CN117905651A - Structure monitoring method based on deformation analysis - Google Patents
Structure monitoring method based on deformation analysis Download PDFInfo
- Publication number
- CN117905651A CN117905651A CN202311796239.XA CN202311796239A CN117905651A CN 117905651 A CN117905651 A CN 117905651A CN 202311796239 A CN202311796239 A CN 202311796239A CN 117905651 A CN117905651 A CN 117905651A
- Authority
- CN
- China
- Prior art keywords
- data
- period
- blade
- rotation
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012544 monitoring process Methods 0.000 title claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 89
- 230000000737 periodic effect Effects 0.000 claims abstract description 50
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 5
- 238000012935 Averaging Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Landscapes
- Wind Motors (AREA)
Abstract
The invention relates to the technical field of wind driven generator monitoring, in particular to a structure monitoring method based on deformation analysis. The invention obtains the rotation period of each blade; collecting data at a preset angle position in each section of rotation period to obtain period data of each section of rotation period of each blade; obtaining initial similarity according to the data integral difference between the periodic data with the same period length; obtaining a time fluctuation condition according to the variation degree of the rotation period time interval; obtaining the length change of the rotation period, and further obtaining the influence degree of wind speed; obtaining final similarity by correcting the influence degree of wind speed and obtaining an abnormal period data cluster; and determining the abnormal position of the blade according to the strain data difference at the same position between the abnormal blade and other blades. According to the method, the influence of the wind speed on the strain force data is calculated so as to correct the clustering condition, and the clustering result can be used for more accurately monitoring whether the blade structure is abnormal or not.
Description
Technical Field
The invention relates to the technical field of wind driven generator monitoring, in particular to a structure monitoring method based on deformation analysis.
Background
Wind power generation is widely used in China as one of clean energy sources, and the wind turbine generator can be damaged due to rapid rotation of blades in the power generation process. The blade needs to be detected and maintained after working for a period of time, and serious consequences caused by damage to the whole unit due to damage of the blade during working are avoided. Because the blade needs to stop working when checking, the working state and economic benefit of the blade are affected, the strain distribution condition of the blade can be detected by installing the strain gauge sensor on the blade, and the blade can be monitored in real time according to the strain variation condition of the blade.
In the prior art, a clustering algorithm is often used for classifying the collected strain forces of the blades, and abnormal analysis is carried out on the strain force data according to the clustered clusters. However, in the clustering process, the influence of the external environment on the strain data is not considered to a certain extent, so that the normal strain data is clustered into abnormal data point clusters, and the accurate monitoring of the blade structure is influenced.
Disclosure of Invention
In order to solve the technical problem that normal strain force data is clustered into abnormal data point clusters and accurate monitoring of a blade structure is affected because the strain force data can not be influenced to a certain extent by an external environment in the clustering process, the invention aims to provide a structural monitoring method based on deformation analysis, and the adopted technical scheme is as follows:
a method of structural monitoring based on deformation analysis, the method comprising:
Obtaining a blade strain dataset for each blade in the wind turbine; the blade strain data set includes all strain data for different locations on each blade;
obtaining each section of rotation period of each blade; acquiring all strain force data sets of the blades acquired at different preset rotation angle positions in each rotation period of each blade to form period data in each rotation period of each blade;
Obtaining initial similarity of the periodic data with the same length of any two rotation periods according to the data integral difference of the periodic data with the same length of any two rotation periods; according to the change degree of the time interval of the adjacent blades passing through the same position in any two rotation periods with the same length, obtaining the time fluctuation condition of the period data with the same period length in the corresponding rotation period; acquiring cycle data of any two same cycle lengths, wherein the cycle data correspond to the rotation cycle length change of each blade rotation cycle; obtaining the wind speed influence degree of the period data with the same length of any two rotation periods according to the time fluctuation condition and the change of the length of the rotation periods;
Correcting the initial similarity by utilizing the wind speed influence degree to obtain a final similarity; clustering the periodic data with the same period length by utilizing the final similarity degree to determine an abnormal periodic data cluster; and determining the abnormal position on the abnormal blade according to the strain force data difference at the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster.
