A wind turbine data processing method and device

Abstract

The present invention provides a method and device for processing wind turbine data, which relates to the field of wind turbine data processing. The method comprises: dividing wind speed data into multiple intervals according to a set step length, obtaining wind turbine operation data corresponding to each interval; calculating the actual gradient change amplitude of the wind turbine operation data in each interval; when the actual gradient change amplitude of a certain wind turbine operation data is lower than the reference gradient change amplitude of the set power curve, marking the certain wind turbine operation data as abnormal and eliminating it; based on the 3σ criterion, calculating the average value and standard deviation corresponding to the wind turbine operation data in each interval; determining the confidence interval of each interval based on the average value and standard deviation; according to the confidence interval, eliminating the abnormal data in the wind turbine operation data in each interval. The method is used to improve the problem of insufficient abnormal data identification and cleaning of traditional wind turbines.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine generator set data processing, and in particular to a wind turbine generator set data processing method and device. Background Art

[0002] In the actual operation process, wind turbines will mistakenly record some abnormal data due to sensor failure, transmission and storage errors, and the influence of control strategies such as power restrictions and shutdowns for maintenance. These abnormal data are mixed with normal operation data. However, abnormal data is not a true reflection of the operating status of wind turbines. Therefore, under the background of big data applications, in order to ensure data quality and provide reliable data guarantee for wind turbine control and prediction research, the recognition and cleaning technology of abnormal data has gradually attracted people's attention. Among them, a lot of research work has been carried out at home and abroad on the recognition and cleaning technology of abnormal power data of wind turbines.

[0003] At present, the identification methods of abnormal data of wind turbines are mainly divided into methods based on statistical analysis, image analysis, machine learning, and detection methods combined with the similarity of operating status. For example, the anomaly identification method based on prediction model and fuzzy theory has proved its accuracy through case analysis. In addition, there are methods for outlier detection using standard deviation method, box plot method, etc., as well as the detection of data anomalies by calculating the state similarity between wind turbines, establishing a similarity matrix, and combining Gaussian mixture model. Common statistical analysis-based methods are represented by 3σ method and quartile method, which calculate the relevant statistics of the data and then use statistical principles to exclude data outside the statistical range. However, when faced with a large amount of concentrated accumulation abnormal data, the statistics of the statistical analysis-based method are easily affected by these abnormal data, resulting in poor anomaly identification effect.

[0004] Therefore, there is an urgent need to provide a new wind turbine data processing method to improve the above problems. Summary of the invention

[0005] The object of the present invention is to provide a method and device for processing wind turbine data, so as to improve the problem of insufficient abnormal data identification and cleaning of traditional wind turbines.

[0006] In a first aspect, an embodiment of the present invention provides a method for processing wind turbine data, the method comprising: The wind speed data is divided into multiple intervals according to a set step size, and the wind turbine operation data corresponding to each interval is obtained; the actual gradient change amplitude of the wind turbine operation data in each interval is calculated; when the actual gradient change amplitude of the wind turbine operation data is lower than the reference gradient change amplitude of the set power curve, the wind turbine operation data is marked as abnormal and eliminated; based on the 3σ criterion, the mean value and standard deviation corresponding to the wind turbine operation data in each interval are calculated; the confidence interval of each interval is determined based on the mean value and standard deviation; according to the confidence interval, the abnormal data in the wind turbine operation data in each interval is eliminated.

[0007] In a possible implementation scheme, according to the confidence interval, abnormal data in the wind turbine operation data within each interval is eliminated, including: according to the confidence interval, determining the neighborhood radius and the minimum number of neighborhood samples for each interval; based on the neighborhood radius and the minimum number of neighborhood samples, performing DBSCAN clustering on the wind turbine operation data to identify and clean abnormal data.

[0008] In another possible implementation, based on the neighborhood radius and the minimum number of neighborhood samples, DBSCAN clustering is performed on the wind turbine operation data to identify and clean abnormal data, including: For any wind turbine operating data: calculate the number of data that conform to a specific direction within the corresponding neighborhood radius, and compare it with the minimum number of neighborhood samples. If the number of data is greater than the minimum threshold, mark the data as normal, otherwise, mark it as abnormal; wherein the minimum number of neighborhood samples is the number of data within the interval whose angle between the direction of data change and the cluster propagation direction is less than the threshold.

