A method for calculating wind speed, power and electricity quantity in a wind farm

Abstract

The present invention discloses a method for calculating wind speed, power and electricity quantity in a wind farm, which relates to the technical field of wind speed survey in wind farms. High-resolution meteorological data of the wind farm area is obtained, the data is processed, and the wind speed data is corrected based on the parameters of the convective boundary layer to generate corrected wind speed data; a dynamic wind speed prediction model based on time series is constructed to predict the corrected wind speed data; a wind energy utilization optimization model is constructed based on the wake effect and layout synergy effect between wind turbines; for extreme weather conditions, a robust power prediction model is constructed based on historical weather data and an anomaly detection mechanism, and the prediction model parameters are dynamically adjusted in combination with an online learning mechanism to complete the calculation of the wind farm power. The present invention provides a systematic solution from wind speed correction to power prediction, which is applicable to the operation management of wind farms under complex environments and dynamic operating conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm wind speed survey, and particularly to a method for calculating wind speed, power and electricity in a wind farm. Background Art

[0002] Currently, as an important form of renewable energy utilization, the accurate prediction and optimization of the power generation capacity of wind farms are crucial. However, the dynamic changes in wind speed, the wake effect between wind turbines, and the uncertainty of wind speed and power under extreme weather conditions make the existing wind speed prediction and wind energy utilization methods have large deviations in complex environments. Traditional wind speed profile models are mainly based on static parameters, ignoring dynamic influencing factors such as turbulence intensity and mixing layer height, resulting in insufficient accuracy of wind speed calculation. In addition, the layout design of wind turbines mostly does not fully consider the wake effect and the synergy between wind turbines, which is likely to cause a reduction in wind energy utilization rate. For the prediction of wind speed and power under extreme weather conditions, existing methods are difficult to adapt to the rapidly changing environmental characteristics, resulting in low reliability of the power prediction model under abnormal conditions. Therefore, there is an urgent need for a comprehensive method that combines dynamic wind speed correction, optimized layout of wind turbines, and robust power prediction to improve the efficiency and stability of wind farm operation. Summary of the Invention

[0003] In view of the above existing problems, the present invention is proposed.

[0004] Therefore, the present invention provides a method for calculating wind speed, power and electricity in a wind farm, and proposes solutions to three key problems in the existing wind farm operation technology. First, in wind speed calculation, how to accurately describe the influence of factors such as turbulence intensity and mixing layer height on wind speed through a dynamic wind speed correction method, and achieve the dynamic adjustment of wind speed with height and time. Second, in the optimization of wind turbine layout, how to optimize the layout parameters of the turbines based on the wake effect and the synergy between the turbines, improve the wind energy utilization efficiency, and at the same time reduce the power loss caused by the wake. Finally, under extreme weather conditions, how to use real-time data and historical samples to construct a robust power prediction model, accurately predict the relationship between wind speed and power, and dynamically adjust the model to adapt to the rapidly changing extreme environment. By comprehensively solving the above problems, the present invention realizes the overall optimization from wind speed correction to power prediction.

[0005] To solve the above technical problems, the present invention provides the following technical solution. A method for calculating wind speed, power and electricity in a wind farm includes:

[0006] Obtain high-resolution meteorological data of the wind farm area, clean the data, process outliers and align them in space and time, and correct the wind speed data based on the parameters of the convective boundary layer to generate corrected wind speed data;

[0007] Construct a dynamic wind speed prediction model based on time series to predict the corrected wind speed data;

[0008] Construct a wind energy utilization optimization model based on the wake effect and layout synergy effect between wind turbines;

[0009] In response to extreme weather conditions, a robust power prediction model is constructed based on historical weather data and anomaly detection mechanism. The prediction model parameters are dynamically adjusted in combination with the online learning mechanism to complete the calculation of wind farm power.