Further, the method for acquiring the overall difference of the data comprises the following steps:
Calculating the difference between the average values of the strain force data sets of the blades at each corresponding angle position in the period data in any two rotation periods with the same length as the corresponding position difference;
And averaging all the corresponding position differences between the period data in two rotation periods with the same length to obtain the data overall difference.
Further, the method for acquiring the initial similarity degree comprises the following steps:
And carrying out negative correlation mapping and normalization processing on the data overall difference to obtain the initial similarity degree.
Further, the method for acquiring the variation degree of the time interval comprises the following steps:
Acquiring all time intervals of the adjacent two blades passing through the same position in each rotation period;
The variance between all of the time intervals contained in each rotation period is taken as the degree of variation of the time intervals within each rotation period.
Further, the method for acquiring the time fluctuation condition comprises the following steps:
and taking the maximum value of the variation degree of the time interval between any two rotation periods with the same length as the time fluctuation condition in any two rotation periods with the same length.
Further, the method for acquiring the rotation period length variation includes:
The first rotation period of each blade is removed, and the difference between the time length of the remaining each rotation period of each blade and the time length of the previous rotation period is taken as the rotation period length change of the remaining each rotation period of each blade.
Further, the method for acquiring the wind speed influence degree comprises the following steps:
Taking the difference between the rotation period length changes of any two rotation periods with the same length as a period change difference;
taking the product of the time fluctuation condition and the periodic variation difference as the wind speed influence degree.
Further, the method for obtaining the final similarity degree comprises the following steps:
obtaining the final similarity according to a final similarity calculation formula, wherein the final similarity calculation formula is as follows:
D i ′＝norm(Di×max(εi)×(|Δl1-Δl2 |)) are described; wherein D i ′ represents the final degree of similarity between the ith pair of periodic data; d i denotes an initial degree of similarity between the ith pair of periodic data; epsilon i represents the variation degree of the time interval in the rotation period of each period data in the ith pair of period data; Δl 1 denotes a rotation period length change of the first period data between the ith pair of period data; Δl 2 denotes a rotation period length change of the second period data between the ith pair of period data; max (epsilon i) represents the time fluctuation condition of the ith pair of periodic data; the i Δl 1-Δl2 represents the periodic variation difference between the i-th pair of periodic data; norm () represents a normalization function; max () represents the maximum function.
Further, determining an abnormal position on the abnormal blade from a difference in strain data at the same position between the corresponding abnormal blade and the other blades in the abnormal period data cluster includes:
for one abnormal period data in the abnormal period data cluster, selecting one rotation angle position in the abnormal period data as a reference rotation angle position;
before the abnormal blades pass through the reference rotation angle position in time sequence, the strain force data sets when other blades pass through the reference rotation angle position are used as analysis strain force data sets, and the number of the analysis strain force data sets is the same as that of the other blades;
calculating and analyzing the data average value of the strain force data at each position in the strain force data set, normalizing the data difference between the strain force data at each position in the abnormal strain force data set and the data average value at the corresponding position in the analysis strain force data set, and considering the position as the abnormal position of the abnormal blade when the normalized data difference is larger than a preset first threshold value.
Further, the first threshold is set to 0.75.
