A wind field safety management system and method based on meteorological data analysis

Abstract

This invention relates to the field of wind farm safety management and discloses a wind farm safety management system and method based on meteorological data analysis. The system includes collecting wind speed, wind direction, wind shear, turbulence, topography, and sensitive locations; mapping sensitive locations to a weighted matrix; continuously processing and combining these data to form a wind farm safety status dataset; generating micro-region wind force response maps using numerical simulations and historical measurement data; dividing the wind farm into micro-regions that can be dynamically updated with real-time meteorological changes; calculating the wind flow parameters for each micro-region; mapping the micro-region wind flow to each wind turbine location; establishing a wind turbine structural capacity model and determining the controllable power range; establishing a wind flow flexible deformation tensor; deriving the wind flow redistribution and corresponding energy field under candidate power conditions; constructing an energy minimization optimization model with constraints on sensitive locations, equipment capacity, and wind shear; solving for the optimal wind turbine power; and generating a unified wind flow optimization scheme. This invention has the advantage of improving wind flow guidance effectiveness.

Description

Technical Field

[0001] This invention relates to the field of wind farm safety management, specifically a wind farm safety management system and method based on meteorological data analysis. Background Technology

[0002] During wind farm operation, the power output and structural safety of wind turbines are significantly affected by the complex meteorological conditions within the wind farm. Existing technologies typically rely on empirical formulas or single-unit load models for power scheduling and turbine safety management, which struggles to comprehensively reflect the spatial distribution characteristics of various factors such as wind speed, wind direction, wind shear, and turbulence within a micro-region. Furthermore, wind turbines within the wind farm experience shielding effects, eddy current interference, and localized wind pressure disturbances. Current methods lack systematic quantitative analysis of airflow within the turbine array, failing to achieve dynamic monitoring and control of localized airflow and energy distribution in micro-regions. In addition, the impact of sensitive locations and complex terrain on localized wind pressure and the stress on turbine blades and towers is not fully considered, making it difficult to generate comprehensive quantitative data for turbine power optimization. Traditional wind farm management also lacks mechanisms for dynamic micro-region division and airflow prediction that combine real-time meteorological data with historical measurement data, making it impossible to optimize the overall turbine power distribution while ensuring safety. Therefore, designing a wind farm safety management system and method based on meteorological data analysis to improve airflow guidance is essential. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a wind field safety management system and method based on meteorological data analysis, which has the advantage of improving wind flow guidance effect and solves the problems mentioned in the background technology.

[0004] To achieve the aforementioned goal of improving wind flow guidance, this invention provides the following technical solution: a wind field safety management method based on meteorological data analysis, comprising the following steps: Collect wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculate energy terms according to wind kinetic energy, wind shear energy and turbulent pulsation energy. Map sensitive locations into weighted matrices, process them continuously and combine them to form a wind field safety status dataset. Based on the wind field safety status dataset, and using numerical simulation and historical measurement data to generate micro-region wind force response maps, spatial interpolation and gradient analysis of the wind flow field are performed. The wind field is divided into micro-regions that can be dynamically updated with real-time meteorological changes, and the wind flow parameters of each micro-region are calculated to construct the micro-region wind flow structure. Based on the micro-area airflow structure, the micro-area airflow is mapped to the location of each fan, the diversion pressure, local load and aerodynamic disturbance are calculated, the fan structural capacity model is established and the controllable power range is determined, and the fan diversion pressure sequence is generated. Based on the wind turbine diversion pressure sequence, a flexible deformation tensor of airflow is established to deduce the airflow redistribution and corresponding energy field under candidate power conditions, thus forming an airflow redistribution data field. Based on wind farm safety status data, micro-area airflow structure, wind turbine diversion pressure, and airflow redistribution data field, an energy minimization optimization model with sensitive location, equipment capacity, and wind shear constraints is constructed. The optimal wind turbine power is obtained by solving the model, and a unified airflow optimization scheme is generated.

[0005] The preferred process for continuously processing and combining data to form a wind farm safety status dataset is as follows: By normalizing the wind shear intensity, turbulence fluctuation amplitude, and historical fault frequency at each sensitive location, a weight matrix is ​​constructed. Continuous interpolation is performed on the matrix, and the information distribution is smoothed using three-dimensional spatial interpolation methods; By superimposing and combining different wind speed levels, wind direction changes, and terrain effects, a complete wind field safety status dataset is generated.

[0006] Preferably, the process of generating micro-area wind response maps using numerical simulation and historical measurement data is as follows: The weighted matrix information, energy terms, and continuous spatial data of each sensitive location in the wind farm safety status dataset are used as the initial conditions for the microgrid points. Based on the wind energy distribution of micro-regions and the weight of sensitive locations, computational fluid dynamics is used to calculate the theoretical wind pressure response of each micro-region under different wind speeds, wind directions and wind shear conditions. At the same time, the simulation results are corrected and dynamically adjusted by combining the actual wind pressure records of historical measurement data. The generated micro-area wind pressure response is mapped to a micro-area grid network to form a quantitative micro-area wind force response benchmark matrix corresponding to the wind field safety state dataset; The quantized micro-region wind response baseline matrix is ​​normalized and coupled with the original wind field safety state dataset to form a micro-region wind response map.