[0009] In a possible implementation manner, the step length is used to control the width of the wind speed range. The larger the step length, the more relevant data there are, and the smaller the step length, the less relevant data there are.

[0010] In another possible implementation, the reference gradient variation amplitude and the actual gradient variation amplitude are represented by cosine similarity.

[0011] In a possible implementation manner, the wind turbine operating data includes one or more of the following information: wind speed, active power, generator speed, generator torque, and pitch angle.

[0012] In a second aspect, the present invention further provides a wind turbine data processing device, comprising a memory and a processor, wherein the memory is used to store programs, and the processor is used to call the programs in the memory to execute the method embodiments as described in any one of the first aspects.

[0013] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a computer program. When the computer program runs on a processor, the processor executes the method executed in the first aspect.

[0014] In a fourth aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory. The memory is used to store one or more computer programs; when the one or more computer programs stored in the memory are executed by the processor, the processor can implement the method of any possible embodiment of the first aspect.

[0015] In a fifth aspect, an embodiment of the present invention further provides a computer program product, which, when executed on a processor, enables the processor to execute a method of any possible embodiment of any of the above aspects.

[0016] The embodiments of the present invention provide a method and device for processing wind turbine data, which have the following beneficial effects: 1. Effectively identify and clean abnormal data: The wind turbine data processing method proposed in the present invention can accurately identify and clean abnormal data in wind turbine operation data. The method divides wind speed data into multiple intervals, and according to the data change characteristics in each interval, combines the 3σ criterion and the DBSCAN clustering algorithm to achieve accurate identification and cleaning of abnormal data. This not only improves data quality, but also provides more reliable data support for wind turbine control and prediction research.

[0017] 2. Improve the accuracy of abnormal data identification: In view of the problem that the traditional abnormal data identification method based on statistical analysis is not effective in processing a large amount of accumulated abnormal data, the present invention effectively improves the accuracy of abnormal data identification by introducing the comparison between the actual gradient change amplitude and the reference gradient change amplitude, and the abnormal data identification strategy based on DBSCAN clustering. In particular, when processing complex and changeable wind turbine operation data, the method of the present invention shows higher adaptability and robustness.

[0018] 3. Enhance data processing efficiency: The wind turbine data processing method of the present invention also focuses on improving data processing efficiency. By reasonably dividing the wind speed data intervals and identifying and cleaning abnormal data based on the data characteristics within the intervals, the computational bottleneck that may occur when processing large amounts of data with traditional methods is avoided. At the same time, the DBSCAN clustering algorithm is used to identify abnormal data, which further improves the speed and efficiency of data processing.

[0019] In summary, the wind turbine data processing method and device of the present invention have achieved remarkable results in improving the problems of insufficient traditional abnormal data identification and cleaning, and have broad application prospects and important promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 A schematic flow chart of a wind turbine data processing method provided by an embodiment of the present invention; Figure 2 A schematic diagram of a power-wind speed characteristic curve of a wind turbine provided by an embodiment of the present invention; Figure 3 A schematic diagram of a clustering method of a DBSCAN algorithm provided in an embodiment of the present invention; Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In the description of the embodiments of the present invention, the terms used in the following embodiments are only for the purpose of describing specific embodiments, and are not intended to be used as limitations on the present invention. As used in the specification of the present invention and the appended claims, the singular expressions "a", "said", "above", "the" and "this" are intended to also include expressions such as "one or more", unless there is a clear contrary indication in the context. It should also be understood that in the following embodiments of the present invention, "at least one", "one or more" refer to one or more (including two). The term "and / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist; for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.

[0023] References to "one embodiment" or "some embodiments" etc. described in this specification mean that one or more embodiments of the present invention include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways. The term "connection" includes direct connection and indirect connection, unless otherwise specified. "First" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.