The invention has the following beneficial effects:
According to the invention, all the strain force data sets of the blades in the wind driven generator are obtained, and the strain force data in the data sets can reflect the structural characteristics of each blade; the time interval when two adjacent blades pass through the preset position and the rotation period of each blade are acquired, so that the influence of the subsequent wind speed on the period change is facilitated; acquiring all strain force data sets of the blades acquired under different preset angle positions in each section of rotation period of each blade to form period data in each section of rotation period of each blade, so that the similarity degree between different period data can be conveniently analyzed later, and the corresponding data set of each strain force data set of the blade between any two period data with the same length can be ensured; obtaining initial similarity degree between the period data according to the overall difference of the data between different period data with the same length of any two rotation periods, wherein the initial similarity degree reflects whether strain force at corresponding positions in different rotation periods changes or not under the influence of no wind speed; obtaining time fluctuation conditions in any two rotation periods with the same length according to the change degree of the time interval in one rotation period, and obtaining the rotation period length change between every two adjacent rotation periods of each blade, wherein the time fluctuation conditions and the rotation period length change reflect the influence of wind speed on period data; obtaining wind speed influence degree according to time fluctuation condition and rotation period length change, correcting initial similarity degree by using the wind speed influence degree to obtain final similarity degree, reducing influence of wind speed on similarity degree among period data, and enabling a subsequent clustering result to be more accurate; clustering the periodic data with the same period length by utilizing the final similarity degree to determine an abnormal periodic data cluster; and determining the abnormal position on the abnormal blade according to the strain force data difference at the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster. According to the method, the influence of the wind speed on the strain force data is calculated so as to correct the clustering condition, and the clustering result can be used for more accurately monitoring whether the blade structure is abnormal or not.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a structure monitoring method based on deformation analysis according to an embodiment of the present invention;
fig. 2 is a schematic diagram illustrating a strain sensor according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description refers to specific embodiments, structures, features and effects of a structure monitoring method based on deformation analysis according to the present invention, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the structure monitoring method based on deformation analysis provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a structure monitoring method based on deformation analysis according to an embodiment of the present invention is shown, where the method includes:
Step S1: obtaining a blade strain dataset for each blade in the wind turbine; the blade strain data set includes all strain data for different locations on each blade.
The embodiment of the invention provides a structure detection method based on deformation analysis, which aims at carrying out deformation analysis on a blade structure of a wind driven generator so as to detect the blade structure, so that strain data of blades are firstly required to be obtained, and all strain data of each blade are formed into a blade strain data set of each blade.
In one embodiment of the invention, the blades of the wind driven generator are required to be stressed uniformly so as to ensure that the wind driven generator can be balanced and stable, so that the number of the blades of the wind driven generator is 3 and the structures of the blades are completely the same. A strain sensor is installed on each blade of a wind driven generator to ensure that strain data collected by the strain sensors can reflect structural characteristics of the blade, and particularly referring to FIG. 2, a strain sensor installation schematic diagram is provided, and strain data collected by the strain sensors can reflect strain change conditions of the blade during working. The strain sensors are installed in each blade in the same way to ensure that the locations where all the blades are structurally inspected are the same. It should be noted that, the installation mode of the strain sensor may be set by an operator according to a specific implementation scenario, which is a technical means known to those skilled in the art, and is not limited herein.
To this end, a blade strain data set is obtained for each blade.
Step S2: obtaining each section of rotation period of each blade; and acquiring all strain force data sets of the blades acquired at different preset rotation angle positions in each rotation period of each blade to form period data in each rotation period of each blade.
The blades of the wind driven generator can periodically rotate in the working process, the strain force changes in the rotating process can also periodically change, and when the strain force change conditions on the blades are subjected to abnormal analysis, whether the strain force data are abnormal or not is judged according to the change conditions of the strain force data of different rotation periods, so that in the embodiment of the invention, each section of rotation period of each blade needs to be acquired first. Because the positions of different blades are different, the corresponding time periods of the same rotation period of the different blades have hysteresis, and for the different blades, the blade strain data sets are divided according to the corresponding time points of the uniform rotation period, so that the clustering result is inaccurate due to the influence of the position of the blades in the subsequent clustering operation, and therefore, for the different blades, each rotation period of each blade needs to be obtained by adopting different period division modes.
In one embodiment of the invention, a fixed laser collector is arranged on a tower of the wind driven generator, when the blades rotate to pass through the position of the laser collector, the passing time points of the blades are recorded, the time interval of two adjacent blades passing through the laser collector is obtained, according to the actual situation, the number of the blades is 3, the time period of each blade passing through the position of the laser collector twice continuously is taken as one rotation period of each blade, and each rotation period comprises the time interval of three adjacent two blades passing through the laser collector, so that each rotation period and the corresponding rotation period length of each rotation period can be obtained, and the subsequent analysis on the influence caused by wind speed is facilitated.