[0007] Preferably, the process of constructing the micro-area airflow structure is as follows: The wind response map of a micro-region is mapped onto a micro-region grid network. Using the wind pressure response, wind speed gradient and wind direction change information recorded in the map, three-dimensional spatial interpolation and gradient calculation are performed on the wind flow field of each micro-region. By combining the historical response characteristics of different wind speeds and wind directions in the micro-region wind response map, the micro-region boundary is defined, and the micro-region units are dynamically divided based on local pressure distribution, turbulent fluctuations and topographic resistance. Within each micro-region, the main wind direction, local load intensity, and energy distribution are calculated using wind response map data to form a complete micro-region wind flow structure.

[0008] Preferably, the process for calculating the diversion pressure, local load, and aerodynamic disturbance is as follows: The complete micro-area airflow structure parameters are projected to the location of each fan using a spatial mapping algorithm; Based on the geometric parameters of the wind turbine blades, the tower height, and the rotation speed, the diversion pressure and local load distribution are calculated. Micro-scale simulations of the forces acting on the surfaces of wind turbine blades and towers are performed to generate local aerodynamic disturbance datasets.

[0009] Preferably, the process of generating the fan diversion pressure sequence is as follows: Based on the material properties of the wind turbine, the stiffness of the blades and the bending resistance of the tower, and combined with the micro-area local aerodynamic disturbance dataset, the maximum load that the wind turbine can withstand under different wind speeds, disturbances and local aerodynamic influences is calculated. By mapping the micro-diverting pressure and local aerodynamic disturbances to the wind turbine blades and tower, the safe and controllable power range under each power level is determined. The controllable power generation sequence of each wind turbine under different micro-region conditions is used to form a diversion pressure sequence.

[0010] Preferably, the process of establishing the airflow flexible deformation tensor based on the fan diversion pressure sequence is as follows: The controllable power of each wind turbine under different micro-region conditions and the corresponding diversion pressure sequence are numerically combined to generate a micro-region power-pressure matrix in time and space order; Based on the airflow pressure distribution and micro-area energy parameters in the matrix, the effects of micro-area airflow on blade stress, tower load, and energy transfer within the array are calculated step by step. The micro-area airflow effect is accumulated according to the spatial layout of the wind turbine array to form a flexible energy distribution map within the wind turbine array; Serial calculations are performed on different power combinations to record the energy transfer and offset of micro-area airflow between blades and tower, generating a quantifiable airflow flexible deformation tensor.

[0011] Preferably, the process of forming the wind redistribution data field is as follows: The wind flow flexible deformation tensor is mapped to the wind flow structure information of each micro-region, and the temporal distribution of wind speed, wind direction and local energy field of the micro-region is generated according to the controllable power sequence of the wind turbine. For each candidate power combination, the redistribution of airflow in micro-intervals is calculated step by step, including local wind speed gradient, turbulent kinetic energy change and micro-interval kinetic energy transfer; The energy density and airflow effect of each micro-zone are accumulated according to the spatial layout of the wind turbine array to form the mutual influence matrix between micro-zones, and the flow direction and local pressure distribution of the micro-zones are adjusted according to the boundary conditions. By integrating the accumulated energy changes and flow adjustment results in the micro-intervals, a complete wind redistribution data field is generated.

[0012] Preferably, the process of solving for the optimal fan power and generating a unified airflow optimization scheme is as follows: The local wind speed, wind direction, energy density and gradient information recorded in the micro-area airflow structure, wind turbine diversion pressure and airflow redistribution data field are mapped to the blade and tower bearing range of each wind turbine, and the energy response characteristics of each wind turbine under different airflow conditions are quantified. Based on wind farm safety status data and wind turbine controllable power range, an energy minimization objective function is established, and the micro-area energy accumulation effect of the wind redistribution data field is incorporated into the power optimization calculation. Weight constraints are applied to sensitive locations, and power upper limit constraints are applied to micro-regions with strong wind shear or high local turbulence. The constraints are adjusted in combination with the wind turbine equipment capacity and safety redundancy. The power of each wind turbine is iteratively adjusted using mixed integer programming and nonlinear optimization algorithms, and the energy accumulation of the micro-region is updated until the optimal power sequence is obtained through convergence. The solution results are compared and verified with the airflow redistribution data field to generate the final unified airflow optimization scheme.

[0013] A wind farm safety management system based on meteorological data analysis includes: Data acquisition module: Collects wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculates energy terms to generate a wind field safety status dataset; Micro-area analysis module: Based on safety status data, combined with numerical simulation and historical measurements, it generates micro-area wind response maps and divides micro-areas to construct micro-area wind flow structures; Pressure calculation module: Maps the micro-area airflow to the fan location, calculates the diversion pressure, local load and aerodynamic disturbance, determines the controllable power range of the fan and generates the diversion pressure sequence; Energy modeling module: Based on the diversion pressure sequence, the airflow flexible deformation tensor and airflow redistribution data field are extrapolated to quantify the relationship between micro-area airflow and fan energy; Power optimization module: Constructs an energy minimization optimization model with constraints on sensitive location, equipment capacity, and wind shear, solves and generates a unified airflow optimization scheme; Compared with existing technologies, the present invention provides a wind farm safety management system and method based on meteorological data analysis, which has the following beneficial effects: This invention collects and quantifies multi-source meteorological data, including wind speed, wind direction, wind shear, turbulence, topography, and sensitive locations. It then combines numerical simulations and historical measurement data to generate micro-region wind response maps, enabling dynamic micro-region division and precise wind flow structure modeling of the wind field. By mapping micro-region wind flow to wind turbine locations, it calculates diversion pressure, local loads, and aerodynamic disturbances, establishing a wind turbine structural capability model and a controllable power range sequence. This allows the wind turbine to achieve optimal power allocation under different micro-region conditions. Furthermore, based on the wind turbine diversion pressure sequence, it establishes a flexible deformation tensor for wind flow, deducing wind flow redistribution and corresponding energy fields under candidate power conditions, forming a complete wind flow redistribution data field, providing precise constraints for energy minimization optimization. This method not only effectively improves the rational distribution of wind flow within the wind turbine array and reduces the impact of local eddies and disturbances on the blades and tower, but also enhances the wind energy transfer efficiency between micro-regions through refined wind flow guidance, optimizing the overall energy utilization of the wind farm, improving wind farm operational stability and power generation performance. It also has the advantages of improving wind flow guidance effects and the coordinated operation level of the wind turbine array. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the method of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0015] 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 some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: Please refer to Figure 1 As shown in the embodiment of the present invention, a wind farm safety management method based on meteorological data analysis includes the following steps: S1: Collect wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculate energy terms according to wind kinetic energy, wind shear energy and turbulent pulsation energy. Map sensitive locations into weighted matrices, process them continuously and combine them to form a wind field safety status dataset.

