Wind turbine generator power data cleaning method and system and storage medium

Abstract

The wind turbine generator power data cleaning method provided by the invention comprises the following steps: acquiring a data file of a wind turbine generator and removing invalid data in the data file; determining a data point with the maximum wind speed in the fluctuation interval of the rated power as an ideal data point, and determining a boundary inflection point according to a power curve near the cut-in wind speed and a power scatter rising period curve; determining a mechanism auxiliary line according to the ideal data points and the boundary inflection points, and eliminating data points which are greatly deviated from power scatter points in a normal operation power curve range in the data file based on the mechanism auxiliary line so as to determine a cleaned first data file; performing segmentation processing on the first data file according to a preset wind speed interval to determine wind speed sections, and eliminating abnormal values in each wind speed section through a quartile method to determine a cleaned second data file; and cleaning the second data file based on a Gaussian probability density method to determine a cleaned third data file.

Description

Technical Field

[0001] The present invention relates to the field of wind power, and in particular to a method for cleaning power data of a wind turbine generator set, a system for cleaning power data of a wind turbine generator set, and a computer-readable storage medium. Background Art

[0002] With the rapid development of my country's wind power industry, a large number of newly built wind farms have been connected to the grid, and the pilot projects of electricity spot trading have gradually spread across the country, and the requirements for wind power prediction accuracy are gradually increasing. However, the efficiency of old wind farms built in the early stage has declined, and various problems have become increasingly prominent when facing transformation and upgrading. Among them, the cleaning of wind turbine data and the fitting of power curves are crucial for energy efficiency evaluation and improving wind power prediction accuracy.

[0003] In the prior art, there are two types of methods for cleaning wind turbine data. The first type is to use statistical methods, such as density, median, variance, etc. However, statistical methods are not ideal for data with a large proportion of abnormal data, and a large amount of normal data will be eliminated. In addition, the parameters of wind turbines with different data characteristics need to be manually adjusted by technicians, which has a great impact on the data processing results. The second type is a method based on data mining. In a large amount of wind turbine data, a machine learning method is used to learn the abnormalities of feature performance, but this method cannot effectively identify high-density accumulation-type abnormal data.

[0004] In order to overcome the above-mentioned defects of the prior art, the art is in urgent need of a wind turbine power data cleaning method, a wind turbine power data cleaning system, and a computer-readable storage medium. By combining the mechanism and mathematical methods, the wind turbine power data can be effectively cleaned under the condition of meeting the operating trend of the wind turbine, and the cleaning effect is excellent. Summary of the invention

[0005] A brief summary of one or more aspects is given below to provide a basic understanding of these aspects. This summary is not an exhaustive overview of all conceived aspects, and is neither intended to identify the key or critical elements of all aspects nor to define the scope of any or all aspects. Its only purpose is to give some concepts of one or more aspects in a simplified form as a prelude to a more detailed description that will be given later.

[0006] In order to overcome the above-mentioned defects existing in the prior art, the present invention provides a wind turbine power data cleaning method, a wind turbine power data cleaning system, and a computer-readable storage medium. By combining the mechanism and mathematical methods, the wind turbine power data can be effectively cleaned under the condition of conforming to the operating trend of the wind turbine, and has an excellent cleaning effect.

[0007] Specifically, the above-mentioned wind turbine power data cleaning method provided according to the first aspect of the present invention includes the steps of: obtaining the data file of the wind turbine and eliminating invalid data in the data file; determining the data point with the maximum wind speed within the fluctuation range of the rated power as the ideal data point, and determining the boundary inflection point according to the power curve near the cut-in wind speed and the power scatter point rising period curve; determining the mechanism auxiliary line according to the ideal data point and the boundary inflection point, and eliminating the data points in the data file that deviate greatly from the power scatter points within the normal operating power curve range based on the mechanism auxiliary line to determine the first data file after cleaning; segmenting the first data file according to the preset wind speed interval to determine the wind speed interval of the first data file, and eliminating the outliers in each wind speed interval by the interquartile range method to determine the second data file after cleaning; and cleaning the second data file based on the Gaussian probability density method to determine the third data file after cleaning.

