Method and system for the rapid evaluation of generator vibration data from a wind turbine

Abstract

The present application provides a method and a system for the rapid evaluation of generator vibration data from a wind turbine, comprising the following steps: Step 1: Acquiring the daily operating data of the wind turbine under test, wherein the daily operating data includes timestamps, operating power, and generator vibration signals; Step 2: Creating a two-dimensional scatter plot based on the daily operating data; Step 3: Combining the created two-dimensional scatter plot with a predefined regression function to evaluate the vibration state of the wind turbine under test; the present application uses a distortion-free compression method for data processing and thus enables an accurate assessment of the generator vibration state for the current day.

Description

A method and system for quickly evaluating vibration data of wind turbine generators

[0001] This application claims priority to a Chinese patent application filed with the Patent Office of China on February 26, 2024, with application number 202410210347.2 and invention name “A method and system for rapid evaluation of vibration data of a wind turbine generator”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of new energy wind power generation, and specifically to a method and system for quickly evaluating vibration data of a wind turbine generator. Background Art

[0003] During operation, wind turbines are subject to constantly changing impact loads, making their blades, hubs, and components of the transmission chain rigidly connected to the hub—including the main shaft, gearbox, high-speed shaft, and generator—prone to failure. Statistics show that failures in wind turbine transmission chains cause the longest downtime and have the greatest impact on power production. Failures in the generator, a key rotating mechanical and electrical component in the transmission chain, can lead to serious accidents such as nacelle fires, runaway vehicles, and tower collapses. Therefore, existing approaches rely on real-time monitoring of generator vibration data during wind turbine operation and use vibration monitoring and analysis techniques for fault diagnosis and early warning. These methods often employ methods such as time-domain analysis, frequency-domain analysis, and time-frequency superposition analysis to predict potential wind turbine hazards and future failures, enabling timely action and improving wind turbine reliability. However, existing generator vibration data is sampled at a high frequency, resulting in a massive amount of data. Real-time data analysis is computationally intensive, and offline analysis suffers from limited timeliness. A rapid assessment method for wind turbine generator vibration data is urgently needed to reduce the data analysis workload while ensuring rapid and accurate assessment of generator status.

[0004] Summary of the Invention

[0005] The purpose of this application is to provide a method and system for rapid evaluation of vibration data of wind turbine generators, which solves the defects of existing wind turbine fault warning methods such as large amount of real-time analysis data and poor offline analysis timeliness. This application can quickly process massive high-frequency generator vibration data, detect potential operating status deviations of wind turbine generators in advance, and avoid the adverse consequences caused by faults, namely extreme events.

[0006] In order to achieve the above objectives, the technical solution adopted in this application is:

[0007] The present application provides a method for quickly evaluating vibration data of a wind turbine generator, comprising the following steps:

[0008] Step 1: Acquire the current day operation data of the wind turbine to be tested, wherein the current day operation data includes a time tag, operating power, and generator vibration signal;

[0009] Step 2: construct a two-dimensional plane scatter plot based on the running data of the day;

[0010] Step 3: Combining the obtained two-dimensional plane scatter plot with a preset regression function to evaluate the vibration state of the wind turbine to be tested;

[0011] Among them, the preset regression function construction method is:

[0012] S31, obtaining the operating data of the wind turbine to be tested in the previous year to form a matrix, wherein the operating data includes a time tag, operating power, and generator vibration signal;

[0013] S32, dividing the matrix according to the time label to obtain multiple sub-matrices;

[0014] S33, performing non-distortion compression on the data of each sub-matrix to obtain non-distortion compressed data corresponding to each sub-matrix;

[0015] S34, performing feature extraction on the non-distorted compressed data to obtain a preset regression function.

[0016] Optionally, in step 2, a two-dimensional plane scatter plot is constructed based on the current day's operating data. The specific method is:

[0017] The operating power in the operating data of the day is used as the x-axis, and the generator vibration signal in the operating data of the day is used as the y-axis to construct a two-dimensional plane scatter plot.

