

[0168] Time granularity 1s Total number of days 183 Total seconds 15,811,200 Total number of rows 15,811,200 Number of missing data rows 0 Data completeness 100％

[0169] Table 4

[0170]

[0171] Average wind speed at different altitude levels, for example Figure 2 Key trend: Average wind speed increases logarithmically with altitude, reaching 6.82 m / s at 120 m, a 78.5% increase compared to 10 m. Wind energy density (W / m³) 2 The increase was more significant: 120m (186.9) was 5.65 times that of 10m (33.09); Turbulence characteristics: the turbulence intensity at low altitude (10m: 0.213) was 1.87 times that at high altitude (120m: 0.114), confirming the need to focus on monitoring the 0-200m near-ground layer; verifying the necessity of layered sampling of UAV formations (significant increase in data above 70m).

[0172] like Figure 3 The wind shear index of the wind measuring tower is 0.2279, which conforms to the power law distribution (fitting equation y = 2.3178x^0.2279), indicating that the surface roughness in this area is moderate.

[0173] According to IEC standards, the measured maximum wind speed is generally 0.8 times the maximum wind speed occurring once every 50 years. According to European Standard II, the maximum wind speed occurring once every 50 years is 5 times the annual average wind speed. The maximum wind speeds occurring once every 50 years calculated according to these two standards are shown in Table 5 below:

[0174] Table 5

[0175]

[0176] Since the actual data used is for a period of less than one year, the above calculation results for the maximum wind speed that occurs once every 50 years are for reference only.

[0177] Figure 4 The wind rose diagram provided in this application embodiment is shown. The dominant wind direction shows significant concentration at all altitude levels, with the dominant wind direction at 120m being northwest (315°±22.5°). Vertical variation: At low altitudes (10-50m), wind direction dispersion is relatively high due to topographic influence; at high altitudes (100m+), wind direction stability is improved, which is beneficial for optimizing wind turbine layout. Layout recommendation: Designing the wind turbine axis direction according to the dominant wind direction at 120m can reduce wake loss by 12%-15%. Temperature, humidity, pressure, and air density statistics are as follows: Statistics at 10m altitude are shown in Table 6.

[0178] Table 6

[0179] Minimum value average value Maximum value Temperature (°C) 6.27 24.03 40.55 Relative humidity (%) 13.9 66.0 98.7 Atmospheric pressure (hPa) 992.2 1009.0 1028.3 <![CDATA[Air density (kg / m 3 )]]> 1.104 1.183 1.278

[0180] Statistics at a height of 120m are shown in Table 7:

[0181] Table 7

[0182]

[0183]

[0184] In summary, the decision recommendations include: Turbine selection: For the extreme wind speed requirement of 41.82 m / s at a height of 120 m, an IEC Category III turbine (safety threshold of 42.5 m / s) must be selected; Control strategy: Activate the dynamic pitch system in areas with turbulence intensity >15% (mainly distributed below 50 m); Power generation estimation: Based on an average wind speed of 6.82 m / s at 120 m and an air density of 1.177 kg / m³. 3 Calculations show that a single unit (5MW) can achieve an annual equivalent operating hours of 3,150 hours; Data fusion: Using wind tower data as a baseline, supplemented by UAVs: Horizontal direction: densifying observation points in ridge / canyon areas; Vertical direction: increasing sampling density by 50-120m layers (layers every 20m).

[0185] In summary, this application provides a wind energy resource prediction method. This method involves acquiring three-dimensional wind field observation data of a target area; performing spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence; fusing multi-source data into the spatiotemporal sequence to construct a multi-dimensional wind field matrix; establishing a wind energy potential prediction model based on the multi-dimensional wind field matrix; and predicting the wind energy data to be predicted based on the wind energy potential prediction model to obtain the wind energy resource prediction result. By constructing a machine learning model, this method provides efficient, accurate, and reliable resource assessment for wind energy development in various terrain regions.

[0186] Based on the same technical concept, embodiments of this application also provide a wind energy resource prediction system, such as... Figure 5 As shown, the system includes:

[0187] Data acquisition module 501 is used to acquire three-dimensional wind field observation data of the target area;

[0188] The spatiotemporal sequence module 502 is used to perform spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence;

[0189] The wind field matrix module 503 is used to perform multi-source data fusion on the spatiotemporal sequence to construct a multi-dimensional wind field matrix.

[0190] Model building module 504 is used to build a wind energy potential prediction model based on the multidimensional wind field matrix;

[0191] The prediction module 505 is used to predict the wind energy data to be predicted based on the wind energy potential prediction model, and obtain the wind energy resource prediction result.