It should be noted that, the embodiment of the present invention provides an expression form of a rotation period and a corresponding rotation period length: acquiring a time interval aggregate { t 1,t2,t3…tz } when two adjacent blades pass through the position of the laser collector; wherein z represents the time interval serial number of the position where the collected adjacent two blades pass through the laser collector; numbering the blades of the wind driven generator according to the sequence of the positions of the blades passing through the laser collector, wherein the blades are respectively a No. 1 blade, a No. 2 blade and a No. 3 blade; the length of the rotation period corresponding to the first rotation period of the blade 1 is t 1+t2+t3, the length of the rotation period corresponding to the first rotation period of the blade 2 is t 2+t3+t4, and the length of the rotation period corresponding to the first rotation period of the blade 3 is t 3+t4+t5; thereby obtaining a rotation period length of each rotation period of each blade corresponding to each rotation period.
It should be noted that, the method of obtaining the time point when the blade passes through the preset position is not unique, and the method is a technical means known to those skilled in the art, and can be set by the implementation personnel, which is not limited and described herein.
Because the strain force born by the blade comprises aerodynamic force and inertial force born by the blade at the current moment, in order to analyze the influence of wind speed subsequently, the inertia force born by different blade strain force data sets needs to be kept consistent when being compared, so in the embodiment of the invention, all the blade strain force data sets acquired under different preset angle positions in each blade rotation period are obtained to form period data in each blade rotation period.
In one embodiment of the present invention, a method for acquiring periodic data includes:
On the plane where all the blades are located, a Cartesian coordinate system is established by the center of the wind driven generator, and a preset angle position is set: when each blade rotates to a preset angle position on a Cartesian coordinate system, all strain force data of each blade are acquired, all strain force data sets of each blade in each section of rotation period are acquired, cycle data of each blade in each section of rotation period are formed, and follow-up analysis of the strain force data sets of the blade is facilitated.
To this end, cycle data for each blade per segment of rotation cycle is obtained.
Step S3: according to the data integral difference of the periodic data with the same length of any two rotation periods, obtaining the initial similarity degree of the periodic data with the same length of any two rotation periods; according to the change degree of the time interval of the adjacent blades passing through the same position in any two rotation periods with the same length, obtaining the time fluctuation condition of the period data with the same period length in the corresponding rotation period; acquiring cycle data of any two same cycle lengths, wherein the cycle data correspond to the rotation cycle length change of each blade rotation cycle; and obtaining the wind speed influence degree of the period data with the same length of any two rotation periods according to the time fluctuation condition and the change of the length of the rotation periods.
Because the materials of different blades are completely consistent in size, the strain data acquired by the blades when the blades are in the same position in the same external environment are consistent, so that for different rotation periods with the same rotation period length, if the blade structure is not changed, the strain data sets of the blades in each corresponding position in the period data are similar, and the difference between the whole period data is small. Therefore, in the embodiment of the invention, the data overall difference of the periodic data with the same length of any two rotation periods is obtained, and the initial similarity degree of the periodic data with the same length of any two rotation periods is further obtained.
Preferably, in one embodiment of the present invention, the method for acquiring the overall difference of data includes:
calculating the difference between the average values of the strain force data sets of the blades at each corresponding angle position in the period data in any two rotation periods with the same length as the corresponding position difference; and averaging all corresponding position differences between the period data in two rotation periods with the same length to obtain the data overall difference.
Preferably, in one embodiment of the present invention, the method for obtaining the initial similarity includes:
and carrying out negative correlation mapping and normalization processing on the overall difference of the data to obtain the initial similarity.