[0010] As a preferred solution of the method for calculating wind speed, power and electricity of a wind farm described in the present invention, the high-resolution meteorological data includes wind speed, wind direction, air pressure and temperature data, and the data sources include satellite remote sensing, ground meteorological station observations and data collected by radar systems, and the data are aligned in time and space through an interpolation algorithm.

[0011] As a preferred solution of the method for calculating wind speed, power and electricity of a wind farm according to the present invention, the correction of wind speed data based on the parameters of the convective boundary layer includes the expression of the wind speed profile model being:

[0012] ,

[0013] in, is the height Basic wind speed under , without turbulence correction; represents the friction speed; represents the Karman constant; Indicates the surface roughness;

[0014] Introducing turbulence intensity and mixing layer height ,right Make dynamic adjustments:

[0015] ,

[0016] The effect of turbulence on wind speed is corrected by an exponential function, and the correction amplitude is proportional to the turbulence intensity and inversely proportional to the height of the mixing layer.

[0017] The cumulative effect of turbulence on wind speed depends not only on height but also on shear stress. and turbulence variance The role of is expressed in integral form as:

[0018] ,

[0019] in, is the similarity function, choose Calculation formula to express height Nonlinear effects on the stability length ; Represents the turbulent shear stress, combined with the friction velocity and the dynamic variation of aerodynamic parameters.

[0020] As a preferred embodiment of the method for calculating wind speed, power and electricity in a wind farm according to the present invention, wherein: the correction of wind speed data based on the parameters of the convective boundary layer includes integrating the corrected wind speed and the turbulent integral effect to obtain an improved dynamic correction wind speed formula:

[0021] ,

[0022] wherein, is the height and time the corrected wind speed at; is the friction velocity; is the von Kármán constant is the surface roughness; is the wind speed measurement height; is the time; is the turbulent intensity correction coefficient; is the mixing layer height, dynamically related to time ; is the shear stress; is the air density; is the turbulent effect adjustment coefficient; is the variance of the turbulent field at height ; is the stability length at time t; is the integration variable, representing the position from the surface height to the target height ;

[0023] As a preferred embodiment of the method for calculating wind speed, power and electricity in a wind farm according to the present invention, wherein: the construction of the dynamic wind speed prediction model based on time series includes obtaining the complex dynamic relationship of wind speed change by processing the corrected wind speed time series data based on the long short-term memory network LSTM or Transformer structure in deep learning, combined with local terrain features and seasonal climate patterns;

[0024] Define the input feature embedding matrix as , where all features are normalized so that their ranges are unified to ;

[0025] Time step and ​The correlation between them is calculated through the self-attention mechanism

[0026] ,

[0027] where:

[0028] Query matrix:

[0029] Key matrix:

[0030] Value matrix:

[0031] Weight matrix is a learnable parameter, is the dimension of the key vector;

[0032] For LSTM, the hidden state at time step is updated as:

[0033] ,

[0034] where, represents the Sigmoid activation function, are the learnable parameters of the network;

[0035] The final wind speed prediction value is mapped by the output of LSTM or Transformer through a fully connected layer as:

[0036] ,

[0037] where, are the weights and biases of the fully connected layer.

[0038] As a preferred solution of the wind farm wind speed, power and electricity calculation method described in the present invention, where: the construction of the wind energy utilization optimization model includes, based on the wake effect and the synergy effect of wind turbines, comprehensively calculating the influence of the airflow interaction between wind turbines on the overall power generation capacity, where the wake effect in the wind energy utilization optimization model is described by the Jensen model or the Gaussian wake model, where the Jensen model is based on the linear assumption of wake speed decay, and its wake speed decay formula is:

[0039] ,

[0040] where, represents the wind speed at a certain point in the wake area; is the wind speed not affected by the wake, is the induction factor of the unit, is the wake expansion coefficient, is the horizontal distance between the wind turbine and the wake influence point. The Gaussian model simulates the spatial decay characteristics of the wake through Gaussian distribution, and the expression is

[0041] ,

[0042] where, represents the wind speed at a point in the wake region in three-dimensional space; represents the horizontal lateral coordinate, which is used to describe the lateral distribution of the wake; is the position of the wake center axis, is the wake expansion width parameter, which reflects the spatial distribution of the wake;

[0043] The synergy effect optimizes the layout of wind turbines through a genetic algorithm. The objective function is defined as the total power generation of the group, which is expressed as

[0044] ,

[0045] where, represents the total power generation of all wind turbines in the wind farm; represents the total number of wind turbines in the wind farm; is the th wind turbine's power generation, is the corrected wind speed at the position of this wind turbine, is the wind direction angle of the wind turbine.