[0024] In the embodiments of the present invention, words such as "exemplarily" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or implementation described as "exemplarily" or "for example" in the embodiments of the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or implementations. Specifically, the use of words such as "exemplarily" or "for example" is intended to present related concepts in a specific way.

[0025] In view of the problems existing in the prior art, the present invention provides a method for processing wind turbine data. The method introduces the power-wind speed characteristic curve of the wind turbine to determine the direction of cluster expansion, and combines the 3σ criterion to adaptively adjust the key parameters of the DBSCAN clustering algorithm, thereby realizing automatic identification and cleaning of abnormal data.

[0026] like Figure 1 As shown, an embodiment of the present invention provides a method for processing wind turbine data. The process of the method can be executed by an electronic device. The method includes the following steps: S101, dividing the wind speed data into multiple intervals according to a set step size, and obtaining wind turbine operation data corresponding to each interval; S102, calculating the actual gradient change amplitude of the wind turbine operating data in each interval; S103, when the actual gradient change amplitude of the operating data of a certain wind turbine generator set is lower than the reference gradient change amplitude of the set power curve, the operating data of the certain wind turbine generator set is marked as abnormal and removed; S104, calculating the mean value and standard deviation of the wind turbine operating data in each interval based on the 3σ criterion; S105, determining a confidence interval for each interval based on the mean value and the standard deviation; S106: Eliminate abnormal data in the wind turbine operating data within each interval according to the confidence interval.

[0027] It should be understood that before executing the above method steps, data set preparation is performed first, such as collecting wind turbine operating data, including wind speed, active power, generator speed, generator torque, pitch angle and other data information. In addition, parameter settings are also required, such as setting the step size (Stride) to 0.1 to select relevant data around the wind speed; for example, setting the reference gradient change amplitude (Threshold) to 0.99 to screen the angle between the data gradient change and the power curve gradient within the neighborhood radius.

[0028] In a possible implementation scheme, in the above step S106, the neighborhood radius and the minimum number of neighborhood samples of each interval are determined according to the confidence interval; based on the neighborhood radius (Eps) and the minimum number of neighborhood samples (MinPts), DBSCAN clustering is performed on the wind turbine operation data to identify and clean abnormal data.

[0029] Specifically, in a possible embodiment, based on the neighborhood radius and the minimum number of neighborhood samples, DBSCAN clustering is performed on the wind turbine operation data to identify and clean abnormal data, including: exemplary, such as Figure 3 As shown, for any wind turbine operating data: calculate the number of data that conform to a specific direction within the corresponding neighborhood radius, and compare it with the minimum neighborhood sample number; if the number of data is greater than the minimum threshold, mark the data as normal, otherwise, mark it as abnormal; wherein the minimum neighborhood sample number is the number of data within the interval whose angle between the direction of data change and the cluster propagation direction is less than a threshold.

[0030] It is worth noting that if Figure 2 The wind turbine power-wind speed characteristic curve shown is one of the core indicators of wind farm operation and performance evaluation, which can intuitively describe the relationship between wind speed and power generation. This curve is not only of great significance for evaluating the power generation performance of wind turbines, but also a key basis for monitoring and diagnosing data anomalies in the SCADA system.

[0031] According to the physical characteristics of wind turbines, the relationship between output power P and wind speed v can be expressed as: ; Where P represents the output power of the wind turbine; V, V i , V r , V o are actual, cut-in, rated, and cut-out wind speeds respectively; P r is the rated output power, P(v) is the dynamic output power generated by the unit when the wind speed is between the cut-in wind speed and the rated wind speed; The wind energy utilization coefficient is closely related to the pitch angle and tip speed ratio of the unit, and it reflects the ability of the wind turbine blades to absorb and convert natural wind energy into electrical energy. According to the Betz limit theory, its maximum value does not exceed 0.593, while in actual operation, the wind energy utilization coefficient of the unit is usually between 0.3 and 0.5; is the air density (kg / m3); R is the impeller radius (m).