[0008] Preferably, in an embodiment of the present invention, the data file includes a time field, a power field, and a wind speed field.

[0009] Preferably, in an embodiment of the present invention, the data file includes at least 3 months of historical data files, and the time resolution of the data file is 10 minutes.

[0010] Preferably, in an embodiment of the present invention, the data file is a historical data file of less than 3 months, and the time resolution of the data file is 1 minute.

[0011] Preferably, in an embodiment of the present invention, the invalid data includes data of wind speed less than 0 m / s and data of power greater than a preset power threshold range.

[0012] Preferably, in an embodiment of the present invention, the fluctuation range of the rated power is 90% to 110% of the rated power.

[0013] Preferably, in one embodiment of the present invention, the step of segmenting the first data file according to a preset wind speed interval to determine the wind speed interval segments of the first data file, and eliminating abnormal values ​​in each of the wind speed interval segments by the interquartile range method includes: after arranging the data in each wind speed interval segment, determining the data at the 25th% position as the first quartile and the data at the 75th% position as the third quartile; determining the interquartile range based on the difference between the third quartile and the first quartile; and determining the normal value range in each wind speed interval segment of the first data file based on the interquartile range to eliminate the abnormal values.

[0014] Preferably, in one embodiment of the present invention, the step of cleaning the second data file based on the Gaussian probability density method to determine the cleaned third data file includes: determining the probability density value of each data point in the wind speed interval segment through a Gaussian probability density function based on the mean and standard deviation in the wind speed interval segment of the second data file; and eliminating abnormal values ​​in the wind speed interval segment whose probability density values ​​are lower than a preset probability density threshold to determine the cleaned third data file.

[0015] In addition, the wind turbine power data cleaning system provided according to the second aspect of the present invention includes a memory and a processor. The memory stores computer instructions. The processor is connected to the memory and is configured to execute the computer instructions stored in the memory to implement the wind turbine power data cleaning method provided in any one of the above embodiments.

[0016] In addition, the computer-readable storage medium provided according to the third aspect of the present invention stores computer instructions. When the computer instructions are executed by the processor, the wind turbine power data cleaning method provided by any one of the above embodiments is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above features and advantages of the present invention can be better understood after reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings. In the drawings, the components are not necessarily drawn to scale, and components with similar related properties or features may have the same or similar reference numerals.

[0018] Figure 1 A flow chart of a wind turbine generator power data cleaning method provided in some embodiments of the present invention is shown;

[0019] Figure 2 A schematic diagram showing a mechanism auxiliary line for cleaning wind turbine power data provided in accordance with some embodiments of the present invention is shown;

[0020] Figure 3A-3B A diagram showing the effect of wind turbine power data cleaning provided in some embodiments of the present invention;

[0021] Figures 4A to 4B A diagram showing the effect of wind turbine power data cleaning provided by some embodiments of the present invention; and

[0022] Figure 5 A schematic diagram of a wind turbine generator power data cleaning system provided according to some embodiments of the present invention is shown.

[0023] Reference numerals:

[0024] 100: Wind turbine power data cleaning method;

[0025] 201: ideal data point;

[0026] 202: Boundary inflection point;

[0027] 210: Mechanism auxiliary line;

[0028] 500: Wind turbine power data cleaning system;

[0029] 510: memory;

[0030] 511: Computer readable storage medium;

[0031] 520: processor; and

[0032] S110~S150: steps. DETAILED DESCRIPTION

[0033] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. Note that the aspects described below in conjunction with the accompanying drawings and specific embodiments are only exemplary and should not be construed as limiting the scope of protection of the present invention in any way.

[0034] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0035] In addition, the terms "upper", "lower", "left", "right", "top", "bottom", "horizontal" and "vertical" used in the following description should be understood as the directions shown in the paragraph and the related drawings. Such relative terms are only used for the convenience of description and do not mean that the device described therein must be manufactured or operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0036] It is understood that although the terms "first", "second", "third", etc. may be used herein to describe various components, regions, layers and / or parts, these components, regions, layers and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers and / or parts. Therefore, the first component, region, layer and / or part discussed below may be referred to as a second component, region, layer and / or part without departing from some embodiments of the present invention.