[0017] The process of continuously processing and combining data in S1 to form the wind farm safety status dataset is as follows: By normalizing the wind shear intensity, turbulence fluctuation amplitude, and historical fault frequency at each sensitive location, a weight matrix is ​​constructed. Data cleaning and missing value imputation were performed on the wind shear intensity, turbulence fluctuation amplitude, and historical fault frequency collected at each sensitive location. The range normalization method was used to map each indicator to a unified dimension. The information entropy of each indicator was calculated according to the entropy method, and the initial weights were determined accordingly. Principal component analysis was used to reduce the dimensionality of the indicator correlation to adjust the weight allocation. The normalized indicator values ​​were multiplied by the corresponding weights and summarized on the spatial coordinates to obtain a weight matrix indexed by the sensitive location and arranged by grid points.

[0018] Continuous interpolation is performed on the matrix, and the information distribution is smoothed using three-dimensional spatial interpolation methods; Interpolation calculations are performed in three-dimensional space using a weight matrix as input. Appropriate interpolation methods are selected and semi-variogram analysis is conducted to determine geostatistical parameters. Kriging interpolation is preferred for spatial estimation to preserve spatial correlation. If necessary, inverse distance weighted interpolation or cubic spline interpolation is used as supplementary methods for different spatial scales. The interpolation process includes three steps: range modeling, weight calculation, and spatial estimation. Cross-validation is performed on the interpolation residuals, and the variogram model is adjusted or converted to logarithmic scaling based on the validation results. The interpolation results are a continuously distributed sensitivity field, which is stored as a three-dimensional data raster with a uniform grid resolution. Each cell in the raster contains spatial coordinates and continuous sensitivity values.

[0019] By superimposing and combining different wind speed levels, wind direction changes and terrain effects, a complete wind field safety status dataset is generated. The continuous sensitivity field is superimposed and combined with the wind speed level field, wind direction change field, and terrain influence field. Real-time wind speed data is divided into several levels according to preset wind speed segmentation rules and a level weight field is generated. At the same time, the short-term wind direction change frequency is statistically analyzed according to wind direction sector and a direction influence factor field is generated. The terrain exposure factor and slope aspect shading correction coefficient are calculated using digital elevation data as inputs to the terrain influence field. The fields are synthesized grid by grid according to time series using weighted superposition. The synthesis process synchronizes the time continuity and saves the time series label. The output is a wind field safety status dataset with timestamps.

[0020] S2: Based on the wind field safety status dataset, and using numerical simulation and historical measurement data to generate micro-region wind force response maps, perform spatial interpolation and gradient analysis on the wind flow field, divide the wind field into micro-regions that can be dynamically updated with real-time meteorological changes, calculate the wind flow parameters of each micro-region, and construct the micro-region wind flow structure.

[0021] The process of generating micro-regional wind response maps in S2 using numerical simulation and historical measurement data is as follows: The weighted matrix information, energy terms, and continuous spatial data of each sensitive location in the wind farm safety status dataset are used as the initial conditions for the microgrid points. The wind field safety status dataset is read in the form of a three-dimensional raster. The weighted values, wind kinetic energy, wind shear energy and turbulent pulsation energy of each sensitive location are interpolated to the corresponding grid points according to the micro-region grid resolution. The initial physical quantity field of each micro-region is constructed, including the initial wind speed vector field, the initial turbulent energy field and the initial sensitivity weight field. The altitude and terrain height are registered with the grid coordinates to generate the boundary height and terrain roughness distribution.

[0022] Based on the wind energy distribution of micro-regions and the weight of sensitive locations, computational fluid dynamics is used to calculate the theoretical wind pressure response of each micro-region under different wind speeds, wind directions and wind shear conditions. At the same time, the simulation results are corrected and dynamically adjusted by combining the actual wind pressure records of historical measurement data. For each micro-region, a geometric mesh is constructed and refined at representative wind speed ranges and wind direction sectors. The mesh refinement focuses on covering the area around sensitive locations and the blade influence zone. Boundary conditions are set in the solver, including upstream inflow profile, no-slip surface, outlet pressure conditions, and turbulent inlet parameters. An appropriate turbulence closure method, such as a two-equation turbulence model, is selected, and the time step or iterative convergence criterion is set. Steady-state or transient numerical solutions are performed to obtain the spatiotemporal distribution of wind pressure. The theoretical wind pressure obtained from the solution is compared point by point with historical wind pressure records. Deviation correction and data assimilation methods are used to adjust the inflow profile. If necessary, the solution settings are relinearized according to the time window to achieve consistency between simulation and observation.

