A wind farm power generation power prediction method, system, device and medium

Abstract

The present application relates to the technical field of wind farm data processing, in particular to a wind farm power generation power prediction method and system, equipment and medium, the method and system provided by the present application mainly include establishing a wind turbine power prediction model, dividing the wind turbines of 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 then judging whether to retain or delete the predicted power of the wind turbine to be predicted calculated through spatial data. 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, and the wind turbine to be predicted that meets the historical operation database rule is no longer predicted using the neural network of the power prediction model, thereby minimizing the work burden of the operation and maintenance system and providing more load for the operation and maintenance system to handle more important matters.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind farm data processing, in particular to a wind farm power generation prediction method and system, equipment and medium. BACKGROUND

[0002] Wind power generation is an important form of renewable energy, and wind turbine power prediction is the key to optimizing wind power generation equipment operation and grid scheduling. Traditional wind turbine power prediction methods include physical model-based prediction and statistical learning-based prediction. Physical model-based prediction generally considers the relationship between wind turbines and meteorological conditions, and statistical learning-based methods 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, deep learning-based methods have been widely used in wind turbine power prediction, such as using recurrent neural networks (RNN) and long short-term memory networks (LSTM) to capture the complex relationships between time series data.

[0003] However, the current wind farm has a large number of installed units, and the operation and maintenance management burden is large. Using existing conventional technology to predict the power of each wind turbine will bring greater burden to the operation and maintenance system, which may affect the development of other work. SUMMARY

[0004] The purpose of the present application is to provide a wind farm power generation prediction method and system, equipment and medium to solve the above technical problems in the prior art.

[0005] Embodiments of the present application are implemented by the following technical solutions:

[0006] In a first aspect, the present application provides a wind farm power generation prediction method, comprising:

[0007] Obtaining historical operation data of wind turbines, the operation data including wind speed and power generation, establishing a wind turbine power prediction model, training the wind turbine power prediction model with historical operation data, and outputting the optimal wind turbine power prediction model;

[0008] Dividing the wind turbines of the wind farm into regions to obtain wind turbine groups in several regions, establishing a wind turbine selection evaluation model, evaluating the wind turbines in the several regions separately through the wind turbine evaluation model to obtain evaluation values, and selecting the wind turbine with the highest evaluation value in a single region as the target wind turbine in the region;

[0009] Obtaining future predicted wind speed data output by a weather forecasting system, and outputting the predicted power of the target wind turbine through the wind turbine power prediction model;

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

[0011] The predicted power of the wind turbine to be predicted is compared with the historical operating data, and 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 difference threshold, the predicted power of the current wind turbine to be predicted is deleted, and the predicted power of the wind turbine to be predicted is output 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 difference threshold, the predicted power of the wind turbine to be predicted is retained and calculated through spatial data.

[0012] Preferably, the establishment of the wind turbine power prediction model includes:

[0013] The historical operating data of the wind turbines is processed by removing the data of wind curtailment and correcting it using the LOF algorithm to obtain the processed historical operating data of the wind turbines.

[0014] The processed historical wind turbine operation data is normalized to obtain a dataset, which is then divided into a training set, a validation set, and a test set.

[0015] Features are extracted from the training set to obtain the feature input set, and wavelet thresholding is used to denoise the wind power time series in the dataset.

[0016] Determine the input order of historical wind power data and meteorological data after noise reduction. Construct the prediction model input vector from the meteorological factor feature input set and the historical wind power order. Construct the wind turbine power prediction model through the input vector.

[0017] Preferably, the wavelet threshold denoising includes:

[0018] Perform wavelet transform on the original signal, determine the wavelet coefficients, and then perform thresholding.

[0019] w λ =sign(w)(|w|-λ)

[0020] In the formula, w λ The wavelet coefficients are processed by the threshold, w is the initial wavelet coefficients, and λ is the threshold.

[0021] Wavelet reconstruction is performed by obtaining the wavelet coefficients after thresholding.