[0037] As mentioned above, in the prior art, there are two types of methods for cleaning wind turbine data. The first type is to use statistical methods, such as density, median, variance, etc. However, statistical methods are not ideal for data with a large proportion of abnormal data, and a large amount of normal data will be eliminated. In addition, the parameters of wind turbines with different data characteristics need to be manually adjusted by technicians, which has a great impact on the data processing results. The second type is a method based on data mining. In a large amount of wind turbine data, a machine learning method is used to learn the abnormalities of feature performance, but this method cannot effectively identify high-density accumulation-type abnormal data.

[0038] In order to overcome the above-mentioned defects of the prior art, the art is in urgent need of a wind turbine power data cleaning method, a wind turbine power data cleaning system, and a computer-readable storage medium. By combining the mechanism and mathematical methods, the wind turbine power data can be effectively cleaned under the condition of meeting the operating trend of the wind turbine, and the cleaning effect is excellent.

[0039] In some non-limiting embodiments, the wind turbine generator power data cleaning method provided in the first aspect of the present invention may be implemented via the wind turbine generator power data cleaning system provided in the second aspect of the present invention.

[0040] The following will first describe the working principle of the above-mentioned wind turbine power data cleaning system in conjunction with some embodiments of the wind turbine power data cleaning method. Those skilled in the art will understand that these embodiments of the wind turbine power data cleaning method are only some non-limiting implementation methods provided by the present invention, which are intended to clearly demonstrate the main concept of the present invention and provide some specific solutions that are convenient for the public to implement, rather than to limit all functions or all working modes of the wind turbine power data cleaning system. Similarly, the wind turbine power data cleaning system is also only a non-limiting implementation method provided by the present invention, and does not constitute a limitation on the execution subject and execution order of each step in these wind turbine power data cleaning methods.

[0041] Please refer to Figure 1 , Figure 1 A flow chart of a wind turbine generator set power data cleaning method provided according to some embodiments of the present invention is shown.

[0042] like Figure 1 As shown, the wind turbine generator power data cleaning method 100 may include step S110: acquiring a data file of the wind turbine generator and removing invalid data in the data file.

[0043] The data file of the wind turbine can be the wind turbine SCADA (Supervisory Control and Data Acquisition) historical data. SCADA is a control system widely used in the field of industrial automation and can be used for monitoring and controlling wind turbines.

[0044] The data file includes at least a time field, a power field and a wind speed field. Preferably, the data file may include at least 3 months of historical data files with a time resolution of 10 minutes. If the historical data file is less than 3 months, the time resolution of the acquired historical data file is at least 1 minute.

[0045] The wind turbine power data cleaning system can obtain the data file of the above-mentioned wind turbine and pre-process the data file. Pre-processing can include removing invalid data in the data file, and the invalid data can be data with a wind speed less than 0m / s. Data with a wind speed less than 0m / s is physically unreasonable. These data are usually generated by noise or sensor errors in data acquisition. Therefore, data with a wind speed less than 0m / s can be removed. Invalid data can also be data with a power greater than a preset power threshold range, and the preset power threshold range can be set to a value that exceeds the rated power by a lot. Since there is a situation where the actual power of the data point is slightly higher than the rated power, the data that fluctuates within a reasonable range near the rated power does not need to be removed. By removing data with a power greater than the preset power threshold range, that is, removing data that exceeds the rated power by a lot, invalid data that obviously exceeds the rated power of the wind turbine can be excluded, avoiding erroneous measurements or affecting analysis due to recording problems.

[0046] Therefore, by preprocessing the data files, the initial invalid data is eliminated, the interference of noise data is reduced, and the accuracy of the subsequent cleaning process is ensured.

[0047] like Figure 1 As shown, the wind turbine power data cleaning method 100 may include step S120: determining the data point with the maximum wind speed within the rated power fluctuation range as the ideal data point, and determining the boundary inflection point according to the power curve near the cut-in wind speed and the power scatter point rising period curve.

[0048] And step S130: determine the mechanism auxiliary line according to the ideal data point and the boundary inflection point, and remove the data points in the data file that deviate greatly from the power scatter points within the normal operating power curve based on the mechanism auxiliary line to determine the first data file after cleaning.