In one embodiment of the present invention, the initial similarity calculation formula is as follows:
Wherein D i represents an initial degree of similarity between the ith period data of the same rotation period length; ΔR k represents the corresponding positional difference between the kth corresponding angular position blade strain dataset between the ith period data of the same rotational period length; n represents the number of blade strain data sets in each cycle of data for which the i-th rotation cycle is the same length; k represents the serial number of the blade strain data set at the corresponding angle position between the ith period data with the same rotation period length; Representing i data overall differences between period data of the same rotation period length; exp () represents an exponential function that bases on a natural constant.
In the initial similarity calculation formula, calculating the similarity between the blade strain data sets at corresponding positions in two period data with the same rotation period length through delta R k, and averaging the similarity between the blade strain data sets at all corresponding positions between the two period data to obtain the data overall difference between the period data, wherein the smaller the data overall difference is, the smaller the change of the blade strain data sets at all positions in the period data is, and the more similar the period data of different rotation periods with the same period length is.
In practice, the blade may be subjected to various factors, such as wind speed, structural fatigue of the blade itself, etc., so that a large difference occurs between two different cycle data having the same cycle length. Regarding the influence of fatigue of the blade structure on the periodic data, it is generally considered that the blade structure may be deformed in a period of time corresponding to the periodic data; in addition, as for the influence of wind speed on the period data, the wind force received by the blades of the wind driven generator in different rotation periods is different, and the strain force received by the blades is also different, but in the specified range of the wind driven generator, the wind speed does not influence the structure of the blades, so that the influence of the wind speed on the initial similarity degree between the period data is reduced. Therefore, in the embodiment of the invention, the change degree of the time interval of the adjacent blades passing through the same position is adopted to represent the influence of wind speed on the inside of the rotation period, and the time fluctuation condition in the rotation period is obtained.
Preferably, in one embodiment of the present invention, the method for acquiring the variation degree of the time interval includes:
acquiring all time intervals of the adjacent two blades passing through the same position in each rotation period; the variance between all time intervals contained in each rotation period is taken as the degree of variation of the time intervals within each rotation period.
Preferably, in one embodiment of the present invention, the method for acquiring a time fluctuation condition includes:
And taking the maximum value of the variation degree of the time interval between any two rotation periods with the same length as the time fluctuation condition in any two rotation periods with the same length.
The wind speed affects the interior of the rotation period and also affects the variation between rotation periods. Therefore, according to the embodiment of the invention, the length change of the rotation period of each of any two rotation periods with the same length is obtained, and the influence degree of wind speed is obtained according to the time fluctuation condition and the length change of the rotation period.
Preferably, in one embodiment of the present invention, the method for acquiring the rotation period length variation includes:
Since deformation of the blade structure may occur after the blades are rotated for a long time, the first rotation period of each blade may be eliminated, and the difference between the time length of each remaining rotation period of each blade and the time length of the previous rotation period is regarded as the change in the rotation period length of each remaining rotation period of each blade.
Preferably, in one embodiment of the present invention, a method for acquiring a wind speed influence degree includes:
Taking the difference between the rotation period length changes of any two rotation periods with the same length as a period change difference; taking the product of the time fluctuation condition and the periodic variation difference as the wind speed influence degree.
Step S4: correcting the initial similarity by utilizing the influence degree of wind speed to obtain the final similarity; clustering the periodic data with the same period length by utilizing the final similarity degree to determine an abnormal periodic data cluster; and determining the abnormal position on the abnormal blade according to the strain force data difference at the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster.
The influence degree of the wind speed obtained in the step S3 can obtain the influence of the wind speed on the period data in different rotation periods with the same period, and in order to reduce the misjudgment of the subsequent clustering operation on the normal period data, the embodiment of the invention needs to correct the initial similarity degree by utilizing the influence degree of the wind speed to obtain the final similarity degree.
