Wind turbine generator grid-connected test data processing method, device and equipment under extreme weather and medium

Abstract

The invention belongs to the technical field of data processing, and particularly relates to a wind turbine generator grid-connected test data processing method, device, equipment and medium under extreme weather, and the method comprises the steps: collecting a grid-connected test data set; removing abnormal data in the grid-connected test data set based on an adaptive multi-scale density clustering algorithm to obtain a first data set; and removing abnormal data in the first data set based on a Pauta criterion to obtain a processed data set. According to the method, the adaptive multi-scale density clustering algorithm is combined with the Pauta criterion, so that abnormal test data can be identified and eliminated; a cubic spline function-elastic network regularization dynamic fitting data curve is adopted, smooth curve fitting is achieved, and the method has the advantages of being high in accuracy, good in universality and the like in the aspect of wind turbine generator grid-connected test data processing under the extreme meteorological environment and the extreme meteorological environment.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data processing, and particularly relates to a wind turbine unit grid connection test data processing method, device, equipment and medium under extreme weather. BACKGROUND

[0002] At present, the power system gradually presents the "double high" characteristics of high proportion of new energy and high proportion of power electronic equipment, however, there are some potential hazards in the grid connection of new energy generating units, which are mainly caused by the inherent characteristics of new energy generation, such as intermittency and volatility, and the interaction between new energy generation equipment and the existing power grid.

[0003] Among new energy, renewable energy such as wind energy, light energy and water energy has been widely used. Since the wind turbine unit usually operates in a relatively harsh environment, there are a large number of abnormal data in the monitored wind turbine unit data, and the test data is inaccurate, and the unit regulation pressure increases. SUMMARY

[0004] The purpose of the present application is to provide a wind turbine unit grid connection test data processing method, device, equipment and medium under extreme weather, which solves the problem of increasing abnormal test data in wind turbine unit grid connection test data caused by extreme weather in the background art, and the problem of inaccurate test data.

[0005] In order to achieve the above purpose, the present application adopts the following technical scheme: The present application provides a grid connection test data processing method in the first aspect, comprising: Collecting a grid connection test data set; The grid connection test data set is divided according to the average distance of the nearest neighbor respectively, and high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data are obtained; Three windows are established, and the high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data are respectively corrected by sliding the three windows to obtain density hierarchical correction data, medium-density hierarchical correction data and low-density hierarchical correction data; Merging the high-density hierarchical correction data, medium-density hierarchical correction data and low-density hierarchical correction data to obtain a merged data set; determining the data points appearing at least twice and the data points appearing only in a single scale from the merged data set; deleting the data points appearing only in a single scale from the grid connection test data set to obtain a first data set; Based on the Laplace criterion, the abnormal data in the first data set is removed to obtain a processed data set.

[0006] Preferably, the step of obtaining the processed data set further comprises: The processed data set is converted into a curve through a cubic spline curve fitting model.

[0007] Preferably, in the step of transforming the processed data set into a curve by a cubic spline curve fitting model, an elastic network regularization term is introduced to constrain the cubic spline curve fitting model.

[0008] Preferably, in the step of dividing the grid-connected test data set into high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data according to the average distance of the nearest neighbors, the step comprises: calculating the average distance of the nearest neighbors of the data points in the grid-connected test data set; selecting three scales of high density, medium density and low density, and determining the minimum number of neighborhood radius and neighborhood radius midpoint based on the scale and the average distance of the nearest neighbors; obtaining the division result by adjusting the minimum number of neighborhood radius and neighborhood radius midpoint of the data set.

[0009] Preferably, in the step of obtaining the division result by adjusting the minimum number of neighborhood radius and neighborhood radius midpoint of the data set, the step comprises: in the dense area, the minimum number of neighborhood radius midpoint is increased; in the sparse area, the minimum number of neighborhood radius midpoint is decreased.

