Fatigue damage real-time evaluation method and system for wind power plant

Abstract

The invention discloses a fatigue damage real-time evaluation method and system for a wind power plant, and the method comprises the steps: obtaining the load data of a current moment, and the load data comprises a load value, a load extreme value, the initial position of a sliding window, and an original accumulated fatigue damage value; updating a load extreme value according to the load data, and judging whether the ending position of the sliding window is changed or not; if so, processing the load data in the sliding window by using a rain flow counting method, and extracting a load amplitude, a mean value and a cycle index; and correcting the load amplitude, determining the maximum cyclic load frequency, and combining with the original accumulated fatigue damage value to obtain the accumulated fatigue damage at the current moment. According to the method, the dynamic sliding window method and the rain flow counting method are combined, the problem that real-time calculation cannot be achieved through a traditional rain flow counting method is solved, and on the basis of guaranteeing precision, rapid real-time calculation of accumulated fatigue damage of the draught fan can be achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind farm optimization, and in particular relates to a fatigue damage real-time assessment method and system for a wind farm. Background Art

[0002] The wind power industry is in a stage of rapid development. Large-scale wind turbines and large-scale stations are continuously put into operation, and wind power generation technology is becoming increasingly mature. However, the wake effect has become an important factor affecting the efficiency of wind farms. When the incoming wind passes through the upstream wind turbine, part of the kinetic energy is absorbed and converted, resulting in a decrease in the inflow wind speed of the downstream wind turbine, which in turn affects its output power. At the same time, the wind turbine wake intensifies the wind speed fluctuation of the inflow wind, increases the turbulence intensity, and increases the fatigue damage of the load on the wind turbine. In order to increase the overall output power of the wind farm and reduce the average fatigue damage, the wind turbine power can be actively controlled during operation.

[0003] In the existing wind farm active power and fatigue optimization model, there are two main methods for calculating cumulative fatigue damage. The first is an approximate method. Since the traditional rain flow counting method cannot achieve real-time calculation, the existing technology usually uses the load (such as shaft torque) directly as an indicator of fatigue size, or assumes that fatigue is proportional to the wind turbine output and defines the fatigue coefficient as a fatigue indicator. However, the calculation accuracy of these approximate methods is obviously insufficient. The second method is a machine learning method, which uses the fatigue data calculated by the rain flow counting method as a training set and realizes real-time operation optimization by offline training of neural networks. Although this method has certain advantages in theory, it has extremely high requirements for the data set, is highly dependent, and the method itself is too complicated, and there are many limitations in practical applications. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a real-time fatigue damage assessment method and system for wind farms to solve the above problems in the prior art.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for real-time fatigue damage assessment of a wind farm, comprising:

[0006] Obtaining load data at the current moment, wherein the load data includes a load value, a load extreme value, a starting position of a sliding window, and an original accumulated fatigue damage value;

[0007] updating the load extreme value according to the load data, and determining whether the end position of the sliding window changes;

[0008] If there is a change, the load data in the sliding window is processed by the rain flow counting method to extract the load amplitude, mean value and number of cycles;

[0009] The load amplitude is corrected, and the maximum number of cyclic loads is determined, and the accumulated fatigue damage at the current moment is obtained by combining the original accumulated fatigue damage value.

[0010] Preferably, the formula for updating the load extreme value is:

[0011]

[0012] In the formula, x t Expressed as the load at time t, x max and x min Expressed as the load extreme value, and They represent the starting and ending positions of the jth sliding window, respectively. win Expressed as the maximum window length.

[0013] Preferably, performing rain flow counting on the load data in the sliding window comprises:

[0014] Preprocessing the load data in the sliding window to obtain preprocessed load data;

[0015] Shearing and reconstructing the preprocessed load data to obtain a load time series;

[0016] According to the load time series, a three-point window is selected;

[0017] According to the three-point window, determining whether there is a closed cycle of rain flow in the window;

[0018] If there is a closed cycle of rainflow, the amplitude, mean and number of cycles of each rainflow cycle are extracted.

[0019] Preferably, obtaining the accumulated fatigue damage at the current moment includes:

[0020] The load amplitude is corrected based on the Goodman curve, the maximum number of cyclic loads is determined using the SN curve, the fatigue damage value at the current moment t is calculated and corrected in combination with the Palmgren-Miner linear cumulative damage theory, and the original cumulative fatigue damage value is superimposed to obtain the cumulative fatigue damage at the current moment.

[0021] Preferably, correcting the load amplitude based on the Goodman curve includes:

[0022] When the average load is not 0, the Goodman curve is used for correction, and the correction expression is:

[0023]

[0024] In the formula, S ai is the cycle amplitude; S iis the load amplitude when the equivalent mean value is 0; S mi is the cyclic mean; σ b It is the maximum load value of the material at tensile fracture.