[0192] This application also provides an electronic device corresponding to the method provided in the foregoing embodiments. Please refer to... Figure 6The diagram illustrates an electronic device provided by some embodiments of this application. The electronic device 20 may include: a processor 200, a memory 201, a bus 202, and a communication interface 203, wherein the processor 200, the communication interface 203, and the memory 201 are connected via the bus 202; the memory 201 stores a computer program that can run on the processor 200, and when the processor 200 runs the computer program, it executes the method provided by any of the foregoing embodiments of this application.

[0193] The memory 201 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one physical port (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0194] Bus 202 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory 201 is used to store programs. After receiving an execution instruction, the processor 200 executes the program. The method disclosed in any of the foregoing embodiments of this application can be applied to the processor 200, or implemented by the processor 200.

[0195] The processor 200 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 200 or by instructions in software form. The processor 200 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 201. The processor 200 reads the information in memory 201 and, in conjunction with its hardware, completes the steps of the above method.

[0196] The electronic devices and methods provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods they employ, operate, or implement.

[0197] This application also provides a computer-readable storage medium corresponding to the method provided in the foregoing embodiments. Please refer to... Figure 7 The computer-readable storage medium shown is an optical disc 30, on which a computer program (i.e., a program product) is stored, which, when run by a processor, executes the methods provided in any of the foregoing embodiments.

[0198] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made under the concept of the present invention using the contents of the present invention specification and drawings, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method of wind energy resource prediction, characterized in that, The method includes: Acquire three-dimensional wind field observation data of the target area, including generating layered trajectory planning instructions based on the terrain data of the target area and the vertical gradient data of the wind measuring tower; the vertical gradient data of the wind measuring tower includes wind energy parameters at several preset heights within a preset height range; The drone formation is invoked to perform cruises at several vertically spaced altitude layers within the preset altitude range according to the layered flight path planning instructions, while maintaining the cruise speed within a preset range; the drone formation includes several drones equipped with weather sensors; Real-time three-dimensional wind field observation data are collected by various drones, and the collection time, spatial coordinates and attitude data are recorded; the sampling density is dynamically adjusted according to the vertical distribution law of wind speed, with the sampling frequency of high altitude layer set to be greater than that of low altitude layer. The three-dimensional wind field observation data are subjected to spatiotemporal alignment processing to generate a spatiotemporal sequence; Multi-source data fusion is performed on the spatiotemporal sequence to construct a multi-dimensional wind field matrix; A wind energy potential prediction model is established based on the aforementioned multidimensional wind field matrix; Based on the wind energy potential prediction model, the wind energy data to be predicted is predicted, and the wind energy resource prediction results are obtained. The method for establishing a wind energy potential prediction model based on the multidimensional wind field matrix includes: extracting multidimensional wind field features from the multidimensional wind field matrix; the multidimensional wind field features include mean wind speed, wind direction frequency, turbulence intensity, and wind shear index; the extraction of multidimensional wind field features is performed based on preset safety boundary conditions, which are set according to historical maximum wind speed statistics; establishing a mapping relationship between wind energy density and the multidimensional wind field features using a machine learning algorithm to obtain the wind energy potential prediction model; the parameter settings of the machine learning algorithm meet the following optimization conditions: the number of nodes in the input layer matches the dimension of the multidimensional wind field features; the output layer includes wind energy density and corresponding probability distribution parameters; the training process is constrained by the turbulence intensity index to ensure that the model output conforms to the laws of fluid mechanics; the effectiveness of the wind energy potential prediction model is evaluated through physical statistical verification and accuracy verification; the physical statistical verification includes fitting the wind energy distribution output by the model with the Weibull distribution to verify the rationality of the parameters; the accuracy verification includes ensuring that the model prediction error is below a set threshold through cross-validation.

2. The method as described in claim 1, characterized in that, The three-dimensional wind field observation data undergoes spatiotemporal alignment processing to generate a spatiotemporal sequence, including: Based on a preset physical threshold range and the setting conditions of the wind measurement tower data, abnormal data points in the three-dimensional wind field observation data are removed; the physical thresholds include wind speed, wind direction and temperature; the setting conditions of the wind measurement tower data include setting the data missing rate to be lower than a set missing rate threshold within a continuous time period. A satellite timing system is used to unify the time reference of each UAV, and the local coordinate system data of each UAV is converted to a unified geographic coordinate system; The spatiotemporal sequence is generated by combining real-time atmospheric parameter correction observation data.