[0023] The generated micro-area wind pressure response is mapped to a micro-area grid network to form a quantitative micro-area wind force response benchmark matrix corresponding to the wind field safety state dataset; The numerical solution results for each representative working condition are sampled and volume-weighted averaged at micro-grid points. Quantitative indicators such as time-averaged wind pressure, fluctuation amplitude, and transient peak value at the grid points are extracted. These indicators are arranged and combined according to the grid index to generate a response benchmark table and stored as a multi-time, multi-working-condition benchmark array according to spatial coordinates. Consistency checks are performed on the benchmark array, including spatial smoothing and time steady-state checks. Outlier values ​​are removed and the calculation methods of alternative values ​​are recorded. Finally, the benchmark array is converted into a storage format consistent with the wind field safety state dataset.

[0024] The quantized micro-region wind response baseline matrix is ​​normalized and coupled with the original wind field safety state dataset to form a micro-region wind response map. The values ​​of each wind pressure index in the baseline matrix are mapped to a unified scale using linear normalization or quantile normalization methods, while retaining the original time series labels and representative operating condition identifiers. The normalized response values ​​are fused point-to-point with the sensitivity weights, energy terms, and terrain correction coefficients in the wind field safety state dataset according to the micro-area grid coordinates. The final response value of each grid point under each operating condition is calculated using weighted superposition or multi-layer fusion rules, and a micro-area wind force response map with timestamps is generated.

[0025] The process of constructing the micro-regional airflow structure in S2 is as follows: The wind response map of a micro-region is mapped onto a micro-region grid network. Using the wind pressure response, wind speed gradient and wind direction change information recorded in the map, three-dimensional spatial interpolation and gradient calculation are performed on the wind flow field of each micro-region. The wind response map is registered with the micro-area grid points according to its spatial coordinates. The calibration steps include coordinate system unification, elevation registration, and time alignment. For each representative working condition in the map, wind pressure response, wind speed vector, and transient fluctuation index are extracted at the grid point location as initial samples. Spatial interpolation is performed on the three-dimensional grid. The interpolation process includes semivariogram estimation, spatial correlation test, and weight calculation. Interpolation operators that can maintain spatial correlation are preferred for estimation. If necessary, inverse distance weighting and spline interpolation are used locally to compensate for boundary effects. After the interpolation is completed, the numerical gradient of the wind speed component at the grid point is calculated. The gradient calculation uses the central difference quotient and combines it with the grid scale for error correction to obtain the three-dimensional wind speed gradient field and wind pressure gradient field.

[0026] By combining the historical response characteristics of different wind speeds and wind directions in the micro-region wind response map, the micro-region boundary is defined, and the micro-region units are dynamically divided based on local pressure distribution, turbulent fluctuations and topographic resistance. Using wind pressure response intensity, turbulence fluctuation intensity, and terrain drag coefficient at micro-grid points as criteria, initial candidate boundaries are determined by contour extraction and connected component analysis. Temporal stability judgment is introduced during boundary identification, and the continuity and repeatability of candidate boundaries are verified by historical response characteristics. Merging and segmentation rules are applied to candidate boundaries. The merging rule is based on the neighborhood similarity threshold, and the segmentation rule is triggered by local gradient abrupt changes or energy density faults. At the same time, time window averaging or lag mechanisms are used to suppress meaningless frequent divisions for short-term boundary jitter. Finally, a list of micro-region units is output according to local pressure distribution, turbulence fluctuation, and terrain occlusion characteristics.

[0027] Within each micro-region, the main wind direction, local load intensity, and energy distribution are calculated using wind response map data to form a complete micro-region wind flow structure. Vector averaging is performed on the wind speed vectors of all grid points within the micro-region to obtain the dominant flow direction. The variance and covariance of the wind speed components at each grid point are calculated to obtain the turbulent kinetic energy index. Dynamic pressure is calculated by multiplying the squared instantaneous velocity at each grid point by the air density. The dynamic pressure is then integrated on the corresponding 3D grid of the micro-region using either a unit area or unit volume ratio to obtain the local load intensity. Simultaneously, the load values ​​are projected and accumulated along the blade action surface direction and the tower outer surface direction to form the load distribution. After the load calculation is completed, based on the wind speed magnitude, turbulent fluctuation amplitude, and wind shear variable at the grid points, a load distribution is calculated using either a unit volume ratio and a unit volume ratio. The kinetic energy density per unit area is calculated, and the kinetic energy density is superimposed with the pulsating energy term to obtain the total energy density of the micro-region. Based on the gradient change of the total energy density along the dominant flow direction, the energy flux is calculated. The energy transfer path is described by the direction and amplitude of the energy flux. After obtaining the dominant flow direction, load distribution, total energy density and energy flux, the above quantity fields are integrated in a unified coordinate system and grid format to generate a multi-parameter wind flow description file containing the flow direction field, load field, velocity gradient field, turbulent energy field and energy flux field. All spatial distribution information is stored according to the micro-region boundary, thus forming a complete micro-region wind flow structure.

[0028] S3: Based on the micro-area airflow structure, map the micro-area airflow to each fan location, calculate the diversion pressure, local load and aerodynamic disturbance, establish the fan structural capacity model and determine the controllable power range, and generate the fan diversion pressure sequence.