[0025] Preferably, the formula for determining the maximum number of cyclic loads using the SN curve is:

[0026] S m ×N=C

[0027] Where S is the load amplitude; N is the maximum number of cyclic loads under this load amplitude; m is the Wohler index, and C is a constant.

[0028] Preferably, the Palmgren-Miner linear cumulative damage theory:

[0029]

[0030] Where n i is the load amplitude S i The number of cycles under the condition, D is the cumulative fatigue damage degree;

[0031] D t =k×D+D t-1

[0032] Where D t is the cumulative fatigue damage at time t, D t-1 is the cumulative fatigue damage at the previous moment, k is the damage correction coefficient, and l win same.

[0033] In a second aspect, the present application also provides a real-time fatigue damage assessment system for a wind farm, comprising:

[0034] A data acquisition module is used to acquire the load data at the current moment, wherein the load data includes the load value, the load extreme value, the starting position of the sliding window and the original accumulated fatigue damage value;

[0035] An extreme value updating module, used for updating the load extreme value according to the load data, and determining whether the end position of the sliding window changes;

[0036] The rain flow counting module is used to process the load data in the sliding window using the rain flow counting method to extract the load amplitude, mean value and number of cycles;

[0037] The fatigue damage calculation module is used to correct the load amplitude and determine the maximum number of cyclic loads, and obtain the cumulative fatigue damage at the current moment in combination with the original cumulative fatigue damage value.

[0038] In a third aspect, the present invention further discloses a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0039] In a fourth aspect, the present invention further discloses a computer program product, comprising a computer program, which implements the steps of the method described in the first aspect when executed by a processor.

[0040] Compared with the prior art, the present invention has the following advantages and technical effects:

[0041] The present invention provides a real-time fatigue damage assessment method for a wind farm, comprising: firstly acquiring load data at the current moment, wherein the load data comprises a load value, a load extreme value, a starting position of a sliding window and an original cumulative fatigue damage value; secondly updating the load extreme value according to the load data, and judging whether the end position of the sliding window has changed; if it has changed, performing rain flow counting method processing on the load data in the sliding window, extracting the load amplitude, mean value and number of cycles; finally correcting the load amplitude, determining the maximum number of load cycles, and obtaining the cumulative fatigue damage at the current moment in combination with the original cumulative fatigue damage value.

[0042] The present invention combines the dynamic sliding window method with the rain flow counting method, solves the problem that the traditional rain flow counting method cannot calculate in real time, and can realize the rapid real-time calculation of the accumulated fatigue damage of the wind turbine on the basis of ensuring accuracy. The method of the present invention can be used for the quantification of fatigue in wind farm optimization, with high calculation accuracy and good optimization effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0044] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0045] Figure 2 is a parameter calculation diagram of an embodiment of the present invention;

[0046] Figure 3 Calculation result diagram of an embodiment of the present invention;

[0047] Figure 4 This is a fatigue growth diagram of a portion of a fan according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0049] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0050] Embodiment 1

[0051] The data set used in this embodiment is a wind farm group data set, including 100 wind turbines in two wind farms and 2000 seconds of shaft torque data. The data of wind farm 1 is used to calculate the window length l win and correction factor k.

[0052] The torque of 100 wind turbines in wind farm 1 is used as the load, and the fatigue damage is first calculated using the traditional rain flow counting method as reference data. The method of this embodiment is used for real-time calculation, and the load is input in time sequence, with time t ranging from 1 to 2000s.

[0053] like Figure 1 As shown, this embodiment provides a real-time fatigue damage assessment method for a wind farm, including:

[0054] Step 1: Input the load x at the current time t t , load extreme value x max and x min , the starting position of the sliding window and the original cumulative fatigue damage value D t-1 ;

[0055] Step 2: Update the load extreme value according to the input load data, and determine whether the end position of the sliding window has changed; if it has not changed, the cumulative fatigue damage is equal to the cumulative damage value at the previous moment, and the calculation ends; if it has changed, go to step 3;

[0056] As an innovative implementation method, the window update formula is as follows:

[0057]

[0058] In the formula, is the end position of the jth sliding window. If the judgment condition is met, go to step 3 to calculate the data in the window. win Indicates the maximum length of the window, l win The value ranges from 5 to 200 steps (discrete integers of 5); the appropriate window length is selected based on the degree of fit between the calculation result and the reference data.