[0010] Preferably, in the step of establishing three windows and correcting the high-density hierarchical data, the medium-density hierarchical data and the low-density hierarchical data by sliding the three windows to obtain the density hierarchical correction data, the medium-density hierarchical correction data and the low-density hierarchical correction data, the step comprises: for the high-density hierarchical data, short-term fluctuation abnormal data is detected by high-frequency scale, and the high-density hierarchical correction data is obtained by removing the short-term fluctuation abnormal data; for the medium-density hierarchical data, periodic abnormal data is detected by medium-frequency scale, and the medium-density hierarchical correction data is obtained by removing the periodic abnormal data; for the low-density hierarchical data, long-term trend abnormal data is detected by low-frequency scale, and the low-density hierarchical correction data is obtained by removing the long-term trend abnormal data. Preferably, in the step of merging the high-density hierarchical correction data, the medium-density hierarchical correction data and the low-density hierarchical correction data to obtain a merged data set, determining the data points appearing at least twice and the data points appearing only at a single scale from the merged data set, and deleting the data points appearing only at a single scale from the grid-connected test data set to obtain a first data set, the step further comprises: for the data points filtered in all scales, checking whether the data point belongs to the boundary of the high-density hierarchical correction data, the medium-density hierarchical correction data and the low-density hierarchical correction data, and if it is a boundary, reclassifying it to the high-density hierarchical data, the medium-density hierarchical data or the low-density hierarchical data. Otherwise filter. In a second aspect, the application provides a wind turbine generator set grid-connection test data processing device under extreme weather, characterized in that, comprising: The acquisition module is configured to acquire a grid-connection test data set. The division module is configured to divide the grid-connection test data set according to the average distance of the nearest neighbors, to obtain high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data. The correction module is configured to establish three windows, and correct the high-density hierarchical data, the medium-density hierarchical data and the low-density hierarchical data by sliding the three windows, to obtain density hierarchical correction data, medium-density hierarchical correction data and low-density hierarchical correction data. The first processing module is configured to combine the high-density hierarchical correction data, the medium-density hierarchical correction data and the low-density hierarchical correction data, to obtain a combined data set; determine data points appearing at least twice and data points appearing only in a single scale from the combined data set; and delete the data points appearing only in the single scale from the grid-connection test data set, to obtain a first data set. The second processing module is configured to remove abnormal data in the first data set based on the Laplace criterion, to obtain a processed data set.

[0011] In a third aspect, the application provides an electronic device, characterized by comprising a processor and a memory, wherein the processor is configured to execute a computer program stored in the memory to implement the wind turbine generator set grid-connection test data processing method under extreme weather.

[0012] In a fourth aspect, the application provides a computer readable storage medium, characterized by storing at least one instruction, wherein the at least one instruction is executed by a processor to implement the wind turbine generator set grid-connection test data processing method under extreme weather.

[0013] Compared with the prior art, the application has the following beneficial effects: The adaptive multi-scale density clustering algorithm is combined with the Laplace criterion, so that abnormal test data can be identified and removed. The cubic spline function-elastic network regularization is used to dynamically fit the data curve, so that smooth curve fitting is achieved, and the wind turbine generator set grid-connection test data processing under extreme weather has the advantages of high accuracy and good universality. BRIEF DESCRIPTION OF DRAWINGS

[0014] The drawings accompanying the specification of the present application form a part thereof, serve to provide further understanding of the application, and together with the specification explain the application. The use of these drawings in explaining the application does not constitute an inappropriate limitation. In the drawings: Figure 1This is a flow chart of a method for processing wind turbine grid-connected test data under extreme weather conditions according to embodiment 1 of the present invention; Figure 2 This is a schematic diagram of a method for processing wind turbine grid-connected test data under extreme weather conditions according to Example 1 of the present invention; Figure 3 This is a wind turbine structure topology diagram of a method for processing wind turbine grid-connected test data under extreme weather conditions according to Example 1 of the present invention; Figure 4 This is a comparison diagram of the AM-DBSCAN effects of the wind turbine grid-connected test data processing method under extreme weather conditions in Example 1 of the present invention; Figure 5 This is a curve fitting diagram of the method for processing wind turbine grid-connected test data under extreme weather conditions according to Example 1 of the present invention; Figure 6 This is a structural block diagram of a data processing device for wind turbine grid-connected testing under extreme weather conditions according to an embodiment of the present invention; Figure 7 This is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0016] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise indicated, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present invention.