[0029] The process of calculating the split pressure, local load, and aerodynamic disturbance in S3 is as follows: The complete micro-area airflow structure parameters are projected to the location of each fan using a spatial mapping algorithm; Coordinate registration is performed based on the three-dimensional coordinates of the wind turbine in the field and the coordinates of the micro-grid points. The wind speed vector, wind direction gradient, turbulence intensity, and kinetic energy density of the micro-grid points are interpolated onto the wind turbine inlet plane and rotor swept surface using inverse distance weighted interpolation or Kriging interpolation methods. During interpolation, layered interpolation is used in the vertical direction to reflect the height profile changes, and the near-ground grid data is corrected by combining terrain correction factors. The wind turbine inlet profile time series and corresponding turbulence statistics are obtained. The interpolation results are phase aligned and spectral verified to ensure that key pulsating frequency bands are preserved in the frequency domain, and finally, the wind turbine inlet profile dataset is formed.

[0030] Based on the geometric parameters of the wind turbine blades, the tower height, and the rotation speed, the diversion pressure and local load distribution are calculated. The inlet profile of the wind turbine is coupled with the wind turbine geometry. The blades are divided into segments according to the blade chord length and torsion. Instantaneous dynamic pressure is calculated based on the local inflow velocity and angle of attack, and the local lift and drag coefficients are obtained by referring to the blade aerodynamic characteristic table. The lift and drag are further decomposed and projected onto the blade chord length and normal direction to obtain the unit force. The inlet field is sliced ​​and mapped on the outer surface of the tower using a surface partitioning method, and the local surface pressure is calculated. When considering vortex interference and upstream wind turbine blocking effect, empirical disturbance correction is applied to the inlet profile or the velocity profile is adjusted based on Gaussian wake reduction. The adjusted velocity field is input into the blade and tower force calculation process to obtain the split pressure and local load distribution including wake effect.

[0031] Micro-scale simulation of the forces on the surface of wind turbine blades and tower is performed to form a local aerodynamic disturbance dataset. The blade and tower surfaces are discretized into several stress elements, and the instantaneous pressure and shear stress of each element are calculated in time series. The pressure is obtained by multiplying the local velocity square term by the air density and combining it with the pulsation correction coefficient. The shear stress is estimated by the near-wall turbulence pulsation parameters and the surface friction coefficient. For the working condition with eddy current excitation, the time-domain numerical integration method is used to accumulate the pulsation load to reflect the unsteady response. For high-frequency pulsation, spectral weighted filtering is used to separate the steady-state and pulsation components. The element stress of each time step is summarized by spatial index to generate the stress time series of the wind turbine blade and tower. At the same time, the local pressure coefficient, pulsation amplitude and spectral characteristics are recorded. Finally, the local aerodynamic disturbance dataset is output in the form of a structured data table.

[0032] The process of generating the fan diversion pressure sequence in S3 is as follows: Based on the material properties of the wind turbine, the stiffness of the blades and the bending resistance of the tower, and combined with the micro-area local aerodynamic disturbance dataset, the maximum load that the wind turbine can withstand under different wind speeds, disturbances and local aerodynamic influences is calculated. The elastic modulus, density, and yield strength of the blade material, as well as the bending stiffness and geometric section information of the tower steel structure, are obtained from the material database. Using the instantaneous pressure, shear stress, and turbulent pulsation energy in the micro-area aerodynamic disturbance dataset, finite element modeling is performed on the blade and tower elements. Loads are applied according to wind speed, wind direction, and disturbance conditions, and the stress and strain distribution of each element are calculated. The maximum allowable load is solved iteratively, while considering the fatigue limitations and local safety factors of the blades and tower, and finally the ultimate load that the wind turbine can withstand under different micro-area aerodynamic conditions is obtained.

[0033] By mapping the micro-diverting pressure and local aerodynamic disturbances to the wind turbine blades and tower, the safe and controllable power range under each power level is determined. The wind speed vector, pressure gradient, and turbulence fluctuation data of micro-area grid points are projected onto the blade chord length element and tower profile using a spatial mapping method. Based on the stress response characteristics of the blades and tower, instantaneous dynamic pressure and local disturbances are converted into blade torque and tower bending moment. Using the blade aerodynamic characteristic table and tower strength limit, the safe power range corresponding to the blade rotation speed and output power under different load conditions is screened to form the controllable range of each power level of each wind turbine under micro-area conditions. The data is then organized according to the time series.

[0034] The controllable power generation sequence of each wind turbine under different micro-zone conditions is used to form a diversion pressure sequence; For each wind turbine, the controllable power range is discretized according to the spatiotemporal distribution of micro-regions. The power levels are sorted and time-sequentialized by combining wind speed changes, wind direction disturbances and eddy current effects. For each time step, a shunt pressure record matching the blade output power and tower load is generated. At the same time, the corresponding micro-region location and airflow conditions are marked. Finally, the controllable power of each wind turbine in the entire wind farm under each micro-region condition is spliced ​​in time sequence to form a continuous shunt pressure sequence.

[0035] S4: Based on the wind turbine diversion pressure sequence, establish the airflow flexible deformation tensor, deduce the airflow redistribution and corresponding energy field under candidate power conditions, and form an airflow redistribution data field.

[0036] The process of establishing the airflow flexible deformation tensor based on the fan diversion pressure sequence in S4 is as follows: The controllable power of each wind turbine under different micro-region conditions and the corresponding diversion pressure sequence are numerically combined to generate a micro-region power-pressure matrix in time and space order; The controllable power and diversion pressure values ​​of each fan under each micro-zone condition are matched according to the micro-zone grid coordinates. The power and pressure data are quantified and the units are unified. The data of each fan are arranged and combined according to the time series and spatial layout to generate a matrix containing the power and pressure information of all micro-zones.

