Wind power plant generation power prediction method and system, equipment and medium

Abstract

The invention relates to the technical field of wind power plant data processing, in particular to a wind power plant generation power prediction method and system, equipment and a medium, and the method and system mainly comprise the steps of building a fan power prediction model, carrying out regional division on fans of a wind power plant, and carrying out power prediction on the fans of the wind power plant. And selecting the fan with the highest evaluation value in the single region as a target fan in the region, calculating the predicted power of the to-be-predicted fan through the spatial data, and after judgment, retaining or deleting the predicted power of the to-be-predicted fan calculated through the spatial data. Through the method, the power prediction of each fan by using the neural network of the power prediction model is reduced to a certain extent, the to-be-predicted fan meeting the historical operation database rule after judgment is not predicted by using the neural network of the power prediction model any more, the workload of an operation and maintenance system is reduced as much as possible, and the power prediction efficiency is improved. And more loads are provided for the operation and maintenance system to process other more important items.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm data processing, and in particular to a method and system, equipment and medium for predicting power generation of a wind farm. Background Art

[0002] Wind power generation is an important form of renewable energy, and wind turbine power prediction is the key to optimizing the operation of wind power generation equipment and grid dispatching. Traditional wind turbine power prediction methods include prediction based on physical models and prediction based on statistical learning. Prediction based on physical models usually considers the relationship between wind turbines and meteorological conditions, while methods based on statistical learning establish prediction models by analyzing historical data, including time series analysis, artificial neural networks, support vector machines, etc. In recent years, with the development of deep learning technology, methods based on deep learning have also been widely used in the field of wind turbine power prediction, such as using models such as recurrent neural networks (RNN) and long short-term memory networks (LSTM) to capture the complex relationship between time series data.

[0003] However, the current wind farms have a large installed capacity and the operation and maintenance management burden is heavy. Using existing conventional technologies to predict the power of each wind turbine will bring a greater burden to the operation and maintenance system, which may affect the implementation of other work. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system, device and medium for predicting the power generation of a wind farm to solve the above-mentioned technical problems in the prior art.

[0005] The embodiments of the present invention are implemented by the following technical solutions:

[0006] In a first aspect, the present invention provides a method for predicting power generation of a wind farm, comprising:

[0007] Acquire historical operation data of the wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output the optimal wind turbine power prediction model;

[0008] The wind turbines in the wind farm are divided into regions to obtain wind turbine groups in several regions, and a wind turbine selection evaluation model is established. The wind turbines in several regions are evaluated separately by the wind turbine evaluation model to obtain evaluation values, and the wind turbine with the highest evaluation value in a single region is selected as the target wind turbine in the region;

[0009] Obtain the future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through the wind turbine power prediction model;

[0010] Acquire the spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted through the spatial data;

[0011] The predicted power of the wind turbine to be predicted is compared with the historical operating data. If a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operating data is greater than the threshold, the predicted power of the current wind turbine to be predicted is deleted, and the output predicted power of the wind turbine to be predicted is performed through the wind turbine power prediction model. If the difference between the predicted power of the wind turbine to be predicted and the historical operating data is not greater than the threshold, the predicted power of the wind turbine to be predicted calculated through the spatial data is retained.

[0012] Preferably, the step of establishing a wind turbine power prediction model includes:

[0013] The abandoned wind data is removed from the historical operation data of the wind turbine, and the data is corrected by the LOF algorithm to obtain the processed historical operation data of the wind turbine;

[0014] The processed historical operation data of the wind turbine is normalized to obtain a data set, and divided into a training set, a validation set, and a test set;

[0015] The features of the training set are extracted to obtain the feature input set, and the wind power time series in the data set is denoised by wavelet threshold;

[0016] The input order of the historical wind power data and meteorological data after noise reduction is determined, the input vector of the prediction model is constructed by the input set of meteorological factor characteristics and the historical wind power order, and the wind turbine power prediction model is constructed through the input vector.

[0017] Preferably, the wavelet threshold denoising includes:

[0018] Perform wavelet transform on the original signal, determine the wavelet coefficients and perform threshold processing:

[0019] w λ =sign(w)(|w|-λ)

[0020] In the formula, w λ is the wavelet coefficient after threshold processing, w is the initial wavelet coefficient, and λ is the threshold;

[0021] Get the wavelet coefficients after threshold processing for wavelet reconstruction:

[0022]

[0023] In the formula, SNR is the signal denoising evaluation index, s 2 (i) is the wavelet coefficient after threshold processing, s(i) is the pure signal, is the denoised signal, and N is the number of denoised signal samples.