[0049] High wind speed and low power points will seriously affect the data cleaning effect, and then affect the subsequent effects of clustering, density, fitting and other methods, so they need to be eliminated first. The wind turbine power data cleaning system can eliminate data points that deviate greatly from the power scatter points within the normal operating power curve by establishing auxiliary lines, thereby eliminating high wind speed and low power abnormal points in the preprocessed data file to determine the first data file after cleaning.

[0050] Since the power scatter point trend of each model or even unit is different, it is impossible to find abnormal points by specifying high wind speed thresholds and low power thresholds. The wind turbine power data cleaning system provided by the present invention can establish auxiliary lines through two data points that conform to the operation mechanism of the wind turbine, and effectively eliminate a large number of high wind speed and low power abnormal points that deviate too much from the power scatter point trend during actual operation. Here, the mechanism auxiliary line and the data points used to establish the mechanism auxiliary line can be adaptively valued according to the actual power scatter points of different models and units.

[0051] The wind turbine power data cleaning system can first determine the rated power of the wind turbine, and determine the data point with the maximum wind speed within the fluctuation range of the rated power as the ideal data point.

[0052] Preferably, the wind turbine power data cleaning system can select a fluctuating power scatter point where the power is maintained within the power range of ±10% of the rated power, that is, the rated power fluctuation range can be 90% to 110% of the rated power. Then, among the fluctuating power scatter points where the power is within the fluctuation range of the rated power, an ideal data point with the maximum wind speed is found.

[0053] Each model has a theoretical power curve, and the power reaches the rated (full) power after reaching a rated wind speed. For example, a certain model generates electricity at 5000kW after the wind speed exceeds 10m / s. However, in actual situations, after the wind speed exceeds 10m / s, the power of a certain model may not be maintained at 5000kW, but fluctuates within a range near it. Therefore, the wind turbine power data cleaning system can find the data point with the largest wind speed among all the fluctuating power scatter points within the range of 10% above and below the rated power of 5000kW that should be achieved as the data point closest to the ideal.

[0054] Then, the wind turbine power data cleaning system can determine the boundary inflection point based on the power curve near the cut-in wind speed and the power scatter point rising period curve. Near the cut-in wind speed interval, the power curve rises slowly; when the power scatter point is in the rising period interval, the power curve rises quickly. The inflection point of these two power curves is the boundary inflection point of the power curve near the cut-in wind speed and the power scatter point rising period curve.

[0055] According to the above-mentioned ideal data points and boundary inflection points, the slope and intercept are calculated, thereby the wind turbine power data cleaning system can establish a mechanism auxiliary line for preliminarily identifying outliers in the power curve.

[0056] Afterwards, data points with large deviations from power scatter points within the normal operating power curve range are removed from the data file based on the mechanism auxiliary line to determine the first data file after cleaning. Here, data points with large deviations from power scatter points within the normal operating power curve range (i.e., high wind speed and low power abnormal points) can be located below the mechanism auxiliary line.

[0057] Please refer to Figure 2 , Figure 2 A schematic diagram of a mechanism auxiliary line for cleaning wind turbine power data provided according to some embodiments of the present invention is shown.

[0058] like Figure 2 As shown, the rated power of this model is 5000kW, and the wind turbine power data cleaning system can find the ideal data point 201 with the maximum wind speed among all the fluctuating power scatter points within the range of 10% above and below the rated power of 5000kW. In addition, the cut-in wind speed interval can be 2.5-5m / s. In this wind speed interval, the power curve rises slowly, that is, the power curve with a wind speed interval of 2.5-5m / s can be determined as the power curve near the cut-in wind speed; correspondingly, in the wind speed interval of 5-10m / s, the power curve rises rapidly, that is, the power curve with a wind speed interval of 5-10m / s can be determined as the power scatter point rising period curve. According to the power curve near the cut-in wind speed and the power scatter point rising period curve, the boundary inflection point 202 can be determined. By connecting the ideal data point 201 and the boundary inflection point 202, the mechanism auxiliary line 210 can be determined. The data points below the mechanism auxiliary line 210 are abnormal points, that is, data points that deviate greatly from the power scatter points within the normal operating power curve range. After the outliers are removed from the data file, the first data file after cleaning can be determined.