[0037] Based on the airflow pressure distribution and micro-area energy parameters in the matrix, the effects of micro-area airflow on blade stress, tower load, and energy transfer within the array are calculated step by step. By utilizing the pressure distribution data in the micro-region power-pressure matrix, combined with the geometric parameters and material mechanical properties of the blade chord length element and the tower section element, the force on each element is calculated using the finite element method. Furthermore, by integrating and accumulating the energy transfer path of the airflow within the array using the micro-region kinetic energy density and turbulent pulsation energy, the local load changes of the blades and tower and the energy transfer data within the array under the action of the airflow are gradually obtained.

[0038] The micro-area airflow effect is accumulated according to the spatial layout of the wind turbine array to form a flexible energy distribution map within the wind turbine array; Based on the relative position of the wind turbine in the array, the local airflow, pressure distribution and load data of each micro-region are mapped onto the array grid. The energy distribution of each micro-region is superimposed using a spatial accumulation algorithm to obtain the total energy effect of the blades and tower at each array position, and a flexible energy distribution map covering the entire wind field is generated.

[0039] Serial calculations are performed on different power combinations to record the energy transfer and offset of micro-area airflow between blades and tower, generating a quantifiable airflow flexible deformation tensor. For all micro-region power combinations, the power-pressure matrix is ​​calculated in chronological order, and the effects of micro-region airflow on the blades and tower are accumulated sequentially, including energy transfer paths and offsets. At the same time, the energy distribution, transfer direction and amplitude of each micro-region in three-dimensional space are quantified. By combining the time series with spatial coordinates, a structured and quantifiable airflow flexible deformation tensor is finally formed.

[0040] The process of forming the airflow redistribution data field in S4 is as follows: The wind flow flexible deformation tensor is mapped to the wind flow structure information of each micro-region, and the temporal distribution of wind speed, wind direction and local energy field of the micro-region is generated according to the controllable power sequence of the wind turbine. The generated airflow flexible deformation tensor is spatially mapped to the airflow structure information of each micro-region. The micro-region power-pressure data in the tensor is matched one by one with the micro-region wind speed, wind direction and energy distribution information. The time series data of each micro-region is associated with the wind turbine power series through coordinate alignment and index mapping to form the initial conditions of the micro-region airflow state.

[0041] For each candidate power combination, the redistribution of airflow in micro-intervals is calculated step by step, including local wind speed gradient, turbulent kinetic energy change and micro-interval kinetic energy transfer; For each candidate power combination, the wind speed, wind direction, and turbulent kinetic energy within the micro-region are calculated point by point in the time series order. The change in kinetic energy within the micro-region is determined by the wind speed gradient and kinetic energy change in the three-dimensional grid. The kinetic energy transfer is accumulated in the micro-region according to the wind flow direction to form the instantaneous wind flow redistribution data of the micro-region.

[0042] The energy density and airflow effect of each micro-zone are accumulated according to the spatial layout of the wind turbine array to form the mutual influence matrix between micro-zones, and the flow direction and local pressure distribution of the micro-zones are adjusted according to the boundary conditions. Based on the spatial layout of the wind turbine array, the energy density and airflow effect of each micro-region are spatially accumulated. The shielding effect of the upstream wind turbine on the downstream micro-region, the eddy current disturbance and local pressure change are included in the calculation. The boundary conditions of the micro-region are adjusted to ensure the flow continuity. The airflow direction and local pressure distribution are iteratively corrected to obtain the energy accumulation matrix and the interaction relationship between the micro-regions.

[0043] By integrating the accumulated energy changes and flow adjustment results in the micro-intervals, a complete wind redistribution data field is generated; The cumulative energy changes, wind effects, and flow adjustment results of all micro-regions are integrated, and the wind speed, wind direction, local kinetic energy density, and pulsating energy of each micro-region in three-dimensional space are serialized and structured to generate a complete wind redistribution data field.

[0044] S5: Based on wind farm safety status data, micro-area airflow structure, wind turbine diversion pressure and airflow redistribution data field, construct an energy minimization optimization model with sensitive location, equipment capacity and wind shear constraints, solve for the optimal wind turbine power and generate a unified airflow optimization scheme.

[0045] The process of obtaining the optimal wind turbine power and generating a unified airflow optimization scheme in S5 is as follows: The local wind speed, wind direction, energy density and gradient information recorded in the micro-area airflow structure, wind turbine diversion pressure and airflow redistribution data field are mapped to the blade and tower bearing range of each wind turbine, and the energy response characteristics of each wind turbine under different airflow conditions are quantified. The local wind speed, wind direction, kinetic energy density and gradient information recorded in the micro-area airflow structure, wind turbine diversion pressure and airflow redistribution data field are mapped to the blade and tower bearing range of each wind turbine through a spatial mapping method. This quantifies the local stress and energy response characteristics of each wind turbine under different micro-area power conditions and airflow states, forming a dynamic state data matrix of the wind turbine.

[0046] Based on wind farm safety status data and wind turbine controllable power range, an energy minimization objective function is established, and the micro-area energy accumulation effect of the wind redistribution data field is incorporated into the power optimization calculation. Based on wind farm safety status data and the controllable power range of wind turbines, the energy accumulation effect of the wind redistribution data field in each micro-region is incorporated into the mathematical model. An energy minimization objective function is defined, where the power variable corresponds to the controllable power of each wind turbine. The constraints include micro-region wind speed gradient, local wind shear, turbulence fluctuations, and equipment physical performance. The mapping relationship between micro-region energy and power is expressed numerically, providing a precise objective for the optimization algorithm.