[0046] As a preferred solution of the wind farm wind speed, power and electricity calculation method described in the present invention, wherein: the construction of the robust power prediction model includes constructing multiple random split trees, mapping the samples to the leaf nodes through eigenvalue partitioning, and the anomaly score formula of the isolation forest is defined as:

[0047] ,

[0048] where, is the anomaly score of the sample , is the average path length required to be separated in all random trees, is the expected path length of the adjustment factor, representing the standard path length value when the number of sample points in the forest is n; is the total number of samples in the isolation forest; when (threshold), it is determined that the current data point belongs to extreme weather conditions, triggering the prediction mode switch; is the anomaly determination threshold of the isolation forest;

[0049] After detecting extreme weather, the robust power prediction model switches from the standard prediction mode to the extreme weather prediction mode, keeps the real-time data stream input uninterrupted, and reloads the parameterized model under extreme weather conditions according to the current environmental variables;

[0050] The extreme weather prediction mode adopts the random forest or Bayesian regression algorithm. For the random forest, the input is the features after anomaly detection , which represent wind speed, wind direction, temperature, and air pressure respectively. The non-linear relationship between wind speed and power is calculated through the splitting of multiple decision trees, and the power prediction value is given by the average output of all trees

[0051] ,

[0052] Among them, represents the power prediction value; is the th tree, is the total number of decision trees; For Bayesian regression, a regression model for power output is defined:

[0053] ,

[0054] Among them, is the weight vector, is the input feature vector is the noise term, and Bayes' formula is used to update the posterior distribution of the weights:

[0055] ,

[0056] Among them, and ]>are the input and output samples respectively, is the prior distribution.

[0057] As a preferred solution of the method for calculating wind speed, power, and electricity quantity in the wind farm described in the present invention, among them: the dynamic adjustment of the prediction model parameters by combining the online learning mechanism includes that the model dynamically updates the recent real-time data through a sliding window, and the window length is determined by the characteristics of the current wind speed fluctuation. When the input data arrives, the splitting parameters of the random forest or the prior weights of the Bayesian regression are adjusted through the online learning mechanism, and the formula is

[0058] ,

[0059] Among them, is the learning rate, is the updated i-th weight value, The \(i\)-th weight value before update is the gradient update amount of the current window, ensuring the dynamic adaptability of the model under extreme weather conditions.

[0060] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method for calculating wind farm wind speed, power, and electricity.

[0061] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of the method for calculating wind farm wind speed, power, and electricity.

[0062] Advantages of the present invention: By combining dynamic wind speed correction, optimized layout of wind turbines, and robust power prediction, the present invention comprehensively improves the operation efficiency and adaptability of wind farms. In terms of wind speed correction, based on the wind speed profile model, dynamic parameters such as turbulence intensity and mixing layer height are introduced, significantly improving the accuracy and spatio-temporal adaptability of wind speed calculation. In terms of the optimized layout of wind turbines, through wake effect modeling and genetic algorithm optimization, the positions and parameters of the turbines are reasonably adjusted, improving the wind energy utilization rate and reducing the wake effect. In terms of extreme weather power prediction, the isolation forest algorithm is used for anomaly detection, and high-precision prediction is achieved through random forest and Bayesian regression models, enhancing the stability and robustness of the power prediction model under extreme conditions. The present invention provides a systematic solution from wind speed correction to power prediction, which is applicable to the operation management of wind farms in complex environments and dynamic operating conditions, laying a solid foundation for the high-efficiency and intelligent utilization of wind energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0064] Figure 1 It is a schematic flowchart of a method for calculating wind farm wind speed, power, and electricity provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0065] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0066] In the following description, many specific details are set forth to facilitate a thorough understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0067] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that mutually excludes other embodiments.