[0018] Optionally, in S32, the matrix is ​​divided according to the time label to obtain multiple sub-matrices, and the specific method is:

[0019] The data with the same value in the corresponding month in the time label are grouped together to obtain multiple sub-matrices.

[0020] Optionally, before performing non-distortion compression on the data of each sub-matrix, the data of each sub-matrix is ​​preprocessed to obtain a preprocessed sub-matrix.

[0021] Optionally, the specific method of preprocessing the data of each sub-matrix is:

[0022] Delete the rows corresponding to the operating power less than or equal to zero in each sub-matrix to obtain the removed sub-matrix;

[0023] The operating power of each removed sub-matrix is ​​rearranged in ascending order, and divided into multiple small sub-matrices according to the preset power window;

[0024] Remove the meaningless data points in each small sub-matrix to obtain the processed small sub-matrix;

[0025] All processed small sub-matrices are combined and rearranged in the order of time labels to obtain the pre-processed sub-matrix.

[0026] Optionally, in S33, the data of each sub-matrix is ​​non-distortedly compressed using a Midimax compression algorithm to obtain non-distorted compressed data corresponding to each sub-matrix.

[0027] Optionally, in S34, feature extraction is performed on the non-distorted compressed data, specifically by:

[0028] The obtained non-distorted compressed data is sorted and reconstructed in ascending order of operating power to obtain reconstructed data;

[0029] The operating power in the reconstructed data is taken as the x-axis, and the generator vibration signal in the reconstructed data is taken as the y-axis to construct a two-dimensional plane scatter plot;

[0030] The K-means clustering method is used to identify the upper boundary points in the two-dimensional plane scatter plot and obtain a point set;

[0031] The nonlinear regression method is used to construct the regression function corresponding to the point set.

[0032] Optionally, in step 3, the operating status of the wind turbine to be measured is evaluated, specifically by:

[0033] The K-means clustering method is used to identify the upper boundary points of the two-dimensional plane scatter plot and obtain a point set;

[0034] If there are ω1 points in the point set that are located on the upper side of the regression function, then the vibration of the generator set under test is abnormal on that day; this is recorded as class2 (i.e., the second state classification);

[0035] If there are ω2 points in the point set located on the upper side of the regression function, and there are no ω1 points located on the upper side of the regression function, then the vibration of the generator set under test on that day has a slight state deviation; this is recorded as class 1 (i.e., the first state classification);

[0036] If the number of points on the upper side of the regression function in the point set is less than ω2, the vibration of the generator set under test is normal on that day; it is recorded as class0 (i.e., the 0th state classification);

[0037] Among them, ω1 is the dynamic anomaly threshold; ω2 is the dynamic offset threshold; r is the number of point sets, and:

[0038] ω2<ω1 <r。

[0039] Step 4: Record the results obtained from the evaluation of the data that has undergone lossless compression on the same day in Step 3, and assign values to each status classification. Assign 2 to class2 (i.e., the second status classification), 1 to class1 (i.e., the first status classification), and 0 to class0 (i.e., the 0th status classification).

[0040] Calculate the cumulative Score of the status assignments for the previous 30 days every 30 days.

[0041] If Score ≤ 10, do not take any measures for the generator of the unit.

[0042] If 10 < Score ≤ 20, arrange for inspections at an appropriate time using the opportunity maintenance method.

[0043] If 20 < Score ≤ 30, arrange for inspections as soon as possible.

[0044] If Score > 30, the generator has shown an obvious long-term status deviation, and arrange for maintenance.

[0045] By combining the vibration data with a preset regression function, evaluate it every day, synthesize the evaluation results every 30 days, and take corresponding inspection and maintenance measures according to different comprehensive score values; identify the status deviation of the generator to the greatest extent and ensure the safety of the unit.

[0046] A fast evaluation system for the vibration data of a wind turbine generator, comprising:

[0047] The same-day data acquisition unit is used to acquire the operating data of the wind turbine to be measured on the same day. The same-day operating data includes a time tag, operating power, and a generator vibration signal.