[0047] Weight constraints are applied to sensitive locations, and power upper limit constraints are applied to micro-regions with strong wind shear or high local turbulence. The constraints are adjusted in combination with the wind turbine equipment capacity and safety redundancy. Weighted constraints are set for sensitive locations, and power upper limit constraints are applied to micro-regions with strong wind shear or high local turbulence. At the same time, the constraint parameters are adjusted in combination with the design limits and safety redundancy of the wind turbine blades, tower and overall structure. The constraint conditions are coupled with the objective function to form a constraint system that meets the requirements for safe operation of the wind farm.

[0048] The power of each wind turbine is iteratively adjusted using a mixed integer programming and nonlinear optimization algorithm to update the energy accumulation of the micro-region until the optimal power sequence is obtained through convergence. The wind turbine power sequence is initialized to the reference value or the result of the previous calculation. The current power sequence is substituted into the objective function to calculate the energy accumulation and wind flow effect of each micro-region. Based on the calculation results, the power allocation order and magnitude of each wind turbine are adjusted using mixed integer programming and nonlinear optimization algorithms. The updated power sequence is then substituted back into the micro-region energy accumulation calculation to correct the energy impact of each micro-region. The above iterative process is repeated to continuously update the power allocation and micro-region energy accumulation until the iteration converges or reaches the preset accuracy standard. Finally, the converged power sequence is taken as the global optimal wind turbine power output.

[0049] The solution results are compared and verified with the airflow redistribution data field to generate the final unified airflow optimization formula; The obtained wind turbine power results are compared with the wind speed, wind direction and energy density information of the corresponding micro-region of the airflow redistribution data field. The load-bearing capacity of each wind turbine under different micro-region conditions is checked to ensure that the structural and power constraints are met, ensuring data consistency and executability. The global optimal power combination is output and the final data record of the unified airflow optimization scheme is formed.

[0050] Example 2: As Figure 2 As shown, a wind farm safety management system based on meteorological data analysis includes: Data acquisition module: Collects wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculates energy terms to generate a wind field safety status dataset; Micro-area analysis module: Based on safety status data, combined with numerical simulation and historical measurements, it generates micro-area wind response maps and divides micro-areas to construct micro-area wind flow structures; Pressure calculation module: Maps the micro-area airflow to the fan location, calculates the diversion pressure, local load and aerodynamic disturbance, determines the controllable power range of the fan and generates the diversion pressure sequence; Energy modeling module: Based on the diversion pressure sequence, the airflow flexible deformation tensor and airflow redistribution data field are extrapolated to quantify the relationship between micro-area airflow and fan energy; Power optimization module: Constructs an energy minimization optimization model with constraints on sensitive location, equipment capacity, and wind shear, solves and generates a unified airflow optimization scheme; It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A wind farm safety management method based on meteorological data analysis, characterized in that, Includes the following steps: Collect wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculate energy terms according to wind kinetic energy, wind shear energy and turbulent pulsation energy. Map sensitive locations into weighted matrices, process them continuously and combine them to form a wind field safety status dataset. Based on the wind field safety status dataset, and using numerical simulation and historical measurement data to generate micro-region wind force response maps, spatial interpolation and gradient analysis of the wind flow field are performed. The wind field is divided into micro-regions that can be dynamically updated with real-time meteorological changes, and the wind flow parameters of each micro-region are calculated to construct the micro-region wind flow structure. Based on the micro-area airflow structure, the micro-area airflow is mapped to the location of each fan, the diversion pressure, local load and aerodynamic disturbance are calculated, the fan structural capacity model is established and the controllable power range is determined, and the fan diversion pressure sequence is generated. Based on the wind turbine diversion pressure sequence, a flexible deformation tensor of airflow is established to deduce the airflow redistribution and corresponding energy field under candidate power conditions, thus forming an airflow redistribution data field. Based on wind farm safety status data, micro-area airflow structure, wind turbine diversion pressure, and airflow redistribution data field, an energy minimization optimization model with sensitive location, equipment capacity, and wind shear constraints is constructed. The optimal wind turbine power is obtained by solving the model, and a unified airflow optimization scheme is generated.

2. The wind farm safety management method based on meteorological data analysis according to claim 1, characterized in that, The process of continuously processing and combining data to form a wind farm safety status dataset is as follows: By normalizing the wind shear intensity, turbulence fluctuation amplitude, and historical fault frequency at each sensitive location, a weight matrix is ​​constructed. Continuous interpolation is performed on the matrix, and the information distribution is smoothed using three-dimensional spatial interpolation methods; By superimposing and combining different wind speed levels, wind direction changes, and terrain effects, a complete wind field safety status dataset is generated.

3. The wind farm safety management method based on meteorological data analysis according to claim 2, characterized in that, The process of generating micro-region wind response maps using numerical simulations and historical measurement data is as follows: The weighted matrix information, energy terms, and continuous spatial data of each sensitive location in the wind farm safety status dataset are used as the initial conditions for the microgrid points. Based on the wind energy distribution of micro-regions and the weight of sensitive locations, computational fluid dynamics is used to calculate the theoretical wind pressure response of each micro-region under different wind speeds, wind directions and wind shear conditions. At the same time, the simulation results are corrected and dynamically adjusted by combining the actual wind pressure records of historical measurement data. The generated micro-area wind pressure response is mapped to a micro-area grid network to form a quantitative micro-area wind force response benchmark matrix corresponding to the wind field safety state dataset; The quantized micro-region wind response baseline matrix is ​​normalized and coupled with the original wind field safety state dataset to form a micro-region wind response map.