[0059] Step 3: Process the load data in the sliding window using the rain flow counting method to extract the load amplitude, mean value and number of cycles;

[0060] Furthermore, the “three-point” rain flow counting method is used to calculate arrive Load data in the window. The specific steps are as follows:

[0061] Step 301, load data preprocessing;

[0062] The load time series in the sliding window Converted into a series of peak and valley points (i.e. local extreme values), recorded as a sequence Where M≤N. i is a local extreme value, and its judgment condition formula is:

[0063]

[0064] Step 302, load time history reconstruction;

[0065] For the load time series, shear and rearrange at the maximum value so that the maximum point is located at the beginning. The following formula is the load series before and after shear reconstruction:

[0066]

[0067] Step 303, three-point window selection;

[0068] According to the load sequence after shear reconstruction, three consecutive extreme points y are selected in turn. j ,y j+1 ,y j+2 , construct a three-point window (y j ,y j+1 ,y j+2 ).

[0069] Step 304, determining the rain flow closure condition;

[0070] According to the following rules, we can judge whether there is a valid rain flow closed cycle in the window. First, we define the two load amplitudes Δy1 = |y j+1 -y j | and Δy2=|y j+2 -y j+1 |, if Δy1≤Δy2, it is considered that a valid cycle is formed in the window, and the closed cycle amplitude is:

[0071] Δy=Δy1

[0072] Step 305, counting and updating;

[0073] If the closing condition is met in step 304, a complete cycle is recorded, and the cycle amplitude is recorded as Δy, and the mean is:

[0074]

[0075] The closed point (y j ,y j+1 ) is removed from the sequence, and the remaining data continues to iterate steps 303 to 305.

[0076] If the closing condition is not met, keep point y j , continue to process the next set of three points (y j+1 ,y j+2 ,y j+3 ).

[0077] Step 306, non-closed loop processing;

[0078] After traversing all the data, if there are still unclosed points in the sequence, assume that the remaining points form a half cycle, and record their amplitude and mean.

[0079] Step 306, output the result;

[0080] Output the amplitude Δy of each rainflow cycle k , mean y k and the number of cycles n k , and its results are used for subsequent fatigue damage calculations.

[0081] Step 4: Based on the Goodman curve, the load amplitude is corrected, the maximum number of cyclic loads is determined using the SN curve, and the fatigue damage value at the current moment t is calculated and corrected in combination with the Palmgren-Miner linear cumulative damage theory, and the original cumulative damage value is superimposed to obtain the cumulative fatigue damage at the current moment. The specific formula is as follows:

[0082] (1) Goodman curve correction: The basic SN curve is the case where the average load is 0. When the average load is not 0, the Goodman curve is used for correction. The correction expression is:

[0083]

[0084] In the formula, S ai is the cycle amplitude calculated in step 3; S i is the load amplitude when the equivalent mean value is 0; S mi is the cyclic mean calculated in step 3; σ b It is the maximum load value of the material at tensile fracture.

[0085] (2) SN curve:

[0086] S m ×N=C

[0087] Where S is the load amplitude; N is the maximum number of cyclic loads under this load amplitude; m is the Wohler index, which is 10; and C is a constant, which is 9.77×1070.

[0088] (3) Palmgren-Miner linear cumulative damage theory:

[0089]

[0090] Where n i is the load amplitude S i When the cumulative fatigue damage degree D reaches 1, the material fails.

[0091] (4) Cumulative fatigue damage:

[0092] D t =k×D+D t-1

[0093] Where D t and D t-1 is the cumulative fatigue damage between time t and the previous time, k is the damage correction factor, and the discrete data with a step length of 0.1 from 0.1 to 10 is in each window length l win , select the value corresponding to the maximum correlation coefficient between the calculated result and the reference data.

[0094] Through the method of this embodiment, the calculation results are obtained, such as Figure 2 The determination coefficients under different window lengths indicate the calculation accuracy of the method of the present invention. When the window length is selected to be 20, the determination coefficient reaches above 0.8, and the corresponding correction coefficient is 0.9.

[0095] In this embodiment, the data of wind farm 2 is used to verify the effect of the method of this embodiment. The torque of 100 wind turbines in wind farm 2 is used as the load, and the fatigue damage is first calculated using the traditional rain flow counting method as reference data. The method of this embodiment is used for real-time calculation, and the specific steps are the same as above, the window length is fixed to 20, and the correction coefficient is fixed to 0.9.

[0096] In this embodiment, the test formula of the determination coefficient is as follows:

[0097]

[0098] In the formula, R 2 is the coefficient of determination, the closer it is to 1, the better the data fitting effect; The calculation results of the embodiment method are as follows: is the average value; d i This is the result calculated by the traditional rain flow counting method.

[0099] Table 1 is a comparison of the calculation time of the method of this embodiment and the traditional method. The computing platform is MATLAB 2024a, the CPU is Intel i5-12400F, the memory is 16GB DDR4, and the operating system is Windows 10. It can be seen that the calculation time is much shorter than the traditional rain flow counting method.

[0100] Table 1

[0101]

[0102] Figure 3 The fatigue damage of 100 wind turbines is fitted to the reference data. The determination coefficient of the calculation result of each wind turbine is above 0.96. The method of this embodiment can accurately calculate the fatigue damage in real time.