[0048] The scatter plot construction unit is used to construct a two-dimensional plane scatter plot based on the same-day operating data.

[0049] The vibration status evaluation unit is used to combine the obtained two-dimensional plane scatter plot with a preset regression function to evaluate the vibration status of the wind turbine to be measured.

[0050] Compared with the prior art, the beneficial effects of this application are:

[0051] A fast evaluation method for the vibration data of a wind turbine generator provided by this application processes the operating data of the wind turbine to be measured in the previous year using lossless compression data. While retaining the data characteristics, it can be fast and have a small amount of calculation, without introducing non-actual data, and only return a subset of the original data. The feature extraction method used can update the regression function every day. By combining the regression function with the operating data of the generator set on the same day, the vibration status of the generator set on the same day can be obtained, achieving accurate evaluation of generator vibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is the overall technical diagram;

[0053] Figure 2 is a flow chart of the data acquisition and preprocessing method and system;

[0054] FIG3 is a flow chart of a method and system for data lossless compression;

[0055] FIG4 is a flow chart of the feature extraction method and system;

[0056] Figure 5 is a flowchart of the evaluation method and system. DETAILED DESCRIPTION

[0057] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0058] Example 1

[0059] As shown in Figures 1 to 5, this embodiment provides a method and system for rapidly evaluating wind turbine generator vibration data, which includes data acquisition and preprocessing, data non-distortion compression, feature extraction, and final evaluation. Specifically, the following steps are included:

[0060] Step 1: This application uses wind turbine #1 in the target wind farm as an example. The data acquired includes the operating time tag T, operating power P, and vibration signal V collected by the generator vibration acceleration sensor of wind turbine #1 in the previous year as of the day to be analyzed. The sampling frequency here is 10kHz, which can form the data matrix A:

[0061] Where A is a matrix with n rows and 3 columns; n represents the number of data collection points of wind turbine #1 in the previous year; T1, T2, ..., T i ,…,T n It is the time tag in the collected data, and its format is 20XX-MM-DD:HH-FF-SS-UUUUUU, where XX represents the "year" in the time tag, MM corresponds to "month", DD corresponds to "day", HH corresponds to "hour", FF corresponds to "minute", SS corresponds to "second", and UUUUUU corresponds to "ten microseconds"; P1, P2, ..., P i ,…,P n is the operating power of the corresponding time tag; V1, V2, ..., V i ,…,V nThe vibration signal collected by the generator vibration acceleration sensor corresponds to the time tag.

[0062] Step 2: Divide matrix A into twelve sub-matrices as follows:

[0063] Divide the matrix A according to the time tags in the collected data, where the time tags T corresponding to the same value of "month" are divided into the same matrix, and then we can get:

[0064] Among them, A1 is the operating data of wind turbine #1 obtained in the same month of the previous year; A2 is the operating data of wind turbine #1 obtained in the next month of the previous year, and so on. 12 These are the operating data of wind turbine #1 obtained in the last month of this year.

[0065] Step 3: A1, A2, ..., A 12 Perform data preprocessing. Here we take the A1 matrix as an example. The specific method is as follows:

[0066] First, remove the rows corresponding to the power data in the second column of the A1 matrix that is less than or equal to 0 to form the removal sub-matrix A1';

[0067] Secondly, rearrange the operating power of the second column of the A1' matrix from small to large, and further divide it into 10 small sub-matrices: A'1 = [A' 11 ,A′ 12 …A 110 ']

[0068] Among them, A' 11 The data obtained when the power of unit #1 is 0-10% of the rated power; A' 12 The data obtained when the power of unit #1 is 10-20% of the rated power; and so on, A' 110 The data obtained when the power of unit #1 is 90-100% of the rated power.

[0069] The 6σ principle is used to calculate A' 11 ,A' 12 …A 110 'Process the data, remove meaningless points, and use A' 11 For example, the specific method is as follows:

[0070] Calculate A' 11 The mathematical expectation ε' of the third column of vibration data 11 and variance σ' 11 ;

[0071] By A' 11 The first element V' in the third column11 Start judging.