Preferably, in one embodiment of the present invention, the method for obtaining the final similarity degree includes:
Obtaining a final similarity according to a final similarity calculation formula, wherein the final similarity calculation formula is as follows:
Di ′＝norm(Di×max(εi)×(|Δl1-Δl2|))
Wherein D i ′ represents the final degree of similarity between the ith pair of periodic data; d i denotes an initial degree of similarity between the ith pair of periodic data; epsilon i represents the variation degree of the time interval in the rotation period of each period data in the ith pair of period data; Δl 1 denotes a rotation period length change of the first period data between the ith pair of period data; Δl 2 denotes a rotation period length change of the second period data between the ith pair of period data; max (epsilon i) represents the time fluctuation condition of the ith pair of periodic data; the i Δl 1-Δl2 represents the periodic variation difference between the i-th pair of periodic data; norm () represents a normalization function; max () represents the maximum function.
In the final similarity calculation formula, max (epsilon i)×(|Δl1-Δl2 |) represents the wind speed influence degree of the ith pair of periodic data, when the wind speed influence degree is larger, the difference between the ith pair of periodic data is larger, namely the initial similarity is smaller, and in order to reduce the influence of the wind speed on the similarity between the periodic data, the initial similarity between the ith pair of periodic data is adjusted so that the final similarity between the ith pair of periodic data is larger, namely the final similarity and the initial similarity and the wind speed influence degree are in positive correlation.
And clustering the periodic data with the same period length according to the final similarity degree to determine an abnormal periodic data cluster. Because the wind driven generator is a whole, when one blade generates a problem, the stability of the whole wind driven generator can be influenced, and the whole strain force data of the wind driven generator can possibly show an abnormal condition, so that the abnormal position on the abnormal blade needs to be determined through the strain force data difference on the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster.
In one embodiment of the invention, an iterative self-organizing clustering algorithm is adopted to cluster all periodic data, and abnormal periodic data clusters with the same period length are obtained from different periodic data with the same period length; the cycle data in the abnormal cycle data cluster is considered to have abnormal strain data, that is, the blade corresponding to the abnormal cycle data is an abnormal blade, and at this time, the abnormal position on the abnormal blade needs to be further determined.
Preferably, in one embodiment of the present invention, determining the abnormal position on the abnormal blade according to the difference in strain force data at the same position between the corresponding abnormal blade and the other blades in the abnormal period data cluster includes:
for one abnormal period data in the abnormal period data cluster, selecting one rotation angle position in the abnormal period data as a reference rotation angle position; before the abnormal blades pass through the reference rotation angle position in time sequence, the strain force data sets when other blades pass through the reference rotation angle position are used as analysis strain force data sets, and the number of the analysis strain force data sets is the same as that of the other blades; calculating and analyzing the data average value of the strain force data at each sensor position in the strain force data set, normalizing the data difference between the strain force data of each sensor position in the abnormal strain force data set and the data average value of the corresponding sensor position in the analysis strain force data set, and considering the position as the abnormal position of the abnormal blade when the normalized data difference is larger than a preset first threshold value.
It should be noted that, one embodiment of the present invention provides a procedure for determining an abnormal position of an abnormal blade:
The abnormal cycle data is a cycle of starting at 16 points of the blade 1,0 degree is selected as a reference rotation angle in the cycle, before 16 points, data when the blade 2 and the blade 3 rotate to 0 degree is counted, the counted two data are used as analysis strain data sets, the strain data of the same position between the blade 2 and the blade 3 are averaged to obtain a data average value of each position of the analysis strain data sets, normalization processing is carried out on data differences between the strain data of each position in the abnormal strain data set of the blade 1 and the data average value of each position in the analysis strain data set, and when the normalized data differences are larger than a preset first threshold value, the position at the moment is considered to be an abnormal position of the abnormal blade.
Preferably, in one embodiment of the present invention, the first threshold is set to 0.75. It should be noted that, the setting of the first threshold may be set by an implementer according to a specific implementation scenario, which is not limited herein.
It should be noted that the iterative self-organizing clustering algorithm is a technical means well known to those skilled in the art, and will not be described herein. In addition, in other embodiments of the present invention, other algorithms such as a K-means clustering algorithm may be used to cluster the stress data, which is not limited herein.
Thus, structural monitoring of the wind driven generator is completed.