4. The wind farm safety management method based on meteorological data analysis according to claim 3, characterized in that, The process of constructing the micro-area airflow structure is as follows: The wind response map of a micro-region is mapped onto a micro-region grid network. Using the wind pressure response, wind speed gradient and wind direction change information recorded in the map, three-dimensional spatial interpolation and gradient calculation are performed on the wind flow field of each micro-region. By combining the historical response characteristics of different wind speeds and wind directions in the micro-region wind response map, the micro-region boundary is defined, and the micro-region units are dynamically divided based on local pressure distribution, turbulent fluctuations and topographic resistance. Within each micro-region, the main wind direction, local load intensity, and energy distribution are calculated using wind response map data to form a complete micro-region wind flow structure.

5. The wind farm safety management method based on meteorological data analysis according to claim 4, characterized in that, The process of calculating the diversion pressure, local load, and aerodynamic disturbance is as follows: The complete micro-area airflow structure parameters are projected to the location of each fan using a spatial mapping algorithm; Based on the geometric parameters of the wind turbine blades, the tower height, and the rotation speed, the diversion pressure and local load distribution are calculated. Micro-scale simulations of the forces acting on the surfaces of wind turbine blades and towers are performed to generate local aerodynamic disturbance datasets.

6. The wind farm safety management method based on meteorological data analysis according to claim 5, characterized in that, The process of generating the fan diversion pressure sequence is as follows: Based on the material properties of the wind turbine, the stiffness of the blades and the bending resistance of the tower, and combined with the micro-area local aerodynamic disturbance dataset, the maximum load that the wind turbine can withstand under different wind speeds, disturbances and local aerodynamic influences is calculated. By mapping the micro-diverting pressure and local aerodynamic disturbances to the wind turbine blades and tower, the safe and controllable power range under each power level is determined. The controllable power generation sequence of each wind turbine under different micro-region conditions is used to form a diversion pressure sequence.

7. The wind farm safety management method based on meteorological data analysis according to claim 6, characterized in that, The process of establishing the airflow flexible deformation tensor based on the fan diversion pressure sequence is as follows: The controllable power of each wind turbine under different micro-region conditions and the corresponding diversion pressure sequence are numerically combined to generate a micro-region power-pressure matrix in time and space order; Based on the airflow pressure distribution and micro-area energy parameters in the matrix, the effects of micro-area airflow on blade stress, tower load, and energy transfer within the array are calculated step by step. The micro-area airflow effect is accumulated according to the spatial layout of the wind turbine array to form a flexible energy distribution map within the wind turbine array; Serial calculations are performed on different power combinations to record the energy transfer and offset of micro-area airflow between blades and tower, generating a quantifiable airflow flexible deformation tensor.

8. The wind farm safety management method based on meteorological data analysis according to claim 7, characterized in that, The process of forming a wind redistribution data field is as follows: The wind flow flexible deformation tensor is mapped to the wind flow structure information of each micro-region, and the temporal distribution of wind speed, wind direction and local energy field of the micro-region is generated according to the controllable power sequence of the wind turbine. For each candidate power combination, the redistribution of airflow in micro-intervals is calculated step by step, including local wind speed gradient, turbulent kinetic energy change and micro-interval kinetic energy transfer; The energy density and airflow effect of each micro-zone are accumulated according to the spatial layout of the wind turbine array to form the mutual influence matrix between micro-zones, and the flow direction and local pressure distribution of the micro-zones are adjusted according to the boundary conditions. By integrating the accumulated energy changes and flow adjustment results in the micro-intervals, a complete wind redistribution data field is generated.

9. A wind farm safety management method based on meteorological data analysis according to claim 8, characterized in that, The process of obtaining the optimal fan power and generating a unified airflow optimization scheme is as follows: The local wind speed, wind direction, energy density and gradient information recorded in the micro-area airflow structure, wind turbine diversion pressure and airflow redistribution data field are mapped to the blade and tower bearing range of each wind turbine, and the energy response characteristics of each wind turbine under different airflow conditions are quantified. Based on wind farm safety status data and wind turbine controllable power range, an energy minimization objective function is established, and the micro-area energy accumulation effect of the wind redistribution data field is incorporated into the power optimization calculation. Weight constraints are applied to sensitive locations, and power upper limit constraints are applied to micro-regions with strong wind shear or high local turbulence. The constraints are adjusted in combination with the wind turbine equipment capacity and safety redundancy. The power of each wind turbine is iteratively adjusted using mixed integer programming and nonlinear optimization algorithms, and the energy accumulation of the micro-region is updated until the optimal power sequence is obtained through convergence. The solution results are compared and verified with the airflow redistribution data field to generate the final unified airflow optimization scheme.

10. A wind farm safety management system based on meteorological data analysis, applied to the method described in any one of claims 1-9, characterized in that, include: Data acquisition module: Collects wind speed, wind direction, wind shear, turbulence, topography and sensitive locations, and calculates energy terms to generate a wind field safety status dataset; Micro-area analysis module: Based on safety status data, combined with numerical simulation and historical measurements, it generates micro-area wind response maps and divides micro-areas to construct micro-area wind flow structures; Pressure calculation module: Maps the micro-area airflow to the fan location, calculates the diversion pressure, local load and aerodynamic disturbance, determines the controllable power range of the fan and generates the diversion pressure sequence; Energy modeling module: Based on the diversion pressure sequence, the airflow flexible deformation tensor and airflow redistribution data field are extrapolated to quantify the relationship between micro-area airflow and fan energy; Power optimization module: Constructs an energy minimization optimization model with constraints on sensitive location, equipment capacity, and wind shear, solves and generates a unified airflow optimization scheme.

Images