[0017] Example 1 like Figures 1-5 As shown in FIG, the data processing method for wind turbine grid connection test under extreme weather conditions includes: S1. Collecting a grid-connected test data set; the grid-connected test data set includes a number of data points; S2. Based on the adaptive multi-scale density clustering algorithm (AM-DBSCAN), the abnormal data in the network test data set is eliminated to obtain the first data set. The specific process is as follows: S21. Divide the grid-connected test data set according to the average neighbor distance to obtain high-density layer data, medium-density layer data, and low-density layer data, including: Define the neighborhood radius and the minimum number of points in the neighborhood radius MinPts; The average distance of each data point is calculated by traversing the data points in the grid test dataset. The neighborhood radius of different regions is dynamically determined by analyzing the histogram inflection point, and MinPts is adjusted according to the data distribution characteristics.

[0018] For the characteristics of wind power data, in the dense area, MinPts is set to a larger value to avoid excessive splitting; in the sparse area, MinPts is appropriately reduced to capture effective clusters. The grid test dataset is divided into three different density levels by setting different average distances of neighbors, i.e. high-density level data, medium-density level data, and low-density level data.

[0019] S22, three windows are established, and the high-density level data, medium-density level data and low-density level data are respectively corrected by sliding the three windows to obtain density level corrected data, medium-density level corrected data and low-density level corrected data; For high-density level data: detect short-term fluctuation anomalies (such as sensor burst noise, instantaneous overload), use small window and strict density threshold: short-term fluctuation anomalies in wind power data usually appear suddenly and last for a short time, and these anomalies show local and obvious different characteristics from surrounding data in time series. Using a small window can focus on local changes in data, and a strict density threshold can filter out data points that deviate from normal data distribution. Let the time series data be x 1, x 2, …, x n , and the size of the small window is w 1. For the i th data point, its small window is: ; In the formula, w 1 is the window parameter, which is used to control the time granularity of short-term fluctuation detection; W 1i is the window range, which is expanded by points on the left and right to form a symmetric window.

[0020] Under high-density level, set the sampling frequency to 10 minutes / time, and the average distance of neighbors is less than 0.2 m / s. If the wind speed fluctuation of three consecutive points exceeds 2 m / s within a 15-minute window, and the density in the window is lower than the threshold (MinPts=15), it is determined as short-term fluctuation anomaly data, and the short-term fluctuation anomaly data is removed to obtain high-density level corrected data.

[0021] For medium density level data: capture periodic anomalies (such as deviation from the diurnal power curve), window covers a complete period, allow certain density fluctuations, because wind power data has obvious periodicity, such as diurnal power curve changes are caused by periodic changes in environmental factors such as daytime and night wind speed, illumination, etc. When a periodic anomaly occurs, it means that within a complete period, the distribution of data deviates from the normal periodic pattern. Using a window that covers a complete period can ensure comprehensive analysis of data within the entire period, and allowing certain density fluctuations is because within a normal period, the density of data itself will naturally change over time. As long as this change is within a reasonable range, it should not be judged as an anomaly.

[0022] Assuming the period length is T , for the k th period, its small window is: ; , where T is the window parameter, representing the period length of the corresponding data; W 2k is the window range, indicating the data segment of the k th complete period, from k-1 ( T ) kT .

[0023] Under the medium density level, set the sampling frequency to 15 minutes / time, and the near-neighbor average distance to 0.3-0.6 kW. The normal night power should be maintained at 200-300 kW. Within a 24-hour window, if the number of points with power below 150 kW between 22:00 and 06:00 exceeds 20 (MinPts=8, exceeding the threshold), and the density fluctuation within the period exceeds the normal range (such as a 30% decrease in average density), it is determined to be periodic anomaly data, and the periodic anomaly data is removed to obtain the medium density level corrected data.

[0024] For low density level data: identify long-term trend anomalies (such as device performance degradation), window contains multiple days of data, density threshold is loose, because long-term trend anomalies are usually caused by gradual degradation of device performance due to factors such as aging and wear. This change is slow and continuous, and can only be clearly reflected over a long period of time. Using a window that contains multiple days of data can capture this long-term trend, and a loose density threshold can tolerate slow changes in data over a long period of time, avoiding misjudgment of normal long-term trends as anomalies.