[0022]

[0023] In the formula, SNR is the evaluation index for signal denoising, and s(i) is the clean signal. denoised signal, N is the number of denoised signal samples.

[0024] Preferably, the constructing a wind turbine power prediction model through an input vector comprises:

[0025]

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

[0027] Preferably, the regionally dividing wind turbines of a wind farm to obtain wind turbine groups in a plurality of regions comprises:

[0028] obtaining an area value and a wind turbine number value of a current wind farm, the area value being a value in square kilometers, and obtaining a number of divided regions:

[0029]

[0030] wherein Q is the number of regions, s x is the area value of the current wind farm, and R is the wind turbine number value;

[0031] grid dividing a two-dimensional plan view of the wind farm, the number of grids being equal to the number of divided regions.

[0032] Preferably, the establishing a wind turbine selection evaluation model comprises:

[0033]

[0034] wherein a i is a distance of an i-th wind turbine from a region center, b i is a historical failure rate of the i-th wind turbine, N is a total number of wind turbines in the region, P i is an installed power of the i-th wind turbine, I i is an altitude value of the i-th wind turbine.

[0035] Preferably, the calculating a predicted power of a to-be-predicted wind turbine through spatial data comprises:

[0036]

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

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

[0039] The prediction model establishing module is configured to acquire historical operation data of the wind turbine, the operation data comprising wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model by the historical operation data, and output an optimal wind turbine power prediction model.

[0040] The region dividing module is configured to divide the wind turbines of the wind farm into regions to obtain wind turbine groups in a plurality of regions, establish a wind turbine selection evaluation model, respectively evaluate the wind turbines in the plurality of regions by the wind turbine evaluation model to obtain evaluation values, and select a wind turbine with the highest evaluation value in a single region as a target wind turbine in the region.

[0041] The prediction module is configured to acquire future prediction wind speed data output by a weather prediction system, output prediction power of the target wind turbine by the wind turbine power prediction model, acquire spatial data between a to-be-predicted wind turbine and the target wind turbine in the same region, calculate prediction power of the to-be-predicted wind turbine by the spatial data, compare the prediction power of the to-be-predicted wind turbine with historical operation data, set a difference threshold, if the difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is greater than the difference threshold, delete the prediction power of the to-be-predicted wind turbine, output prediction power of the to-be-predicted wind turbine by the wind turbine power prediction model, and if the difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is not greater than the difference threshold, retain the prediction power of the to-be-predicted wind turbine calculated by the spatial data.

[0042] The main control module is connected with the prediction model establishing module, the region dividing module and the prediction module, and is configured to execute the wind farm power generation prediction method.

[0043] In a third aspect, the present application 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 executes the computer program to implement the wind farm power generation prediction method.

[0044] In a fourth aspect, the present application further provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the wind farm power generation prediction method.

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

[0046] The method and system provided by the application mainly include establishing a fan power prediction model, dividing the fan of a wind farm into regions, selecting the fan with the highest evaluation value in a single region as the target fan in the region, calculating the prediction power of the to-be-predicted fan through spatial data, and judging whether to retain or delete the prediction power of the to-be-predicted fan calculated through spatial data. Through the above method, the power prediction of each fan by the neural network of the power prediction model is reduced to a certain extent, and the to-be-predicted fan that meets the historical operation database rule is no longer predicted by the neural network of the power prediction model, so as to reduce the work burden of the operation and maintenance system as much as possible and provide more load for the operation and maintenance system to process other more important matters. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor.

[0048] Figure 1 The flowchart of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0050] The division of the modules appearing in the present application is a logical division, and in actual application, another division mode can be used, for example, multiple modules can be combined or integrated in another system, or some features can be ignored or not executed.

[0051] The independently described modules or sub-modules can be physically separated or not physically separated, can be software implemented or hardware implemented, and part of the modules or sub-modules can be implemented by software, the function of the part of the modules or sub-modules is called by a processor, and the other part of the modules or sub-modules is implemented by hardware, for example, by hardware circuit. In addition, part or all of the modules can be selected to achieve the purpose of the present application according to actual needs.