[0072] If V' 11 >ε' 11 +6×σ' 11 or V' 11 <ε' 11 -6×σ' 11 , then eliminate V' 11 Corresponding row data;

[0073] In other cases, retain the row data;

[0074] Then A' 11 After processing, we get the matrix A″ 11 ;

[0075] Similarly, we can get A″ processed by the 6σ principle 12 …A 110 ″;

[0076] Finally, for A″ 11 ,A″ 12 …A 110 ”, in order to maximize the data authenticity and subsequent compression, no data removal is performed, and the ten matrices A” 11 ,A″ 12 …A 110 ″ are combined in sequence and re-arranged according to the order of the time labels in the first column to obtain the sub-matrix A″1 after preprocessing.

[0077] Similarly, A2,…,A 12 Using the above method to process, we finally get A″2,…A″ 12 .

[0078] Then we have:

[0079] Among them, A” is the pre-processed operating data of wind turbine #1 in the previous year, A″1, A″2,…A″ 12 They are the pre-processed operating data of the #1 wind turbine in the previous year from this month to the previous January.

[0080] Step 4: The sampling frequency of the pre-processed #1 wind turbine operating data from the previous year in step 3 is 10kHz. The data volume is huge, and direct analysis and calculation is computationally intensive. Therefore, data non-distortion compression is required. The specific method is as follows:

[0081] Here, we take the non-distortion compression of the operating data A″1 of wind turbine #1 in the previous year as an example. The operating data A″1 is sorted in the order of time tags, which is a general high-frequency acquisition time domain sequence. According to steps 1-3, A″1 is an m-row 3-column matrix, where the first column is the time tag, the second column is the operating power under the corresponding time tag, and the third column is the vibration signal collected by the vibration acceleration sensor under the corresponding time tag. In addition, m≤100000×3600×24×31=2.6784×10 11

[0082] The Midimax compression algorithm is used to perform lossless compression on A″1. The specific process is as follows:

[0083] S41, input time series data, i.e. A″1 and compression coefficient λ, to the Midimax algorithm. In this example, the compression coefficient λ is much smaller than m, i.e. λ<<m

[0084] S42, split the time series data A″1 into λ non-overlapping windows of equal size, where the size L of each non-overlapping window is calculated as follows:

[0085] Then the original A″1 is divided into λ non-overlapping window corresponding matrices, that is

[0086] S43, sort the matrices corresponding to the λ non-overlapping windows in S42 in ascending order according to the operating power (i.e., the second column of the corresponding matrix), and obtain B1, B2, ..., B λ ;

[0087] S44, for B1, B2…, B λ , here taking B1 as an example, select the rows corresponding to the largest k elements in the third column of each matrix (i.e., the vibration signal collected by the generator vibration acceleration sensor), the rows corresponding to the smallest k elements, and the rows corresponding to the k elements closest to e1 to form the matrix C1, where e1 represents the mathematical expectation of the third column in the matrix B1 (i.e., the vibration signal collected by the generator vibration acceleration sensor). Similarly, C2, C3…C λ , then C1, C2…C λ Both are 3k rows and 3 columns matrices;

[0088] S45, the matrices C1, C2...C obtained in S44 are λ The matrix D1 is formed by reordering the second column of operating power from small to large. D1 is a 3×k×λ row and 3 column matrix. The original m row and 3 column matrix A″1 is compressed into a 3×k×λ row and 3 column matrix D1.

[0089] Similarly, we can complete A″2, A″3, …A″ 12The lossless compression finally gets D2, D3…D 12 .

[0090] Step 5: Extract features from the non-distorted compressed data

[0091] S51, D1, D2...D 12 The matrix D is a 36×k×λ matrix with 3 rows and columns, where the first column is the time tag, the second column is the operating power under the corresponding time tag, and the third column is the vibration signal collected by the vibration acceleration sensor under the corresponding time tag.