[0025] Assuming the window size is w 3 days, and there are m data points per day, its small window is: ; wherein, w 3 is a window parameter, representing the number of days contained in the window, used to control the time span of long-term trend detection; m is the number of data points per day, determined by the sampling frequency; W 3l is the window range, representing the l th window containing 3-day data. w

[0026] Under the low-density level, the sampling frequency is set to 1 hour / time, the near-neighbor average distance is greater than 0.8 kW, and the average power output in the 7-day window continuously decreases from 500 kW last week to 450 kW. The near-neighbor average distance of the data points in the window exceeds 0.8 kW (normal ≤ 0.5 kW), and when the density threshold MinPts = 5, the power points deviate from the normal distribution for 3 consecutive days, which is determined as long-term trend abnormal data. The low-density level correction data is obtained by removing the long-term trend abnormal data.

[0027] As a specific example of the above embodiment, the high-density, medium-density, and low-density scales are selected, and the grid-connected test data set is divided three times based on the near-neighbor average distance. In the steps of obtaining high-density level data, medium-density level data, and low-density level data, respectively, the following steps are included: Randomly initialize a point x i If point x i is a core point, then point x i should contain at least MinPts points within its neighborhood radius, i.e.: ; wherein, is the number of data points within the neighborhood radius of point x i .

[0028] If point x i does not satisfy the above formula, then point x i is a boundary point, and a point is reinitialized. Within the neighborhood of the boundary point: ; wherein, is the number of data points within the neighborhood radius of point .

[0029] After initializing the first point, it is determined whether the points within its neighborhood radius are core points. If they are core points, it is still determined whether the points within their neighborhood radius are core points, until all points are processed. The abnormal points are coordinate points that are neither core points nor boundary points.

[0030] ​S23. Merge the high-density layer correction data, the medium-density layer correction data, and the low-density layer correction data to obtain a merged data set; determine data points that appear at least twice and data points that appear only at a single scale from the merged data set; delete the data points that appear at a single scale from the grid-connected test data set to obtain a first data set.

[0031] As a preferred example of the above embodiment, for data points that have been filtered at all scales, it is checked whether the data point belongs to the boundary of the density level correction data, the medium density level correction data, or the low density level correction data. If it is a boundary, it is reclassified into the high density level data, the medium density level data, or the low density level data. Otherwise filter.

[0032] S3, eliminating abnormal data in the first data set based on the Laida criterion to obtain a second data set; The density-based clustering algorithm DBSCAN can only identify outliers that are outside the clustering area. When the density of the sample set is uneven, that is, the sample density in the cluster group is different, the clustering effect of the density-based clustering algorithm may be affected. Therefore, the Laida criterion is used to identify outliers in the cluster area to improve clustering quality.

[0033] S31. Sample data: ; Where, n is the sample size, x n For the n Sample observations represent specific measurement data.

[0034] S32. Calculate the sample average m , reflecting the central tendency of the data: ; S33. Calculate sample variance , which measures the degree of discreteness of the data: ; According to the formula, three times the standard deviation is set as the confidence interval, and the data outside the confidence interval are deleted to obtain the second data set.

[0035] The second dataset is output as the processed dataset.

[0036] S4. Dynamic spline fitting data curve based on cubic spline function-elastic network regularization. The specific process is as follows: S41. To improve the fitting degree, a cubic spline curve fitting model is adopted. The cubic spline function is: ; wherein, a are polynomial coefficients that determine the shape of the curve within the interval, a 3 controls the cubic term curvature, a 2 controls the quadratic term bending degree, a 1 and a 0 control the linear offset and intercept; t i is the number of nodes that determines the flexibility of fitting.

[0037] The data fitting model is: ; wherein, is the error term of the data fitting function, obeys a normal distribution.

[0038] S42, the complexity of the model is constrained by introducing an elastic network regularization term, so as to prevent overfitting. The elastic network regularization least square estimation of the objective function is: ; S43, the local data volatility is calculated according to the time window σ , and λ 1 and λ 2 in the regularization term are adaptively adjusted, and the adaptive rules are as follows: ; wherein, λ 1 is used to constrain the sparsity of parameters; λ 2 is responsible for controlling the smoothness of the curve, which corresponds to the integral penalty term of the second derivative; k 1 and k 2 are reference adjustment coefficients, which are calibrated according to historical data.