[0068] The present invention is described in detail in conjunction with schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views showing the device structure are enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the protection scope of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0069] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationships indicated by terms such as "upper, lower, inner, and outer" are based on the orientation or positional relationships shown in the accompanying drawings. This is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0070] Unless otherwise clearly specified and defined in the present invention, the terms "mounted, connected, and coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or it may be indirectly connected through an intermediate medium, or it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0071] Example 1, referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides a method for calculating the wind speed, power, and electricity consumption of a wind farm, including:

[0072] S1: Obtain the high-resolution meteorological data of the wind farm area, clean the data, process outliers, and perform spatio-temporal alignment. Then, correct the wind speed data based on the parameters of the convective boundary layer to generate the corrected wind speed data.

[0073] S2: Construct a dynamic wind speed prediction model based on time series to predict the corrected wind speed data.

[0074] S3: Construct an optimization model for wind energy utilization based on the wake effect and layout synergy effect among wind turbines.

[0075] S4: For extreme weather conditions, construct a robust power prediction model based on historical weather data and anomaly detection mechanism, and dynamically adjust the prediction model parameters in combination with the online learning mechanism to complete the calculation of the wind farm power.

[0076] The high-resolution meteorological data includes wind speed, wind direction, air pressure, and temperature data. The data sources include data collected by satellite remote sensing, ground meteorological station observations, and radar systems, and spatio-temporal alignment of the data is performed through interpolation algorithms.

[0077] The correction of the wind speed data based on the parameters of the convective boundary layer includes that the expression of the wind speed profile model is:

[0078] ,

[0079] where, is the basic wind speed at height without considering turbulence correction; represents the friction velocity; represents the von Kármán constant; represents the surface roughness;

[0080] Introduce the turbulence intensity and the mixed layer height , and dynamically adjust as follows:

[0081] ,

[0082] where, the influence of turbulence on the wind speed is corrected through an exponential function, and the correction amplitude is proportional to the turbulence intensity and inversely proportional to the mixed layer height;

[0083] The cumulative influence of turbulence on the wind speed depends not only on height but also on the shear stress and the turbulence variance , and is expressed in integral form as:

[0084] ,

[0085] Among them, is a similarity function, and the calculation formula expresses the height of the non-linear influence on the stability length ; represents the turbulent shear stress, combined with the friction velocity and the dynamic change of aerodynamic parameters.

[0086] The correction of wind speed data by the parameters based on the convective boundary layer includes integrating the corrected wind speed and the turbulent integral influence to obtain an improved dynamic correction wind speed formula:

[0087] ,

[0088] Among them, is the height and time under the corrected wind speed; is the friction velocity; is the von Kármán constant is the surface roughness; is the wind speed measurement height; is the time; is the turbulent intensity correction coefficient; is the mixing layer height, which is dynamically related to time ; is the shear stress; is the air density; is the turbulent influence adjustment coefficient; is the variance of the turbulent field at height ; is the stability length at time t; is the integration variable, representing the position from the surface height to the target height ;

[0089] The construction of the dynamic wind speed prediction model based on time series includes obtaining the complex dynamic relationship of wind speed changes by processing the corrected wind speed time series data based on the long short-term memory network LSTM or Transformer structure in deep learning, combined with local terrain features and seasonal climate patterns;

[0090] Define the input feature embedding matrix as , where all features are normalized so that their ranges are unified to ;

[0091] Time step and The correlation between them is calculated through the self-attention mechanism

[0092] ,

[0093] where:

[0094] Query matrix:

[0095] Key matrix:

[0096] Value matrix:

[0097] Weight matrix is a learnable parameter, is the dimension of the key vector;

[0098] For LSTM, the hidden state at time step is updated as:

[0099] ,

[0100] where, represents the Sigmoid activation function, is a learnable parameter of the network;

[0101] The final wind speed prediction value is mapped by the output of LSTM or Transformer through a fully connected layer as:

[0102] ,

[0103] where, are the weights and biases of the fully connected layer.