[0092] S52, reordering the matrix D according to the second column of the operating power from small to large to form a matrix O;

[0093] S53, constructing a two-dimensional plane scatter plot using the matrix O in S52, where the x-axis is the operating power in the second column of the matrix O, and the y-axis is the vibration signal collected by the vibration acceleration sensor in the third column of the matrix O;

[0094] S54, use K-means clustering method to identify the upper boundary points in the scatter plot in S53 and form a point set. There are v upper boundary points in this set, which are recorded as edge O =[dot O1 ,dot O2 ,…,dot Ov ]: dot O1 =(x O1 ,y O1 ) dot O2 =(x O2 ,y O2 ) dot Ov =(x Ov ,y Ov )

[0095] Among them, dot O1 ,dot O2 ,…,dot Ov are the coordinates of each point;

[0096] S55, using nonlinear regression method, constructs the upper boundary point set edge in S54 O The regression function f O , regression function f O The independent variable is the operating power, and the dependent variable is the vibration signal collected by the vibration acceleration sensor.

[0097] Step 6: Evaluate the wind turbine to be tested based on the features extracted in step 5. Take the wind turbine #1 in the target wind farm as an example.

[0098] S61, every night at 00:00, extract the operating time tag T, operating power P, and vibration signal V collected by the generator vibration acceleration sensor of the day (00:00 to 23:59), where the sampling frequency is 10kHz, to form a data matrix H';

[0099] S62, removing the rows corresponding to the power data in the second column of the H' matrix that are less than or equal to 0, to form the matrix H;

[0100] S63, constructing a two-dimensional plane scatter plot using the matrix H in S62, where the x-axis is the operating power in the second column of the matrix H, and the y-axis is the vibration signal collected by the generator vibration acceleration sensor in the third column of the matrix H;

[0101] S64, use K-means clustering method to identify the upper boundary points in the scatter plot in S63 and form a point set. There are a total of r upper boundary points in this set, which are recorded as edge H =[dot H1 ,dot H2 ,…,dot Hr ], dot H1 =(x H1 ,y H1 ) dot H2 =(x H2 ,y H2 ) dot Hr =(x Hr ,y Hr )

[0102] Among them, dot H1 ,dot H2 ,…,dot Hr are the coordinates of each point;

[0103] S65, analyzing and judging the data obtained in S64,

[0104] If dot H1 ,dot H2 ,…,dot Hr If there are ω1 points located on the upper side of the regression function in S55, it is estimated that the generator vibration on that day is abnormal;

[0105] If dot H1 ,dot H2 ,…,dot Hr If there are ω2 points located on the upper side of the regression function in S55, and there is no ω1 point located on the upper side of the regression function in S55, it is assessed that there is a slight state deviation in the generator vibration on that day;

[0106] If dot H1 ,dot H2 ,…,dotHr If the number of points on the upper side of the regression function in S55 is less than ω2, it is assessed that the generator vibration is normal on that day;

[0107] Where ω1 and ω2 are the dynamic abnormality threshold and dynamic offset threshold respectively, and: ω2<ω1 <r。

[0108] Example 2

[0109] Based on Example 1, this embodiment provides a system for rapidly evaluating vibration data of a wind turbine generator, including:

[0110] The current day data acquisition unit is used to acquire the current day operation data of the wind turbine to be tested, wherein the current day operation data includes a time tag, operating power and generator vibration signal;

[0111] A scatter plot construction unit is used to construct a two-dimensional plane scatter plot based on the current day's operating data;

[0112] The vibration state evaluation unit is used to combine the obtained two-dimensional plane scatter plot with a preset regression function to evaluate the vibration state of the wind turbine to be tested.

[0113] This application uses a non-distortion compression method to process data. It is fast and computationally light while retaining data features. It does not introduce non-actual data and only returns a subset of the original data. The feature extraction method used can update the regression curve every day to achieve accurate assessment of generator vibration.