[0052] Please refer to Figure 1The application provides a wind farm power generation power prediction method, comprising the following steps:

[0053] S101: historical operation data of a wind turbine is acquired, the operation data comprising wind speed and power generation power, a wind turbine power prediction model is established, the wind turbine power prediction model is trained through historical operation data, and an optimal wind turbine power prediction model is output;

[0054] In the embodiment, the wind turbine power prediction model can be trained by using conventional technical means in the prior art, the training set is used for training, the verification set is used for verifying the performance, and the test set is used for testing the prediction result.

[0055] S102: the wind turbines of a wind farm are divided into regions, wind turbine groups in a plurality of regions are obtained, a wind turbine selection evaluation model is established, the wind turbines in the plurality of regions are respectively evaluated through the wind turbine selection evaluation model, an evaluation value is obtained, and the wind turbine with the highest evaluation value in a single region is selected as a target wind turbine in the region;

[0056] In the embodiment, since the scale of the wind farm is large, if the entire wind farm is considered together, the data error is large, therefore, the embodiment is divided into regions, and representative wind turbines are selected to calculate the data of adjacent wind turbines.

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

[0058] S104: spatial data between a to-be-predicted wind turbine and the target wind turbine in the same region is acquired, and the prediction power of the to-be-predicted wind turbine is calculated through the spatial data;

[0059] S105: the prediction power of the to-be-predicted wind turbine is compared with historical operation data, a difference threshold is set, if the difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is greater than the difference threshold, the prediction power of the to-be-predicted wind turbine is deleted, the to-be-predicted wind turbine is output through the wind turbine power prediction model, and if the difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is not greater than the difference threshold, the prediction power of the to-be-predicted wind turbine calculated through the spatial data is retained.

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

[0061] The method and system provided by this 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 that region, calculating the predicted power of the wind turbine to be predicted using spatial data, and then retaining or deleting it after judgment. This method reduces the amount of power prediction for each wind turbine using the neural network of the power prediction model. Wind turbines that conform to the historical operation database patterns after judgment are no longer predicted using the neural network of the power prediction model, minimizing the workload of the operation and maintenance system and allowing it to handle other more important matters.

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

[0063] The historical operating data of the wind turbines is processed by removing the data of wind curtailment and correcting it using the LOF algorithm to obtain the processed historical operating data of the wind turbines.

[0064] The LOF algorithm is a density-based anomaly detection algorithm used to identify outliers in different density regions of a dataset. The LOF algorithm assesses the degree of anomaly of each data point by calculating its local density relative to its neighboring data points; that is, the LOF value of a data point represents its relative degree of anomaly compared to the density of its neighboring data points.

[0065] The processed historical wind turbine operation data is normalized to obtain a dataset, which is then divided into a training set, a validation set, and a test set.

[0066] Features are extracted from the training set to obtain the feature input set, and wavelet thresholding is used to denoise the wind power time series in the dataset.

[0067] Determine the input order of historical wind power data and meteorological data after noise reduction. Construct the prediction model input vector from the meteorological factor feature input set and the historical wind power order. Construct the wind turbine power prediction model through the input vector.

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

[0069] Perform wavelet transform on the original signal, determine the wavelet coefficients, and then perform thresholding.

[0070] w λ =sign(w)(|w|-λ)

[0071] In the formula, w λ The wavelet coefficients are processed by the threshold, w is the initial wavelet coefficients, and λ is the threshold.

[0072] Generally, the threshold processing method mainly includes two kinds: soft threshold function processing and hard threshold function processing.

[0073] Wavelet reconstruction is carried out by acquiring the wavelet coefficients after threshold processing:

[0074]

[0075] In the formula, SNR is a signal denoising evaluation index, s(i) is a pure signal, is a signal after denoising, and N is the sample quantity of the signal to be denoised.