[0039] Under high volatility, increasing λ1 inhibits the spline oscillation caused by noise, and reducing λ2 avoids the interference of sparsity constraint on core feature extraction; under stable state, λ1 weight is basically stable to maintain the basic smoothness constraint, and λ2 reduces the excessive inhibition of non-key features.

[0040] As a preferred example of the above embodiment, as shown in Figure 3 , the new energy unit adopts a permanent magnet direct drive wind turbine, mainly composed of a PMSG, a full-power converter and a filter.

[0041] Embodiment 2 As shown in Figure 6 , based on the same inventive concept as the above embodiment, the application further provides a wind turbine grid-connected test data processing device under extreme weather, comprising: An acquisition module is configured to acquire a grid-connected test data set. The dividing module is configured to divide the grid-connected test data set according to the average distance of the nearest neighbors to obtain high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data. The correcting module is configured to establish three windows, and correct the high-density hierarchical data, the medium-density hierarchical data and the low-density hierarchical data by sliding the three windows respectively to obtain density hierarchical corrected data, medium-density hierarchical corrected data and low-density hierarchical corrected data. The first processing module is configured to combine the high-density hierarchical corrected data, the medium-density hierarchical corrected data and the low-density hierarchical corrected data to obtain a combined data set, determine data points appearing at least twice and data points appearing only in a single scale from the combined data set, and delete the data points appearing only in the single scale from the grid-connected test data set to obtain a first data set. The second processing module is configured to remove abnormal data in the first data set based on the Laplace criterion to obtain a processed data set.

[0042] Embodiment 3 As shown in Figure 6 The application also provides an electronic device 100 for implementing the extreme weather wind turbine grid-connected test data processing method. The electronic device 100 comprises a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0043] The memory 101 can be used to store the computer program 103, and the processor 102 can implement the steps of the extreme weather wind turbine grid-connected test data processing method of embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0044] The memory 101 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function (such as a sound playing function, an image playing function, etc.), and the like; and the data storage area can store data (such as audio data) created according to the use of the electronic device 100, etc. In addition, the memory 101 can include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device.

[0045] The at least one processor 102 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic component, a discrete hardware component, or the like. The processor 102 can be a microprocessor or the processor 102 can also be any conventional processor and the like, and the processor 102 is a control center of the electronic device 100, and connects various parts of the entire electronic device 100 through various interfaces and lines.

[0046] The memory 101 in the electronic device 100 stores a plurality of instructions to implement an extreme weather wind turbine grid-connected test data processing method, and the processor 102 can execute the plurality of instructions to implement: Collecting a grid-connected test data set; Dividing the grid-connected test data set according to the nearest neighbor average distance respectively to obtain high-density hierarchical data, medium-density hierarchical data and low-density hierarchical data; Establishing three windows, and respectively correcting the high-density hierarchical data, the medium-density hierarchical data and the low-density hierarchical data by sliding the three windows to obtain density hierarchical correction data, medium-density hierarchical correction data and low-density hierarchical correction data; Merging the high-density hierarchical correction data, the medium-density hierarchical correction data and the low-density hierarchical correction data to obtain a merged data set; determining data points appearing at least twice and data points appearing only in a single scale from the merged data set; and deleting the data points appearing only in a single scale from the grid-connected test data set to obtain a first data set; Based on the Laplace criterion, the abnormal data in the first data set is removed to obtain a processed data set.

[0047] Embodiment 4 The modules / units integrated in the electronic device 100, if implemented in the form of software function units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, and read-only memory (ROM).

[0048] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0049] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system), and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The means for performing the functions specified in one or more flows and / or blocks.

[0050] These computer program instructions can also be stored in a computer-readable memory capable of guiding the computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction means, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The means for performing the functions specified in one or more flows and / or blocks.