[0059] Please continue to refer to Figure 1 The wind turbine power data cleaning method 100 may include step S140: segmenting the first data file according to a preset wind speed interval to determine the wind speed interval of the first data file, and removing abnormal values ​​in each wind speed interval by the interquartile range method to determine a cleaned second data file.

[0060] According to the wind speed field included in the first data file, the data included in the first data file can be segmented according to the preset wind speed interval to determine the wind speed interval segment. When the data power scatter point density is high, if the first data file includes at least 3 months of historical data, the wind turbine power data cleaning system can use a wind speed interval of 0.5m / s for segmentation; if the historical data in the first data file is less than 3 months, the wind turbine power data cleaning system can use a wind speed interval of 0.1m / s for segmentation.

[0061] The manifestations of abnormal points in different wind speed ranges are not the same. By processing the data in each wind speed range separately, the wind turbine power data cleaning system can ensure the accuracy of subsequent abnormal point identification, thereby better handling the changes in wind turbine power under different wind speeds.

[0062] Afterwards, the wind turbine power data cleaning system can remove outliers in each wind speed interval segment through the interquartile range method to determine the second data file after cleaning. Through the interquartile range method, the wind turbine power data cleaning system can effectively process the data segments in the first data file where the proportion of outliers is not large, avoiding the influence of extreme points on the analysis results.

[0063] Specifically, the wind turbine power data cleaning system can first calculate the quartiles of the power distribution in each wind speed interval. After arranging the data in each wind speed interval, the data at the 25% position is determined to be the first quartile Q1, i.e., the lower quartile; the data at the 75% position is determined to be the third quartile Q3, i.e., the upper quartile.

[0064] The interquartile range (IQR) can be used to indicate the degree of dispersion of the middle 50% of the data distribution. The IQR is determined by the difference between the third quartile Q3 and the first quartile Q1: IQR = Q3-Q1. The larger the value of IQR, the wider the data distribution, and vice versa, the more concentrated the data distribution. Based on the interquartile range, the wind turbine power data cleaning system can determine the range of normal data in each wind speed interval and eliminate outliers based on the range of normal data. Preferably, the upper and lower limits of the range of normal data can be set as:

[0065] Lower limit = Q1-n*IQR;

[0066] Upper limit=Q3+n*IQR;

[0067] Where n is a multiple of IQR.

[0068] Thus, the wind turbine power data cleaning system can determine that data beyond the upper and lower limits of the normal data range are abnormal values, and remove these abnormal values ​​from the first data file, thereby determining the cleaned second data file.

[0069] like Figure 1 As shown, the wind turbine power data cleaning method 100 may include step S150: cleaning the second data file based on the Gaussian probability density method to determine a cleaned third data file.

[0070] The wind turbine power data cleaning method may also include cleaning the second data file based on the Gaussian probability density method. Through the Gaussian probability density method, the wind turbine power data cleaning system can further eliminate abnormal points in the cleaned second data file that the interquartile range method fails to identify, especially for abnormal values ​​that deviate from the mean in distribution, the Gaussian probability density method has a good recognition effect.

[0071] Specifically, the wind turbine power data cleaning system can determine the probability density value of each data point in the wind speed interval segment through a Gaussian probability density function according to the mean and standard deviation in the wind speed interval segment of the second data file.

[0072] The mean μ in each wind speed interval can be expressed as follows:

[0073]

[0074] Among them, P i is the power value of the data point of the ith record, and N is the number of data points in the wind speed interval.

[0075] The standard deviation σ within each wind speed interval can be expressed as follows:

[0076]

[0077] After that, the Gaussian Probability Density Function (PDF) is used to determine the probability density value of each data point in the wind speed interval. Assuming that the power value of the data point in each wind speed interval follows a Gaussian distribution (normal distribution), the power value P of the data point in each wind speed interval is i Substitute it into the Gaussian distribution model and calculate the probability density value of the data point. Its probability density function is:

[0078]

[0079] Where x can be the power value P of the data point i, f(x|μ,σ) is the probability density value of each data point. The probability density value indicates the possibility of a certain power value within a given wind speed range. The larger the probability density value, the more the data point conforms to the Gaussian distribution, that is, the more it conforms to the characteristics of normal data. If the probability density value of a power value is very low, it means that the data point may be an outlier.