[0032] In this embodiment, the 3σ criterion, also known as the Laida criterion, is a data processing method based on the normal distribution assumption, which is used to identify and eliminate outliers. This criterion calculates the standard deviation and determines a confidence interval. Values ​​outside this interval are considered outliers and are eliminated. The specific calculation steps are as follows: a. Determine the sample data set ; b, calculate the sample mean ; c, calculate the sample standard deviation ; d, determine the confidence interval: ; e. Eliminate outliers: Delete the numbers in the data set X that are outside the confidence interval to obtain standardized data.

[0033] In other possible embodiments, the power-wind speed characteristics can also be dynamically adjusted through machine learning algorithms, such as neural networks or support vector machines. In addition, a real-time data stream processing algorithm can be developed to quickly respond to wind speed changes, adjust clustering parameters in real time, and realize instant identification and cleaning of abnormal data. Optionally, a mechanism is designed to dynamically adjust the neighborhood radius (Eps) and the minimum number of neighborhood samples (MinPts) parameters in the DBSCAN clustering algorithm based on real-time feedback of wind turbine operation data to adapt to data characteristics in different wind speed ranges.

[0034] Furthermore, historical data and time series analysis can be used to establish a prediction model to predict possible abnormal data, and preventive measures can be taken to reduce the impact of abnormal data on wind turbine performance evaluation. Alternatively, a comprehensive evaluation system can be built to not only clean the data, but also score the data quality, providing a more reliable basis for subsequent data analysis and decision-making. Intelligent diagnosis and feedback loop: Integrate abnormal data identification with the diagnosis system. When abnormal data is identified, the system can automatically diagnose the possible causes and feedback the diagnosis results to maintenance personnel, forming a closed-loop intelligent diagnosis system. Data processing platform based on cloud computing: Utilize the powerful computing power and storage resources of cloud computing to establish a centralized data processing platform to achieve efficient processing and analysis of large-scale wind turbine data. It can also be combined with AR technology to provide maintenance personnel with real-time fault diagnosis information and operation guidance to improve maintenance efficiency and accuracy.

[0035] In this embodiment, the DBSCAN algorithm performs clustering through two key parameters, namely, the neighborhood radius (Eps) and the minimum number of samples (MinPts). The steps of the DBSCAN algorithm are as follows: (1) Select an unvisited data point p as the starting point, mark it as visited, and add all its density-reachable neighbor points to a set.

[0036] (2) For each data point q in the set, calculate the density of all points in its neighborhood and add the points whose density is greater than or equal to the threshold to the set. If q is a core point, add all points in its neighborhood to the set. If q is a boundary point, only add the points whose density is greater than or equal to the threshold in its neighborhood.

[0037] (3) Repeat step (2) until all data points in the set have been visited.

[0038] (4) For each unvisited data point, if it is not included in any set, it is marked as a noise point. Otherwise, it is assigned to the same cluster as the set it belongs to.

[0039] The embodiments of the present invention provide a method and device for processing wind turbine data, which have the following beneficial effects: 1. Effectively identify and clean abnormal data: The wind turbine data processing method proposed in the present invention can accurately identify and clean abnormal data in wind turbine operation data. The method divides wind speed data into multiple intervals, and according to the data change characteristics in each interval, combines the 3σ criterion and the DBSCAN clustering algorithm to achieve accurate identification and cleaning of abnormal data. This not only improves data quality, but also provides more reliable data support for wind turbine control and prediction research.

[0040] 2. Improve the accuracy of abnormal data identification: In view of the problem that the traditional abnormal data identification method based on statistical analysis is not effective in processing a large amount of accumulated abnormal data, the present invention effectively improves the accuracy of abnormal data identification by introducing the comparison between the actual gradient change amplitude and the reference gradient change amplitude, and the abnormal data identification strategy based on DBSCAN clustering. In particular, when processing complex and changeable wind turbine operation data, the method of the present invention shows higher adaptability and robustness.

[0041] 3. Enhance data processing efficiency: The wind turbine data processing method of the present invention also focuses on improving data processing efficiency. By reasonably dividing the wind speed data intervals and identifying and cleaning abnormal data based on the data characteristics within the intervals, the computational bottleneck that may occur when processing large amounts of data with traditional methods is avoided. At the same time, the DBSCAN clustering algorithm is used to identify abnormal data, which further improves the speed and efficiency of data processing.