[0024] Preferably, constructing a wind turbine power prediction model by inputting a vector comprises:

[0025]

[0026] Where y(t) is the predicted power output value of the model, f() is a nonlinear function, x1(ti) is the average wind speed, y(tj) is the historical wind speed, ε(t) is the noise, i is the model input order of the average wind speed, and j is the model input order of the historical wind speed.

[0027] Preferably, the wind turbines of the wind farm are divided into regions to obtain wind turbine groups in several regions, including:

[0028] Get the area value and number of wind turbines of the current wind farm. The area value is a value in square kilometers. Get the number of divided areas:

[0029]

[0030] In the formula, Q is the number of regions, s is x is the area of ​​the current wind farm, and R is the number of wind turbines;

[0031] The two-dimensional plane diagram of the wind farm is divided into grids, and the number of grids is equal to the number of divided areas.

[0032] Preferably, the establishment of a wind turbine selection evaluation model includes:

[0033]

[0034] In the formula, a i is the distance between the ith wind turbine and the center of the region, b i is the historical failure rate of the i-th wind turbine, N is the total number of wind turbines in the area, P i is the installed power of the i-th wind turbine, I i is the altitude value of the i-th wind turbine.

[0035] Preferably, calculating the predicted power of the wind turbine to be predicted by using spatial data includes:

[0036]

[0037] Where P x The predicted power of the fan to be predicted, ρ is the air density, C p Wind power coefficient, v is the predicted wind speed received by the target wind turbine, θ is the angle between the wind turbine to be predicted and the target wind turbine and the X-axis in the plane coordinate system, ω is the angle between the wind turbine to be predicted and the target wind turbine and the Y-axis in the plane coordinate system, and d is the distance between the wind turbine to be predicted and the target wind turbine.

[0038] In a second aspect, the present invention further provides a wind farm power generation prediction system, comprising:

[0039] A prediction model building module is configured to obtain historical operation data of the wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output an optimal wind turbine power prediction model;

[0040] The regional division module is configured to divide the wind turbines in the wind farm into regions, obtain wind turbine groups in several regions, establish a wind turbine selection evaluation model, evaluate the wind turbines in several regions separately using the wind turbine evaluation model, obtain evaluation values, and select the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region;

[0041] The prediction module is configured to obtain the future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through the wind turbine power prediction model; obtain the spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted through the spatial data; compare the obtained predicted power of the wind turbine to be predicted with the historical operation data, and if a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is greater than the value threshold, then delete the predicted power of the current wind turbine to be predicted, and output the predicted power of the wind turbine to be predicted through the wind turbine power prediction model, and if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is not greater than the value threshold, then retain the predicted power of the wind turbine to be predicted calculated through the spatial data;

[0042] The main control module is connected with the prediction model establishment module, the area division module and the prediction module, and is used to execute the above-mentioned wind farm power generation prediction method.

[0043] In a third aspect, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for predicting power generation of a wind farm when executing the computer program.

[0044] In a fourth aspect, the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, a method for predicting the power generation of a wind farm as described above is implemented.

[0045] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0046] The method and system provided by the present invention mainly include establishing a wind turbine power prediction model, dividing the wind turbines in the wind farm into regions, selecting the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region, calculating the predicted power of the wind turbine to be predicted through spatial data, and retaining or deleting the predicted power of the wind turbine to be predicted calculated through spatial data after judgment. Through the above method, the power prediction of each wind turbine using the neural network of the power prediction model is reduced to a certain extent. After the wind turbine to be predicted meets the rules of the historical operation database after judgment, the neural network of the power prediction model is no longer used for prediction, which reduces the workload of the operation and maintenance system as much as possible, and provides more load for the operation and maintenance system to handle other more important matters. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0050] The division of modules in this application is a logical division. There may be other division methods when implemented in actual applications. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0051] The modules or submodules described independently may be physically separated or not: they may be implemented by software or hardware, and some modules or submodules may be implemented by software, and the processor may call the software to implement the functions of these modules or submodules, and other modules or submodules may be implemented by hardware, such as by hardware circuits. In addition, some or all of the modules may be selected according to actual needs to achieve the purpose of the present application.