In summary, the invention obtains all blade strain data sets in the wind driven generator; the time interval when two adjacent blades pass through the preset position and the rotation period of each blade are acquired, so that the influence of the subsequent wind speed on the period change is facilitated; dividing a blade strain data set of each blade according to the rotation period to obtain period data; obtaining initial similarity degree between the period data according to the overall difference of the data between different period data with the same length of any two rotation periods, wherein the initial similarity degree reflects whether strain force at corresponding positions in different rotation periods changes or not under the influence of no wind speed; obtaining time fluctuation conditions in any two rotation periods with the same length according to the change degree of the time interval in one rotation period, and obtaining rotation period length change between the rotation period length and the previous rotation period length, wherein the time fluctuation conditions and the rotation period length change reflect the influence of wind speed on period data; obtaining wind speed influence degree according to time fluctuation condition and rotation period length change, correcting initial similarity degree by using the wind speed influence degree to obtain final similarity degree, reducing influence of wind speed on similarity degree among period data, and enabling a subsequent clustering result to be more accurate; clustering all strain data at the same position of all the blades according to the final similarity degree, and finding out abnormal strain data points at the acquisition positions in each blade; clustering the periodic data with the same period length by utilizing the final similarity degree to determine an abnormal periodic data cluster; and determining the abnormal position on the abnormal blade according to the strain force data difference at the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster. According to the method, the influence of the wind speed on the strain force data is calculated so as to correct the clustering condition, and the clustering result can be used for more accurately monitoring whether the blade structure is abnormal or not.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (10)
1.A method of structural monitoring based on deformation analysis, the method comprising:
Obtaining a blade strain dataset for each blade in the wind turbine; the blade strain data set includes all strain data for different locations on each blade;
obtaining each section of rotation period of each blade; acquiring all strain force data sets of the blades acquired at different preset rotation angle positions in each rotation period of each blade to form period data in each rotation period of each blade;
Obtaining initial similarity of the periodic data with the same length of any two rotation periods according to the data integral difference of the periodic data with the same length of any two rotation periods; according to the change degree of the time interval of the adjacent blades passing through the same position in any two rotation periods with the same length, obtaining the time fluctuation condition of the period data with the same period length in the corresponding rotation period; acquiring cycle data of any two same cycle lengths, wherein the cycle data correspond to the rotation cycle length change of each blade rotation cycle; obtaining the wind speed influence degree of the period data with the same length of any two rotation periods according to the time fluctuation condition and the change of the length of the rotation periods;
Correcting the initial similarity by utilizing the wind speed influence degree to obtain a final similarity; clustering the periodic data with the same period length by utilizing the final similarity degree to determine an abnormal periodic data cluster; and determining the abnormal position on the abnormal blade according to the strain force data difference at the same position between the corresponding abnormal blade and other blades in the abnormal period data cluster.
2. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for obtaining the overall difference of data comprises:
Calculating the difference between the average values of the strain force data sets of the blades at each corresponding angle position in the period data in any two rotation periods with the same length as the corresponding position difference;
And averaging all the corresponding position differences between the period data in two rotation periods with the same length to obtain the data overall difference.
3. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for obtaining the initial similarity comprises:
And carrying out negative correlation mapping and normalization processing on the data overall difference to obtain the initial similarity degree.
4. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for obtaining the variation degree of the time interval comprises the steps of:
Acquiring all time intervals of the adjacent two blades passing through the same position in each rotation period;
The variance between all of the time intervals contained in each rotation period is taken as the degree of variation of the time intervals within each rotation period.
5. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for obtaining the time fluctuation condition comprises the following steps:
and taking the maximum value of the variation degree of the time interval between any two rotation periods with the same length as the time fluctuation condition in any two rotation periods with the same length.
6. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for acquiring the rotation period length variation comprises:
The first rotation period of each blade is removed, and the difference between the time length of the remaining each rotation period of each blade and the time length of the previous rotation period is taken as the rotation period length change of the remaining each rotation period of each blade.