[0104] The construction of the wind energy utilization optimization model includes, based on the wake effect and synergy effect of wind turbines, comprehensively calculating the impact of the airflow interaction between wind turbines on the overall power generation capacity. Among them, the wake effect in the wind energy utilization optimization model is described by the Jensen model or the Gaussian wake model. The Jensen model is based on the linear assumption of wake speed decay, and its wake speed decay formula is:

[0105] ,

[0106] where, represents the wind speed at a certain point in the wake area; is the wind speed without being affected by the wake, is the induction factor of the unit, is the wake expansion coefficient, is the horizontal distance between the unit and the wake influence point. The Gaussian model simulates the spatial attenuation characteristics of the wake through Gaussian distribution, and the expression is

[0107] ,

[0108] where represents the wind speed at a point within the wake region in three-dimensional space; represents the horizontal lateral coordinate, which is used to describe the lateral distribution of the wake; is the position of the wake center axis, is the wake expansion width parameter, which reflects the spatial distribution of the wake; The synergy effect optimizes the layout of wind turbines through a genetic algorithm. The objective function is defined as the total power generation of the group, which is expressed as

[0109] ,

[0110] where represents the total power generation of all wind turbines in the wind farm; represents the total number of wind turbines in the wind farm; is the power generation of the th unit, is the corrected wind speed at the location of this unit, is the wind direction angle of the unit.

[0111] The construction of the robust power prediction model includes constructing multiple random split trees, mapping the samples to the leaf nodes through eigenvalue partitioning. The outlier score formula of the isolation forest is defined as:

[0112] ,

[0113] where is the outlier score of the sample , is the average path length required to be separated in all random trees, is the expected path length of the adjustment factor, representing the standard path length value when the number of sample points in the forest is n; is the total number of samples in the isolation forest; When (threshold), it is determined that the current data point belongs to extreme weather conditions, triggering the prediction mode switch; is the outlier determination threshold of the isolation forest;

[0114] After detecting extreme weather, the robust power prediction model switches from the standard prediction mode to the extreme weather prediction mode, maintains the uninterrupted input of real-time data stream, and reloads the parameterized model under extreme weather conditions according to the current environmental variables;

[0115] The extreme weather prediction mode uses the random forest or Bayesian regression algorithm. For the random forest, the input is the features after anomaly detection , which represent wind speed, wind direction, temperature, and air pressure respectively. The non-linear relationship between wind speed and power is calculated through the splitting of multiple decision trees, and the power prediction value is given by the average output of all trees

[0116] ,

[0117] Among them, represents the power prediction value; is the th tree, is the total number of decision trees; For Bayesian regression, a regression model for power output is defined:

[0118] ,

[0119] Among them, is the weight vector, is the input feature vector is the noise term, and Bayes' formula is used to update the posterior distribution of the weights:

[0120] ,

[0121] Among them, and are the input and output samples respectively, is the prior distribution.

[0122] The dynamic adjustment of the prediction model parameters by combining the online learning mechanism includes that the model dynamically updates the recent real-time data through a sliding window, and the window length is determined by the characteristics of the current wind speed fluctuation. When the input data arrives, the splitting parameters of the random forest or the prior weights of the Bayesian regression are adjusted through the online learning mechanism, and the formula is

[0123] ,

[0124] Among them, is the learning rate, is the updated i-th weight value, is the i-th weight value before update, is the gradient update amount for the current window, ensuring that the model has dynamic adaptability under extreme weather conditions.