[0052] Please refer to Figure 1The present invention provides a method for predicting wind farm power generation, comprising:

[0053] S101: Acquire historical operation data of a wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output an optimal wind turbine power prediction model;

[0054] In this embodiment, the wind turbine power prediction model can be trained using conventional technical means of the prior art, such as training with a training set, verifying its performance with a validation set, and testing the prediction results with a test set.

[0055] S102: Divide the wind turbines in the wind farm into regions to obtain wind turbine groups in several regions, establish a wind turbine selection evaluation model, evaluate the wind turbines in several regions using the wind turbine evaluation model to obtain evaluation values, and select the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region;

[0056] In this embodiment, since the scale of the wind farm is relatively large, if the entire wind farm is considered together, it will lead to large data errors. Therefore, this embodiment performs regional division and selects representative wind turbines to calculate the data of adjacent wind turbines.

[0057] S103: Acquire future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through a wind turbine power prediction model;

[0058] S104: Acquire spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted based on the spatial data;

[0059] S105: Compare the predicted power of the wind turbine to be predicted with the historical operating data. If a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operating data is greater than the threshold, delete the predicted power of the current wind turbine to be predicted, and output the predicted power of the wind turbine to be predicted through the wind turbine power prediction model. If the difference between the predicted power of the wind turbine to be predicted and the historical operating data is not greater than the threshold, retain the predicted power of the wind turbine to be predicted calculated through the spatial data.

[0060] In this embodiment, the corresponding difference threshold can be set according to the experience in the field, and it can be set within a reasonable range. If the predicted power of a wind turbine to be predicted is greatly different from the historical operating data, it will not be included in the current calculation, and the wind turbine power prediction model will still be used for prediction. The purpose is to provide another possible prediction method for the predicted power of the wind turbine to be predicted, so as to reduce the burden on the operation and maintenance system as much as possible.

[0061] The method and system provided by the present invention mainly include establishing a wind turbine power prediction model, dividing the wind turbines in the wind farm into regions, selecting the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region, calculating the predicted power of the wind turbine to be predicted through spatial data, and retaining or deleting the predicted power of the wind turbine to be predicted calculated through spatial data after judgment. Through the above method, the power prediction of each wind turbine using the neural network of the power prediction model is reduced to a certain extent. After the wind turbine to be predicted meets the rules of the historical operation database after judgment, the neural network of the power prediction model is no longer used for prediction, which reduces the workload of the operation and maintenance system as much as possible, and provides more load for the operation and maintenance system to handle other more important matters.

[0062] In an exemplary embodiment of the present invention, establishing a wind turbine power prediction model includes:

[0063] The abandoned wind data is removed from the historical operation data of the wind turbine, and the data is corrected by the LOF algorithm to obtain the processed historical operation data of the wind turbine;

[0064] The LOF algorithm is a density-based anomaly detection algorithm used to identify anomalies in different density areas in a data set. The LOF algorithm evaluates the degree of anomaly of each data point by calculating the local density of each data point relative to its neighboring data points. That is, the LOF value of a data point represents the relative degree of anomaly of the point relative to the density of its neighboring data points.

[0065] The processed historical operation data of the wind turbine is normalized to obtain a data set, and divided into a training set, a validation set, and a test set;

[0066] The features of the training set are extracted to obtain the feature input set, and the wind power time series in the data set is denoised by wavelet threshold;

[0067] The input order of the historical wind power data and meteorological data after noise reduction is determined, the input vector of the prediction model is constructed by the input set of meteorological factor characteristics and the historical wind power order, and the wind turbine power prediction model is constructed through the input vector.

[0068] In an exemplary embodiment of the present invention, wavelet threshold denoising includes:

[0069] Perform wavelet transform on the original signal, determine the wavelet coefficients and perform threshold processing:

[0070] w λ =sign(w)(|w|-λ)

[0071] In the formula, w λ is the wavelet coefficient after threshold processing, w is the initial wavelet coefficient, and λ is the threshold;

[0072] Generally, there are two main threshold processing methods: soft threshold function processing and hard threshold function processing.

[0073] Get the wavelet coefficients after threshold processing for wavelet reconstruction:

[0074]

[0075] In the formula, SNR is the signal denoising evaluation index, s 2 (i) is the wavelet coefficient after threshold processing, s(i) is the pure signal, is the denoised signal, and N is the number of denoised signal samples.