[0103] Figure 4 This is the cumulative fatigue damage growth diagram of some fans. It can be seen that the growth trend is similar to the original data.

[0104] Embodiment 2

[0105] Based on the same inventive concept, this embodiment also provides a real-time fatigue damage assessment system for a wind farm, comprising:

[0106] A data acquisition module is used to acquire the load data at the current moment, wherein the load data includes the load value, the load extreme value, the starting position of the sliding window and the original accumulated fatigue damage value;

[0107] An extreme value updating module, used for updating the load extreme value according to the load data, and determining whether the end position of the sliding window changes;

[0108] The rain flow counting module is used to process the load data in the sliding window using the rain flow counting method to extract the load amplitude, mean value and number of cycles;

[0109] The fatigue damage calculation module is used to correct the load amplitude and determine the maximum number of cyclic loads, and obtain the cumulative fatigue damage at the current moment in combination with the original cumulative fatigue damage value.

[0110] The real-time fatigue damage assessment system for a wind farm provided by this embodiment has all the advantages of the real-time fatigue damage assessment method for a wind farm provided by the first embodiment.

[0111] Embodiment 3

[0112] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first embodiment are implemented.

[0113] Embodiment 4

[0114] This embodiment also discloses a computer program product, including a computer program, which implements the steps of the method described in the first embodiment when executed by a processor.

[0115] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A real-time fatigue damage assessment method for a wind farm, characterized in that: The following steps are involved: Obtaining load data at the current moment, wherein the load data includes a load value, a load extreme value, a starting position of a sliding window, and an original accumulated fatigue damage value; updating the load extreme value according to the load data, and determining whether the end position of the sliding window changes; If there is a change, the load data in the sliding window is processed by the rain flow counting method to extract the load amplitude, mean value and number of cycles; The load amplitude is corrected, and the maximum number of cyclic loads is determined, and the accumulated fatigue damage at the current moment is obtained by combining the original accumulated fatigue damage value.

2. The method according to claim 1, characterized in that Formula for updating load extreme values: In the formula, x t Expressed as the load at time t, x max and x min Expressed as the load extreme value, and They represent the starting and ending positions of the jth sliding window, respectively. win Expressed as the maximum window length.

3. The method according to claim 1, characterized in that The rainflow counting method is used to process the load data in the sliding window, including: Preprocessing the load data in the sliding window to obtain preprocessed load data; Shearing and reconstructing the preprocessed load data to obtain a load time series; According to the load time series, a three-point window is selected; According to the three-point window, determining whether there is a closed cycle of rain flow in the window; If there is a closed cycle of rainflow, the amplitude, mean and number of cycles of each rainflow cycle are extracted.

4. The method according to claim 1, characterized in that: The accumulated fatigue damage at the current moment includes: The load amplitude is corrected based on the Goodman curve, the maximum number of cyclic loads is determined using the SN curve, the fatigue damage value at the current moment t is calculated and corrected in combination with the Palmgren-Miner linear cumulative damage theory, and the original cumulative fatigue damage value is superimposed to obtain the cumulative fatigue damage at the current moment.

5. The method according to claim 4, characterized in that Correction of the load amplitude based on the Goodman curve includes: When the average load is not 0, the Goodman curve is used for correction, and the correction expression is: In the formula, S ai is the cycle amplitude; S i is the load amplitude when the equivalent mean value is 0; S mi is the cyclic mean; σ b It is the maximum load value of the material at tensile fracture.

6. The method according to claim 4, characterized in that The formula for determining the maximum number of cyclic loads using the SN curve is: S m ×N=C Where S is the load amplitude; N is the maximum number of cyclic loads under this load amplitude; m is the Wohler index, and C is a constant.

7. The method according to claim 1, characterized in that Palmgren-Miner linear cumulative damage theory: Where n i is the load amplitude S i The number of cycles under the condition, D is the cumulative fatigue damage degree; D t =k×D+D t-1 Where D t is the cumulative fatigue damage at time t, D t-1 is the cumulative fatigue damage at the previous moment, k is the damage correction coefficient, and l win same.

8. A real-time fatigue damage assessment system for a wind farm, characterized in that: include: A data acquisition module is used to acquire the load data at the current moment, wherein the load data includes the load value, the load extreme value, the starting position of the sliding window and the original accumulated fatigue damage value; An extreme value updating module, used for updating the load extreme value according to the load data, and determining whether the end position of the sliding window changes; The rain flow counting module is used to process the load data in the sliding window using the rain flow counting method to extract the load amplitude, mean value and number of cycles; The fatigue damage calculation module is used to correct the load amplitude and determine the maximum number of cyclic loads, and obtain the cumulative fatigue damage at the current moment in combination with the original cumulative fatigue damage value.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images