3. The method as described in claim 1, characterized in that, Multi-source data fusion is performed on the spatiotemporal sequence to construct a multi-dimensional wind field matrix, including: A spatial interpolation algorithm is used to grid the wind field data of discrete observation points to generate continuous spatial distribution data. The spatial interpolation algorithm dynamically adjusts the interpolation weights in the vertical direction based on the wind shear index, which is obtained by fitting historical wind measurement data. Wind field data missing time periods are filled in by a time-series prediction model to generate a continuous time series; the inputs of the time-series prediction model include the distribution characteristics of the dominant wind direction and the historical wind field change trend. By integrating the continuous spatial distribution data with the continuous temporal sequence through a probabilistic inference algorithm, a multidimensional wind field matrix with a confidence level that meets a preset condition is generated; the multidimensional wind field matrix includes multidimensional data such as spatial coordinates, timestamps, wind field parameters, and confidence level ratings.

4. The method as described in claim 1, characterized in that, A machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field characteristics, resulting in the wind energy potential prediction model, according to the following formula: It satisfies the following constraints: , 5. The method as described in claim 1, characterized in that, Based on the aforementioned wind energy potential prediction model, predictions are made on the wind energy data to be predicted, resulting in wind energy resource predictions, including: The wind energy data to be predicted is input into the wind energy potential prediction model to extract the target wind field characteristics; and the wind speed of the return period is fitted based on the extreme value distribution model to generate the wind speed extreme value prediction curve; the wind speed extreme value prediction curve includes the maximum wind speed value corresponding to several return periods and is marked with risk thresholds; A wind power density map is generated based on the wind energy density distribution characteristics; the wind energy density distribution characteristics include wind energy density variation curves at vertical height levels and real-time atmospheric parameter correction observation data. The wind turbine deployment restriction area is determined based on the set turbulence intensity threshold and the wind speed extreme value prediction results; the restriction area determination conditions include turbulence intensity exceeding the preset safety threshold and maximum wind speed exceeding the wind resistance limit of the wind turbine. The wind energy resource prediction results are output, including the wind speed extreme value prediction curve, the wind power density map, and the wind turbine deployment restriction area.

6. A wind energy resource prediction system, characterized in that, The system includes: The data acquisition module is used to acquire three-dimensional wind field observation data of the target area, including generating layered trajectory planning instructions based on the terrain data and vertical gradient data of the wind measuring tower of the target area; the vertical gradient data of the wind measuring tower includes wind energy parameters at several preset heights within a preset height range; calling a UAV formation to perform cruise at several vertically spaced height layers within the preset height range according to the layered trajectory planning instructions, and maintaining the cruise speed within a preset range; the UAV formation includes several UAVs equipped with meteorological sensors; collecting three-dimensional wind field observation data in real time through each UAV, and recording the collection time, spatial coordinates and attitude data; dynamically adjusting the sampling density according to the vertical distribution law of wind speed, wherein the sampling frequency of the high-altitude layer is set to be greater than that of the low-altitude layer; The spatiotemporal sequence module is used to perform spatiotemporal alignment processing on the three-dimensional wind field observation data to generate a spatiotemporal sequence. The wind field matrix module is used to perform multi-source data fusion on the spatiotemporal sequence to construct a multi-dimensional wind field matrix. A model building module is used to build a wind energy potential prediction model based on the multidimensional wind field matrix. The building of the wind energy potential prediction model based on the multidimensional wind field matrix includes: extracting multidimensional wind field features from the multidimensional wind field matrix; the multidimensional wind field features include mean wind speed, wind direction frequency, turbulence intensity, and wind shear index; the extraction of multidimensional wind field features is performed based on preset safety boundary conditions, which are set according to historical maximum wind speed statistics; a machine learning algorithm is used to establish a mapping relationship between wind energy density and the multidimensional wind field features to obtain the wind energy potential prediction model; the parameter settings of the machine learning algorithm meet the following optimization conditions: the number of input layer nodes matches the dimension of the multidimensional wind field features; the output layer includes wind energy density and corresponding probability distribution parameters; the training process is constrained by the turbulence intensity index to ensure that the model output conforms to the laws of fluid mechanics; the effectiveness of the wind energy potential prediction model is evaluated through physical statistical verification and accuracy verification; the physical statistical verification includes fitting the wind energy distribution output by the model with a Weibull distribution to verify the rationality of the parameters; the accuracy verification includes using cross-validation to ensure that the model prediction error is below a set threshold. The prediction module is used to predict the wind energy data to be predicted based on the wind energy potential prediction model, and obtain the wind energy resource prediction result.

7. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method as claimed in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores computer-readable instructions that can be executed by a processor to implement the method as described in any one of claims 1-5.

Images