[0076] In an example embodiment of the present application, the fan power prediction model is constructed by inputting a vector, and the method comprises the following steps:

[0077]

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

[0079] In an example embodiment of the present application, the wind farm is divided into regions, and a wind turbine group in each region is obtained, and the method comprises the following steps:

[0080] An area value and a wind turbine quantity value of a current wind farm are obtained, the area value is a value in a unit of square kilometers, and the number of divided regions is obtained:

[0081]

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

[0083] A two-dimensional plan view of the wind farm is grid divided, and the number of grids is equal to the number of divided regions.

[0084] In an example embodiment of the present application, the wind turbine selection evaluation model is established, and the method comprises the following steps:

[0085]

[0086] In the formula, a i is the distance of the i th fan from the center of the region, b i is the historical failure rate of the i th fan, N is the total number of wind turbines in the region, P i is the installed capacity of the i th fan, I i is the altitude value of the i th fan.

[0087] In one example embodiment of the present application, the predicted power of the wind turbine to be predicted is calculated by spatial data, comprising:

[0088]

[0089] wherein P x the predicted power of the wind turbine to be predicted, p is the air density, C p the wind power coefficient, v is the predicted wind speed received by the target wind turbine, theta is the angle between the X-axis in the plane coordinate system of the wind turbine to be predicted and the target wind turbine, omega is the angle between the Y-axis in the plane coordinate system of the wind turbine to be predicted and the target wind turbine, and d is the distance between the wind turbine to be predicted and the target wind turbine.

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

[0091] The prediction model establishing module is configured to obtain historical operation data of the wind turbine, the operation data comprising wind speed and power generation, establish a wind turbine power prediction model, train the wind turbine power prediction model by the historical operation data, and output the optimal wind turbine power prediction model.

[0092] The region division module is configured to divide the wind turbines of the wind farm into regions to obtain wind turbine groups in several regions, establish a wind turbine selection evaluation model, respectively evaluate the wind turbines in the several regions by 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.

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

[0094] The main control module is connected with the prediction model establishing module, the region division module and the prediction module, and is used for executing the wind farm power generation prediction method.

[0095] In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0096] If the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. The computer software product stored in a storage medium includes a plurality of 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 method of each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0097] The above is only the preferred embodiment of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A wind farm power production forecasting method, characterized by, The method comprises the following steps: acquiring historical operation data of a wind turbine, the operation data comprising wind speed and power generation, establishing a wind turbine power prediction model, training the wind turbine power prediction model through historical operation data, and outputting an optimal wind turbine power prediction model; dividing wind turbines of a wind farm into regions to obtain wind turbine groups in a plurality of regions, establishing a wind turbine selection evaluation model, respectively evaluating wind turbines in the plurality of regions through the wind turbine selection evaluation model to obtain evaluation values, and selecting a wind turbine with the highest evaluation value in a single region as a target wind turbine in the region; acquiring future predicted wind speed data output by a weather forecasting system, and outputting predicted power of the target wind turbine through the wind turbine power prediction model; acquiring spatial data between a to-be-predicted wind turbine and the target wind turbine in the same region, and calculating predicted power of the to-be-predicted wind turbine through the spatial data; comparing the predicted power of the to-be-predicted wind turbine obtained with historical operation data, setting a difference threshold, if the difference between the predicted power of the to-be-predicted wind turbine and the historical operation data is greater than the difference threshold, deleting the predicted power of the to-be-predicted wind turbine, outputting predicted power of the to-be-predicted wind turbine through the wind turbine power prediction model, if the difference between the predicted power of the to-be-predicted wind turbine and the historical operation data is not greater than the difference threshold, retaining the predicted power of the to-be-predicted wind turbine calculated through the spatial data; The method further comprises the following steps: acquiring an area value and a wind turbine quantity value of a current wind farm, the area value being a value in square kilometers, and obtaining a number of division regions; In the formula, Q is the number of regions, s x is the area value of the current wind farm, and R is the number of wind turbines. dividing a two-dimensional plan view of the wind farm into grids, the number of the grids being equal to the number of the division regions; The method further comprises the following steps: In the formula, a i is the distance of the i th fan from the center of the area, b i is the historical failure rate of the i th fan, N is the total number of fans in the area, P i is the installed power of the i th fan, I i is the altitude value of the i th fan; The method further comprises the following steps: wherein P x the predicted power of the wind turbine to be predicted, p is the air density, C p the wind power coefficient, v is the predicted wind speed received by the target wind turbine, θ is the angle between the X axis in the plane coordinate system of the target wind turbine and the wind turbine to be predicted, ω is the angle between the Y axis in the plane coordinate system of the target wind turbine and the wind turbine to be predicted, d is the distance between the target wind turbine and the wind turbine to be predicted, and A is the area swept by the wind wheel.