[0051] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0052] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0053] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for processing wind turbine grid-connected test data under extreme weather conditions, characterized in that: include: Collect grid-connected test data sets; Dividing the grid-connected test data set according to the average distance of the nearest neighbor to obtain high-density layer data, medium-density layer data and low-density layer data; Three types of windows are established, and the high-density layer data, medium-density layer data and low-density layer data are corrected respectively by sliding the three windows to obtain density layer corrected data, medium-density layer corrected data and low-density layer corrected data; Merging the high-density layer-corrected data, the medium-density layer-corrected data, and the low-density layer-corrected data to obtain a merged data set; From the merged dataset, identify data points that appear at least twice, as well as data points that appear only at a single scale; Deleting data points that appear at a single scale from the grid-connected test data set to obtain a first data set; The abnormal data in the first data set are eliminated based on the Laida criterion to obtain a processed data set.

2. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 1, characterized in that: After the step of obtaining the processed data set, the method further includes: The processed data set is converted into a curve through a cubic spline fitting model.

3. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 2, wherein: In the step of converting the processed data set into a curve through a cubic spline curve fitting model, an elastic network regularization term is introduced to constrain the cubic spline curve fitting model.

4. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 1, wherein: The step of dividing the grid-connected test data set according to the average neighbor distance to obtain high-density layer data, medium-density layer data and low-density layer data includes: Calculate the average distance between the nearest neighbors of the data points in the grid connection test dataset; Select three scales: high density, medium density, and low density, and determine the neighborhood radius of the data point and the minimum number of points in the neighborhood radius based on the scale and the average distance of the nearest neighbors; The partitioning result is obtained by adjusting the neighborhood radius of the data set and the minimum number of points in the neighborhood radius.

5. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 4, characterized in that: The step of obtaining the partitioning result by adjusting the neighborhood radius of the data set and the minimum number of points in the neighborhood radius includes: In dense areas, increase the minimum number of points in the neighborhood radius; In sparse areas, the minimum number of points in the neighborhood radius is adjusted to a smaller value.

6. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 1, characterized in that: The step of establishing three windows and respectively correcting the high-density layer data, the medium-density layer data, and the low-density layer data by sliding the three windows to obtain density layer corrected data, medium-density layer corrected data, and low-density layer corrected data includes: For high-density level data, short-term fluctuation abnormal data is detected through high-frequency scale, and short-term fluctuation abnormal data is eliminated to obtain high-density level corrected data; For medium-density level data, periodic abnormal data are detected by medium-frequency scale, and the periodic abnormal data are eliminated to obtain the medium-density level corrected data; For low-density layer data, long-term trend anomaly data is detected through low-frequency scale, and long-term trend anomaly data is eliminated to obtain low-density layer corrected data.

7. The method for processing wind turbine grid connection test data under extreme weather conditions according to claim 1, wherein: The merging of the high-density level-corrected data, the medium-density level-corrected data, and the low-density level-corrected data to obtain a merged data set; From the merged dataset, identify data points that appear at least twice, as well as data points that appear only at a single scale; The step of deleting data points that appear at a single scale from the grid connection test data set to obtain the first data set further includes: For data points that have been filtered at all scales, check whether the data point belongs to the boundary of high-density layer correction data, medium-density layer correction data, and low-density layer correction data. If it is the boundary, reclassify it to high-density layer data, medium-density layer data, or low-density layer data; Otherwise filter.

8. A data processing device for wind turbine grid connection test under extreme weather conditions, characterized in that: include: Acquisition module, used to collect grid-connected test data sets; A partitioning module is used to partition the grid-connected test data set according to the average distance of the nearest neighbors to obtain high-density layer data, medium-density layer data and low-density layer data; A correction module is used to establish three windows, and correct the high-density layer data, the medium-density layer data and the low-density layer data respectively by sliding the three windows to obtain density layer correction data, medium-density layer correction data and low-density layer correction data; A first processing module merges the high-density layer corrected data, the medium-density layer corrected data, and the low-density layer corrected data to obtain a merged data set; From the merged dataset, identify data points that appear at least twice, as well as data points that appear only at a single scale; Deleting data points that appear at a single scale from the grid-connected test data set to obtain a first data set; The second processing module removes abnormal data in the first data set based on the Laida criterion to obtain a processed data set.

9. An electronic device, characterized in that: The system comprises a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the method for processing wind turbine grid-connected test data under extreme weather conditions as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, the method for processing wind turbine grid-connected test data under extreme weather conditions according to any one of claims 1 to 7 is implemented.

Images