[0080] Points with too low probability density values ​​may be outliers or noise. These data points rarely appear in actual operation and are not representative for the analysis of the power curve. By setting the probability density threshold, data points with probability density values ​​lower than the probability density threshold in the wind speed interval are removed and marked as outliers, thereby determining the cleaned third data file.

[0081] Please refer to Figure 3A-3B and Figures 4A to 4B , Figure 3A-3B and Figures 4A to 4B A diagram showing the effect of wind turbine generator power data cleaning provided according to some embodiments of the present invention is shown.

[0082] Existing cleaning methods are more limited when the amount of data is smaller. Figure 3A The scatter plot of wind speed and power of the historical power scatter data of wind turbine No. 12 in a certain wind farm within one month is shown. The wind turbine power data cleaning method provided by the present invention also has a good cleaning effect on data files with a small amount of data (less than 3 months) and poor clustering effect. The scatter plot after cleaning is as follows Figure 3B As shown, the wind turbine power data cleaning method provided by the present invention can effectively clean the abnormal values ​​in the historical data within 1 month.

[0083] In addition, when cleaning the power scatter points of old stock wind farms based on SCADA data, the power scatter points of SCADA data are very chaotic when the wind turbine operating status labels cannot be obtained, and there are a large number of inefficient power scatter points in abnormal operating ranges. Therefore, when using mathematical or machine learning methods alone to clean the power scatter points of SCADA data, the feature or clustering effect is not obvious. Figure 4A The scatter plot of wind speed and power of the historical power scatter data of the old 1500kW wind turbine No. 11 in a certain wind farm within one year is shown. The wind turbine power data cleaning method provided by the present invention combines the mechanism and mathematical methods. Under the condition of meeting the operation trend of the wind turbine, it can effectively clean the data and has an excellent cleaning effect on the power scatter of old wind turbines, inefficient wind turbines, and abnormal wind turbines. The scatter plot after cleaning is shown in Figure 1. Figure 4B As shown, the wind turbine power data cleaning method provided by the present invention can effectively clean the abnormal values ​​in the historical data of old wind turbines within 1 year.

[0084] In summary, the wind turbine power data cleaning method and wind turbine power data cleaning system provided by the present invention adopt a mechanism elimination method combined with quartile fuzzy elimination and Gaussian probability density method to establish a set of adaptive wind turbine data cleaning methods, which not only solves the limitations of existing statistical methods and eliminates the need for complex and time-consuming manual cleaning of wind turbines with different data characteristics, but also reduces the instability of existing machine learning black box algorithms during cluster cleaning. In addition, adaptive parameter adjustment can be performed according to the data conditions of different wind turbines.

[0085] Please refer to Figure 5 , Figure 5 A schematic diagram of a wind turbine generator power data cleaning system provided according to some embodiments of the present invention is shown.

[0086] like Figure 5 As shown, the wind turbine power data cleaning system 500 for implementing the wind turbine power data cleaning method may be configured with a memory 510 and a processor 520. The memory 510 includes but is not limited to the computer-readable storage medium 511 provided in the third aspect of the present invention, on which computer instructions are stored. The processor 520 is connected to the memory 510 and is configured to execute the computer instructions stored in the memory 510 to implement the wind turbine power data cleaning method provided in the first aspect of the present invention.

[0087] Although the above methods are illustrated and described as a series of actions for simplicity of explanation, it should be understood and appreciated that these methods are not limited by the order of the actions, because according to one or more embodiments, some actions may occur in a different order and / or concurrently with other actions from those illustrated and described herein or not illustrated and described herein but understandable to those skilled in the art.

[0088] Those skilled in the art will appreciate that information, signals, and data may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips cited throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or optical particles, or any combination thereof.