2. The wind farm power production forecasting method of claim 1, wherein, The method further comprises the following steps: performing outlier rejection and deletion on historical operation data of the wind turbine, and correcting the historical operation data of the wind turbine through a LOF algorithm to obtain processed historical operation data of the wind turbine; normalizing the processed historical operation data of the wind turbine to obtain a data set, and dividing the data set into a training set, a validation set and a test set; extracting features of the training set to obtain a feature input set, and performing wavelet threshold denoising on a wind power time sequence in the data set; determining an input order of denoised wind power historical data and meteorological data, constructing a prediction model input vector from a meteorological factor feature input set and a wind power historical order, and constructing a wind turbine power prediction model through the input vector.

3. A wind farm power production forecasting method according to claim 2, characterized in that, The method further comprises the following steps: performing wavelet transform on an original signal, determining wavelet coefficients, and performing threshold processing on the wavelet coefficients; w λ = sign(w)(|w| - λ) where w λ is the wavelet coefficient after thresholding, w is the initial wavelet coefficient, and λ is the threshold value. performing wavelet reconstruction on the wavelet coefficients processed through the threshold; In the formula, SNR is a signal denoising evaluation index, s(i) is a pure signal, is a denoised signal, and N is the number of denoised signal samples.

4. The method of claim 2, wherein, The method further comprises the following steps: In the formula, y(t) is a predicted power output value of the model, f() is a nonlinear function, x1(t-i) is an average wind speed, y(t-j) is a historical wind speed, ε(t) is noise, i is a model input order of the average wind speed, and j is a model input order of the historical wind speed.

5. A wind farm power production forecasting system, characterized by, The method comprises the following steps: The prediction model establishing module is configured to 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 an optimal wind turbine power prediction model; The region division module is configured to divide the wind turbines of the wind farm into regions to obtain wind turbine groups in a plurality of regions, establish a wind turbine selection evaluation model, respectively evaluate the wind turbines in the plurality of regions through the wind turbine selection evaluation model to obtain evaluation values, and select a wind turbine with the highest evaluation value in a single region as a target wind turbine in the region; The prediction module is configured to acquire future prediction wind speed data output by a weather forecasting system, output prediction power of the target wind turbine through the wind turbine power prediction model, acquire spatial data between a to-be-predicted wind turbine and the target wind turbine in the same region, calculate prediction power of the to-be-predicted wind turbine through the spatial data, compare the obtained prediction power of the to-be-predicted wind turbine with historical operation data, set a difference threshold, if a difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is greater than the difference threshold, delete the prediction power of the to-be-predicted wind turbine, output prediction power of the to-be-predicted wind turbine through the wind turbine power prediction model, and if the difference between the prediction power of the to-be-predicted wind turbine and the historical operation data is not greater than the difference threshold, retain the prediction power of the to-be-predicted wind turbine calculated through the spatial data; The main control module is connected with the prediction model establishing module, the region division module and the prediction module, and is configured to execute the wind farm power prediction method in any one of claims 1 to 4.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the wind farm power prediction method in any one of claims 1 to 4.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores the computer program, and the computer program is executed by the processor to implement the wind farm power prediction method in any one of claims 1 to 4.

Images