[0125] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

[0126] Embodiment 2, the second embodiment of the present invention, which is different from the previous embodiment in that:

[0127] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the essence of the technical solution of the present invention, or the part that contributes to the prior art, or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0128] This application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0129] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including instruction means, and the instruction means implements the functions in Figure 1 one process or multiple processes and / or blocks Figure 1The functions specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide for implementing the steps of the functions specified in one Figure 1 one process or more processes and / or boxes Figure 1 step of the functions specified in one box or more boxes.

[0131] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present application.

[0132] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A method for calculating wind speed, power and electricity quantity in a wind farm, characterized in that: including Obtain high-resolution meteorological data of the wind farm area, clean the data, handle outliers, and perform spatio-temporal alignment. Then, correct the wind speed data based on the parameters of the convective boundary layer to generate corrected wind speed data; Construct a dynamic wind speed prediction model based on time series to predict the corrected wind speed data; Construct an optimization model for wind energy utilization based on the wake effect and layout synergy effect between wind turbines; For extreme weather conditions, construct a robust power prediction model based on historical weather data and an anomaly detection mechanism, and dynamically adjust the prediction model parameters in combination with an online learning mechanism to complete the calculation of the wind farm power; The construction of the wind energy utilization optimization model includes that the wake effect in the wind energy utilization optimization model is described by the Jensen model or the Gaussian wake model. The Jensen model is based on the linear assumption of wake speed decay, and its wake speed decay formula is: where v0 is the wind speed not affected by the wake, a is the induction factor of the unit, k is the wake expansion coefficient, x is the horizontal distance between the unit and the wake influence point, and the Gaussian model simulates the spatial decay characteristics of the wake through a Gaussian distribution, and the expression is: where y0 is the position of the wake center axis, and σ is the wake expansion width parameter, which reflects the spatial distribution of the wake; The synergy effect optimizes the layout of wind turbines through a genetic algorithm. The objective function is defined as the total power generation of the group, which is expressed as: Among them, P i is the power generation power of the i-th unit, v i is the corrected wind speed at the location of the unit, and θ i is the wind direction angle of the unit; The optimization process of the genetic algorithm includes initializing the layout parameters of the wind turbines, that is, the position and wind direction angle, calculating the wind speed correction value of each wind turbine based on the wake model, calculating the total power generation as the fitness function, and iteratively optimizing the layout parameters through selection, crossover, and mutation operations; The construction of the robust power prediction model includes constructing multiple random split trees, mapping the samples to leaf nodes through eigenvalue partitioning, and the anomaly score formula of the isolation forest is defined as: where s(x) is the anomaly score of sample x, E(h(x)) is the average path length required for x to be separated in all random trees, c(n) is the expected path length of the adjustment factor, and n is the number of samples; when s(x)>τ (threshold), it is determined that the current data point x belongs to extreme weather conditions, and the prediction mode is triggered to switch; After detecting extreme weather, the model switches from the standard prediction mode to the extreme weather prediction mode, keeps the real-time data stream input uninterrupted, and reloads the parameterized model under extreme weather conditions according to the current environmental variables; The extreme weather prediction model uses a random forest or Bayesian regression algorithm. For the random forest, the input is the features {v, d, T, P env} after anomaly detection, representing wind speed, wind direction, temperature, and air pressure respectively. The non-linear relationship between wind speed and power is calculated through the splitting of multiple decision trees, and the power prediction value is given by the average output of all trees: Among them, T i is the i-th tree, and N is the total number of decision trees; For Bayesian regression, define the regression model of power output: P(t) = w T x + ∈ where \(w\) is the weight vector, \(x\) is the input feature vector \([v, d, T, P env , is the noise term, and the Bayes formula is used to update the posterior distribution of the weights: p(w∣X,y)∝p(y∣X,w)p(w) where X and y are the input and output samples respectively, and p(w) is the prior distribution.

2. The method for calculating wind farm wind speed, power and electricity quantity according to claim 1, characterized in that: The high-resolution meteorological data includes wind speed, wind direction, air pressure, and temperature data. The data sources include data collected by satellite remote sensing, ground meteorological station observations, and radar systems, and the data is spatio-temporally aligned through an interpolation algorithm.