[0076] In an exemplary embodiment of the present invention, constructing a wind turbine power prediction model by inputting a vector comprises:

[0077]

[0078] Where y(t) is the predicted power output value of the model, f() is a nonlinear function, x1(ti) is the average wind speed, y(tj) is the historical wind speed, ε(t) is the noise, i is the model input order of the average wind speed, and j is the model input order of the historical wind speed.

[0079] In an exemplary embodiment of the present invention, the wind turbines of the wind farm are divided into regions to obtain wind turbine groups in several regions, including:

[0080] Get the area value and number of wind turbines of the current wind farm. The area value is a value in square kilometers. Get the number of divided areas:

[0081]

[0082] In the formula, Q is the number of regions, s is x is the area of ​​the current wind farm, and R is the number of wind turbines;

[0083] The two-dimensional plane diagram of the wind farm is divided into grids, and the number of grids is equal to the number of divided areas.

[0084] In an exemplary embodiment of the present invention, the establishment of a wind turbine selection evaluation model includes:

[0085]

[0086] In the formula, a i is the distance between the ith wind turbine and the center of the region, b i is the historical failure rate of the i-th wind turbine, N is the total number of wind turbines in the area, P i is the installed power of the i-th wind turbine, I iis the altitude value of the i-th wind turbine.

[0087] In an exemplary embodiment of the present invention, calculating the predicted power of the wind turbine to be predicted by using spatial data includes:

[0088]

[0089] Where P x The predicted power of the fan to be predicted, ρ is the air density, C p Wind power coefficient, v is the predicted wind speed received by the target wind turbine, θ is the angle between the wind turbine to be predicted and the target wind turbine and the X-axis in the plane coordinate system, ω is the angle between the wind turbine to be predicted and the target wind turbine and the Y-axis in the plane coordinate system, and d is the distance between the wind turbine to be predicted and the target wind turbine.

[0090] In a second aspect, the present invention further provides a wind farm power generation prediction system, comprising:

[0091] A prediction model building module is configured to obtain historical operation data of the wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output an optimal wind turbine power prediction model;

[0092] The regional division module is configured to divide the wind turbines in the wind farm into regions, obtain wind turbine groups in several regions, establish a wind turbine selection evaluation model, evaluate the wind turbines in several regions separately using the wind turbine evaluation model, obtain evaluation values, and select the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region;

[0093] The prediction module is configured to obtain the future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through the wind turbine power prediction model; obtain the spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted through the spatial data; compare the obtained predicted power of the wind turbine to be predicted with the historical operation data, and if a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is greater than the value threshold, then delete the predicted power of the current wind turbine to be predicted, and output the predicted power of the wind turbine to be predicted through the wind turbine power prediction model, and if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is not greater than the value threshold, then retain the predicted power of the wind turbine to be predicted calculated through the spatial data;

[0094] The main control module is connected with the prediction model establishment module, the area division module and the prediction module, and is used to execute the above-mentioned wind farm power generation prediction method.

[0095] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0096] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0097] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting wind farm power generation, characterized in that: include: Acquire historical operation data of the wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output the optimal wind turbine power prediction model; The wind turbines in the wind farm are divided into regions to obtain wind turbine groups in several regions, and a wind turbine selection evaluation model is established. The wind turbines in several regions are evaluated separately by the wind turbine evaluation model to obtain evaluation values, and the wind turbine with the highest evaluation value in a single region is selected as the target wind turbine in the region; Obtain the future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through the wind turbine power prediction model; Acquire the spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted through the spatial data; The predicted power of the wind turbine to be predicted is compared with the historical operating data. If a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operating data is greater than the threshold, the predicted power of the current wind turbine to be predicted is deleted, and the output predicted power of the wind turbine to be predicted is performed through the wind turbine power prediction model. If the difference between the predicted power of the wind turbine to be predicted and the historical operating data is not greater than the threshold, the predicted power of the wind turbine to be predicted calculated through the spatial data is retained.