[0042] All relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.

[0043] The above system may be executed by a chip or a chip module. The various modules / units included in the various devices and products described in the above embodiments may be software modules / units or hardware modules / units, or may be partially software modules / units and partially hardware modules / units.

[0044] In some other embodiments of the present invention, the present invention discloses an electronic device, such as Figure 4 As shown, the electronic device can integrate the above system, and the hardware composition includes: one or more processors 401; memory 402; display 403; one or more application programs (not shown); and one or more computer programs 404, and the above components can be connected through one or more communication buses 405. The one or more computer programs 404 are stored in the above memory 402 and are configured to be executed by the one or more processors 401, and the one or more computer programs 404 include instructions.

[0045] The present invention also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a computer, the method described in the above method embodiment is implemented. Specific beneficial effects can be found in the above method embodiment.

[0046] The present invention also provides a computer program product, which implements the method described in the above method embodiment when executed by a computer. Specific beneficial effects can be found in the above method embodiment.

[0047] Through the description of the above implementation methods, technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0048] Each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0049] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: flash memory, mobile hard disk, read-only memory, random access memory, disk or optical disk and other media that can store program codes.

[0050] The above is only a specific implementation of the embodiment of the present invention, but the protection scope of the embodiment of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiment of the present invention should be included in the protection scope of the embodiment of the present invention. Therefore, the protection scope of the embodiment of the present invention should be based on the protection scope of the claims.

Claims

1. A method for processing wind turbine data, characterized in that: The following steps are involved: Divide the wind speed data into multiple intervals according to the set step size, and obtain the wind turbine operation data corresponding to each interval; Calculate the actual gradient change amplitude of the wind turbine operating data in each interval; When the actual gradient change amplitude of the operating data of a certain wind turbine generator set is lower than the reference gradient change amplitude of the set power curve, the operating data of the certain wind turbine generator set is marked as abnormal and removed; Based on the 3σ criterion, the mean value and standard deviation of the wind turbine operating data in each interval are calculated; determining a confidence interval for each interval based on the mean and standard deviation; According to the confidence interval, abnormal data in the wind turbine operating data within each interval is eliminated.

2. The method according to claim 1, characterized in that: According to the confidence interval, the abnormal data in the wind turbine operation data within each interval is eliminated, including: According to the confidence interval, determining the neighborhood radius and the minimum number of neighborhood samples for each interval; Based on the neighborhood radius and the minimum number of neighborhood samples, DBSCAN clustering is performed on the wind turbine operating data to identify and clean abnormal data.

3. The method according to claim 2, characterized in that Based on the neighborhood radius and the minimum number of neighborhood samples, DBSCAN clustering is performed on the wind turbine operation data to identify and clean abnormal data, including: For any wind turbine operating data: calculate the number of data that conform to a specific direction within the corresponding neighborhood radius, and compare it with the minimum number of neighborhood samples. If the number of data is greater than the minimum threshold, mark the data as normal, otherwise, mark it as abnormal; wherein the minimum number of neighborhood samples is the number of data within the interval whose angle between the direction of data change and the cluster propagation direction is less than the threshold.

4. The method according to any one of claims 1 to 3, characterized in that: The step length is used to control the width of the wind speed range. The larger the step length, the more relevant data there are, and the smaller the step length, the less relevant data there are.

5. The method according to any one of claims 1 to 3, characterized in that: The reference gradient change amplitude and the actual gradient change amplitude are represented by cosine similarity.

6. The method according to any one of claims 1 to 3, characterized in that: The wind turbine operating data includes one or more of the following information: wind speed, active power, generator speed, generator torque, and pitch angle.

7. A wind turbine data processing device, characterized in that: The invention comprises a memory and a processor, wherein the memory is used to store a program, and the processor is used to call the program in the memory to execute the method according to any one of claims 1 to 6.

8. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein the memory stores a computer program executable on the processor, and when the computer program is executed by the processor, the electronic device executes the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored therein, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The method comprises a program which causes a computer to execute the method according to any one of claims 1 to 6.

Images