3. The method for calculating wind farm wind speed, power and electricity quantity according to claim 2, characterized in that: The correction of the wind speed data based on the parameters of the convective boundary layer includes that the expression of the logarithmic wind speed profile model is: where U base (z) is the basic wind speed at height z without including the turbulence correction; Introduce the turbulence intensity η(z,t) and the mixing layer height H(t) to dynamically adjust U base (z): Among them, the influence of turbulence on wind speed is corrected by an exponential function, and the correction amplitude is proportional to the turbulence intensity and inversely proportional to the mixing layer height; The cumulative effect of turbulence on wind speed depends not only on height, but also on the shear stress τ(z,t) and the turbulence variance and is expressed in integral form as: Among them, is a similarity function, and Φ(x) = 1 - 2x + x 2 is used to express the non-linear influence of the height ξ on the stability length L(t); τ(ξ, t) represents the turbulent shear stress, combined with the friction velocity u * and the dynamic changes of aerodynamic parameters.

4. The method for calculating wind farm wind speed, power and electricity quantity according to claim 3, wherein: The correction of wind speed data based on the parameters of the convective boundary layer includes correcting the wind speed U adj (z, t) and the influence of turbulent integral on U int (z, t) are integrated to obtain an improved dynamic correction wind speed formula: where U(z,t) is the corrected wind speed at height z and time t; u * is the friction velocity; κ is the von Kármán constant with a value of 0.4; z0 is the surface roughness; z is the wind speed measurement height; t is the time; α is the turbulence intensity correction coefficient; η(z,t) is the turbulence intensity field; H(t) is the mixing layer height, dynamically related to time t; τ(ξ,t) is the shear stress; ρ is the air density; β is the turbulence influence adjustment coefficient; is the height variance of the turbulence field; L(t) is the stability length Φ(x) = 1 - 2x + x 2 is the turbulence similarity function; ξ is the integration variable, representing the position from the surface height z0 to the target height z.

5. The method for calculating wind farm wind speed, power and electricity quantity according to claim 4, characterized in that: The construction of the dynamic wind speed prediction model based on time series includes, based on the long short-term memory network LSTM or Transformer structure in deep learning, by processing the corrected wind speed time series data, combining local terrain features and seasonal climate patterns, to obtain the complex dynamic relationship of wind speed changes; The input feature embedding matrix X(t) is defined as: Among them, all features are normalized so that their ranges are unified to [0, 1]; The correlation between time steps t and t+Δt is calculated by the self-attention mechanism: Where: Query matrix: Q = X(t)W Q Key matrix: K = X(t)W K Value matrix: V = X(t)W V Weight matrix W Q ,W K ,W V is a learnable parameter, and d k is the dimension of the key vector; For LSTM, the hidden state at time step t is updated as: h t = σ(W f [h t-1 , X(t)] + b f ) · h t-1 + tanh(W c [h t-1 , X(t)] + b c ) Among them, σ represents the Sigmoid activation function, W f , W c , b f , b c are the learnable parameters of the network; Final predicted wind speed The output of LSTM or Transformer is mapped through a fully connected layer to: Among them, W o , b o are the weights and biases of the fully connected layer.

6. The method for calculating wind farm wind speed, power and electricity quantity according to claim 5, wherein: The combination of the online learning mechanism to dynamically adjust the prediction model parameters includes that the model dynamically updates the most recent real-time data through a sliding window, and the window length L is determined by the characteristics of the current wind speed fluctuation. When the input data arrives, the splitting parameters of the random forest or the prior weights of the Bayesian regression are adjusted through the online learning mechanism, and the formula is: Among them, η is the learning rate, and Δw is the gradient update amount of the current window, ensuring that the model has dynamic adaptability under extreme weather conditions.

7. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method described in any one of claims 1 to 5.

Images