[0089] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or a combination of the two. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps are generally described above in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. The technician may implement the described functionality in different ways for each specific application, but such implementation decisions should not be interpreted as resulting in a departure from the scope of the present invention.

[0090] The various illustrative logic modules and circuits described in conjunction with the embodiments disclosed herein may be implemented or executed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in cooperation with a DSP core, or any other such configuration.

[0091] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read and write information from / to the storage medium. In an alternative, a storage medium may be integrated into a processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and the storage medium may reside in a user terminal as discrete components.

[0092] In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented as a computer program product in software, each function may be stored on or transmitted by a computer-readable medium as one or more instructions or codes. Computer-readable media include both computer storage media and communication media, including any medium that facilitates the transfer of a computer program from one place to another. Storage media may be any available medium that can be accessed by a computer. As an example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer. Any connection is also properly referred to as a computer-readable medium. For example, if the software is transmitted from a website, a server, or other remote source using a coaxial cable, a fiber optic cable, a twisted pair, a digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwaves, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwaves are included in the definition of the medium. Disk and disc as used herein include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, wherein disk often reproduces data magnetically, while disc reproduces data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0093] The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein, but should be granted the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for cleaning wind turbine power data, characterized in that: Includes steps: Obtaining a data file of a wind turbine generator set and removing invalid data from the data file; The data point with the maximum wind speed within the fluctuation range of the rated power is determined as the ideal data point, and the boundary inflection point is determined based on the power curve near the cut-in wind speed and the power scatter point rising period curve; Determine a mechanism auxiliary line according to the ideal data point and the boundary inflection point, and remove data points in the data file that deviate greatly from power scatter points within the normal operating power curve range based on the mechanism auxiliary line to determine a first data file after cleaning; Segmenting the first data file according to a preset wind speed interval to determine the wind speed interval of the first data file, and removing abnormal values ​​in each wind speed interval by the interquartile range method to determine a cleaned second data file; as well as The second data file is cleaned based on a Gaussian probability density method to determine a cleaned third data file.

2. The wind turbine power data cleaning method according to claim 1, characterized in that: The data file includes a time field, a power field, and a wind speed field.

3. The wind turbine power data cleaning method according to claim 1, characterized in that: The data files include historical data files of at least 3 months, and the time resolution of the data files is 10 minutes.

4. The wind turbine power data cleaning method according to claim 1, characterized in that: The data file is a historical data file of less than 3 months, and the time resolution of the data file is 1 minute.

5. The wind turbine power data cleaning method according to claim 1, characterized in that: The invalid data includes data with a wind speed less than 0 m / s and data with a power greater than a preset power threshold range.

6. The wind turbine power data cleaning method according to claim 1, characterized in that: The fluctuation range of the rated power is 90% to 110% of the rated power.

7. The wind turbine power data cleaning method according to claim 1, characterized in that: The step of segmenting the first data file according to the preset wind speed interval to determine the wind speed interval segments of the first data file, and removing abnormal values ​​in each wind speed interval segment by using the interquartile range method comprises: After arranging the data in each wind speed interval, the data at the 25% position is determined as the first quartile, and the data at the 75% position is determined as the third quartile; Determining the interquartile range according to the difference between the third quartile and the first quartile; and A normal value range in each wind speed interval of the first data file is determined based on the interquartile range to eliminate the abnormal value.

8. The wind turbine power data cleaning method according to claim 1, characterized in that: The step of cleaning the second data file based on the Gaussian probability density method to determine the cleaned third data file includes: Determining the probability density value of each data point in the wind speed interval segment by using a Gaussian probability density function according to the mean and standard deviation in the wind speed interval segment of the second data file; and The abnormal values ​​whose probability density values ​​in the wind speed interval are lower than a preset probability density threshold are eliminated to determine the cleaned third data file.

9. A wind turbine power data cleaning system, characterized in that: include: a memory having computer instructions stored thereon; as well as A processor is connected to the memory and is configured to execute computer instructions stored in the memory to implement the wind turbine power data cleaning method according to any one of claims 1 to 8.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed by the processor, the wind turbine power data cleaning method according to any one of claims 1 to 9 is implemented.

Images