[0114] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for rapid evaluation of wind turbine generator vibration data, characterized in that: The following steps are involved: Step 1: Acquire the current day operation data of the wind turbine to be tested, wherein the current day operation data includes a time tag, operating power, and generator vibration signal; Step 2: construct a two-dimensional plane scatter plot based on the running data of the day; Step 3: Combining the obtained two-dimensional plane scatter plot with a preset regression function to evaluate the vibration state of the wind turbine to be tested; Among them, the preset regression function construction method is: S31, obtaining the operating data of the wind turbine to be tested in the previous year to form a matrix, wherein the operating data includes a time tag, operating power, and generator vibration signal; S32, dividing the matrix according to the time label to obtain multiple sub-matrices; S33, performing non-distortion compression on the data of each sub-matrix to obtain non-distortion compressed data corresponding to each sub-matrix; S34, performing feature extraction on the non-distorted compressed data to obtain a preset regression function.

2. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: In step 2, a two-dimensional scatter plot is constructed based on the current day's operating data. The specific method is: The operating power in the operating data of the day is used as the x-axis, and the generator vibration signal in the operating data of the day is used as the y-axis to construct a two-dimensional plane scatter plot.

3. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: In S32, the matrix is ​​divided according to the time label to obtain multiple sub-matrices. The specific method is: The data with the same value in the corresponding month in the time label are grouped together to obtain multiple sub-matrices.

4. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: Before performing non-distortion compression on the data of each sub-matrix, the data of each sub-matrix is ​​preprocessed to obtain a preprocessed sub-matrix.

5. A method for rapid evaluation of wind turbine generator vibration data according to claim 4, characterized in that: The specific method for preprocessing the data of each sub-matrix is: Delete the rows corresponding to the operating power less than or equal to zero in each sub-matrix to obtain the removed sub-matrix; The operating power of each removed sub-matrix is ​​rearranged in ascending order, and divided into multiple small sub-matrices according to the preset power window; Remove the meaningless data points in each small sub-matrix to obtain the processed small sub-matrix; All processed small sub-matrices are combined and rearranged in the order of time labels to obtain the pre-processed sub-matrix.

6. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: In S33, the data of each sub-matrix is ​​non-distortedly compressed using a Midimax compression algorithm to obtain non-distorted compressed data corresponding to each sub-matrix.

7. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: In S34, feature extraction is performed on the non-distorted compressed data. The specific method is: The obtained non-distorted compressed data is sorted and reconstructed in order of operating power from small to large, and the result is to reconstruct the data; The operating power in the reconstructed data is taken as the x-axis, and the generator vibration signal in the reconstructed data is taken as the y-axis to construct a two-dimensional plane scatter plot; The K-means clustering method is used to identify the upper boundary points in the two-dimensional plane scatter plot and obtain a point set; The nonlinear regression method is used to construct the regression function corresponding to the point set.

8. A method for rapid evaluation of wind turbine generator vibration data according to claim 1, characterized in that: In step 3, the operating status of the wind turbine to be tested is evaluated. The specific method is: The K-means clustering method is used to identify the upper boundary points of the two-dimensional plane scatter plot and obtain a point set; If there are ω1 points in the point set located on the upper side of the regression function, then the vibration of the generator set under test is abnormal on that day; If there are ω2 points in the point set located on the upper side of the regression function, and there are no ω1 points located on the upper side of the regression function, then the vibration of the generator set under test on that day has a slight state deviation; If the number of points on the upper side of the regression function in the point set is less than ω2, the vibration of the generator set under test is normal on that day; Among them, ω1 is the dynamic anomaly threshold; ω2 is the dynamic offset threshold; r is the number of point sets, and: ω2<ω1 <r。 9. A wind turbine generator vibration data rapid assessment system, characterized in that: include: The current day data acquisition unit is used to acquire the current day operation data of the wind turbine to be tested, wherein the current day operation data includes a time tag, operating power and generator vibration signal; A scatter plot construction unit is used to construct a two-dimensional plane scatter plot based on the current day's operating data; The vibration state evaluation unit is used to combine the obtained two-dimensional plane scatter plot with a preset regression function to evaluate the vibration state of the wind turbine to be tested.