2. A method for predicting wind farm power generation according to claim 1, characterized in that: The wind turbine power prediction model is established as follows: The abandoned wind data is removed from the historical operation data of the wind turbine, and the data is corrected by the LOF algorithm to obtain the processed historical operation data of the wind turbine; The processed historical operation data of the wind turbine is normalized to obtain a data set, and divided into a training set, a validation set, and a test set; The features of the training set are extracted to obtain the feature input set, and the wind power time series in the data set is denoised by wavelet threshold; The input order of the historical wind power data and meteorological data after noise reduction is determined, the input vector of the prediction model is constructed by the input set of meteorological factor characteristics and the historical wind power order, and the wind turbine power prediction model is constructed through the input vector.

3. A method for predicting wind farm power generation according to claim 2, characterized in that: The wavelet threshold denoising comprises: Perform wavelet transform on the original signal, determine the wavelet coefficients and perform threshold processing: In λ =sign(w)(|w|-λ) In the formula, w λ is the wavelet coefficient after threshold processing, w is the initial wavelet coefficient, and λ is the threshold; Get the wavelet coefficients after threshold processing for wavelet reconstruction: In the formula, SNR is the signal denoising evaluation index, s 2 (i) is the wavelet coefficient after threshold processing, s(i) is the pure signal, is the denoised signal, and N is the number of denoised signal samples.

4. A method for predicting wind farm power generation according to claim 2, characterized in that: The constructing of a wind turbine power prediction model by inputting a vector comprises: Where y(t) is the predicted power output value of the model, f() is a nonlinear function, x1(ti) is the average wind speed, y(tj) is the historical wind speed, ε(t) is the noise, i is the model input order of the average wind speed, and j is the model input order of the historical wind speed.

5. A method for predicting wind farm power generation according to claim 2, characterized in that: The wind turbines of the wind farm are divided into several regions to obtain wind turbine groups in several regions, including: Get the area value and number of wind turbines of the current wind farm. The area value is a value in square kilometers. Get the number of divided areas: In the formula, Q is the number of regions, s is x is the area of ​​the current wind farm, and R is the number of wind turbines; The two-dimensional plane diagram of the wind farm is divided into grids, and the number of grids is equal to the number of divided areas.

6. A method for predicting wind farm power generation according to claim 3, characterized in that: The establishment of the wind turbine selection evaluation model comprises: In the formula, a i is the distance between the ith wind turbine and the center of the region, b i is the historical failure rate of the i-th wind turbine, N is the total number of wind turbines in the area, P i is the installed power of the i-th wind turbine, I i is the altitude value of the i-th wind turbine.

7. A method for predicting wind farm power generation according to claim 3, characterized in that: The predicted power of the wind turbine to be predicted is calculated through spatial data including: Where P x The predicted power of the fan to be predicted, ρ is the air density, C p Wind power coefficient, v is the predicted wind speed received by the target wind turbine, θ is the angle between the wind turbine to be predicted and the target wind turbine and the X-axis in the plane coordinate system, ω is the angle between the wind turbine to be predicted and the target wind turbine and the Y-axis in the plane coordinate system, and d is the distance between the wind turbine to be predicted and the target wind turbine.

8. A wind farm power generation prediction system, characterized in that: include: A prediction model building module is configured to obtain historical operation data of the wind turbine, the operation data including wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model through the historical operation data, and output an optimal wind turbine power prediction model; The regional division module is configured to divide the wind turbines in the wind farm into regions, obtain wind turbine groups in several regions, establish a wind turbine selection evaluation model, evaluate the wind turbines in several regions separately using the wind turbine evaluation model, obtain evaluation values, and select the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region; The prediction module is configured to obtain the future predicted wind speed data output by the weather forecast system, and output the predicted power of the target wind turbine through the wind turbine power prediction model; obtain the spatial data between the wind turbine to be predicted and the target wind turbine in the same area, and calculate the predicted power of the wind turbine to be predicted through the spatial data; compare the obtained predicted power of the wind turbine to be predicted with the historical operation data, and if a difference threshold is set, if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is greater than the value threshold, then delete the predicted power of the current wind turbine to be predicted, and output the predicted power of the wind turbine to be predicted through the wind turbine power prediction model, and if the difference between the predicted power of the wind turbine to be predicted and the historical operation data is not greater than the value threshold, then retain the predicted power of the wind turbine to be predicted calculated through the spatial data; A main control module is connected to the prediction model building module, the area division module and the prediction module, and is used to execute a wind farm power generation prediction method as described in any one of claims 1-7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, a method for predicting power generation in a wind farm according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a method for predicting the power generation of a wind farm according to any one of claims 1 to 7 is implemented.

Images