7. The method for monitoring a structure based on deformation analysis according to claim 5, wherein the method for obtaining the influence degree of wind speed comprises:
Taking the difference between the rotation period length changes of any two rotation periods with the same length as a period change difference;
taking the product of the time fluctuation condition and the periodic variation difference as the wind speed influence degree.
8. The method for monitoring a structure based on deformation analysis according to claim 1, wherein the method for obtaining the final similarity degree comprises:
obtaining the final similarity according to a final similarity calculation formula, wherein the final similarity calculation formula is as follows:
D i ′＝norm(Di×max(εi)×(|Δl1-Δl2 |)) are described; wherein D i ′ represents the final degree of similarity between the ith pair of periodic data; d i denotes an initial degree of similarity between the ith pair of periodic data; epsilon i represents the variation degree of the time interval in the rotation period of each period data in the ith pair of period data; Δl 1 denotes a rotation period length change of the first period data between the ith pair of period data; Δl 2 denotes a rotation period length change of the second period data between the ith pair of period data; max (epsilon i) represents the time fluctuation condition of the ith pair of periodic data; the i Δl 1-Δl2 represents the periodic variation difference between the i-th pair of periodic data; norm () represents a normalization function; max () represents the maximum function.
9. The method of claim 1, wherein determining the abnormal position on the abnormal blade based on the difference in strain data at the same position between the corresponding abnormal blade and the other blades in the abnormal cycle data cluster comprises:
for one abnormal period data in the abnormal period data cluster, selecting one rotation angle position in the abnormal period data as a reference rotation angle position;
before the abnormal blades pass through the reference rotation angle position in time sequence, the strain force data sets when other blades pass through the reference rotation angle position are used as analysis strain force data sets, and the number of the analysis strain force data sets is the same as that of the other blades;
calculating and analyzing the data average value of the strain force data at each position in the strain force data set, normalizing the data difference between the strain force data at each position in the abnormal strain force data set and the data average value at the corresponding position in the analysis strain force data set, and considering the position as the abnormal position of the abnormal blade when the normalized data difference is larger than a preset first threshold value.
10. The method of claim 9, wherein the first threshold is set to 0.75.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311796239.XA CN117905651A (en) | 2023-12-25 | 2023-12-25 | Structure monitoring method based on deformation analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311796239.XA CN117905651A (en) | 2023-12-25 | 2023-12-25 | Structure monitoring method based on deformation analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117905651A true CN117905651A (en) | 2024-04-19 |
Family
ID=90693549
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311796239.XA Pending CN117905651A (en) | 2023-12-25 | 2023-12-25 | Structure monitoring method based on deformation analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117905651A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102341597A (en) * | 2009-03-02 | 2012-02-01 | 苏司兰能源有限公司 | Method for monitoring wind turbines |
| US20170074743A1 (en) * | 2015-09-11 | 2017-03-16 | Mitsubishi Heavy Industries, Ltd. | Method of calibrating load measurement apparatus, load measurement system of wind turbine blade, and wind turbine |
| CN106837709A (en) * | 2017-04-20 | 2017-06-13 | 北京金风科创风电设备有限公司 | Monitoring method and monitoring system for wind generating set blade |
| CN108087210A (en) * | 2017-12-19 | 2018-05-29 | 北京金风科创风电设备有限公司 | Wind generating set blade abnormity identification method and device |
| US20190010923A1 (en) * | 2017-07-10 | 2019-01-10 | WindESCo, Inc. | System and method for augmenting control of a wind turbine assembly |
| CN116498496A (en) * | 2023-03-27 | 2023-07-28 | 国家电投集团江苏海上风力发电有限公司 | Wind generating set monitoring management system and method based on multi-sensor signals |
-
2023
- 2023-12-25 CN CN202311796239.XA patent/CN117905651A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102341597A (en) * | 2009-03-02 | 2012-02-01 | 苏司兰能源有限公司 | Method for monitoring wind turbines |
| US20170074743A1 (en) * | 2015-09-11 | 2017-03-16 | Mitsubishi Heavy Industries, Ltd. | Method of calibrating load measurement apparatus, load measurement system of wind turbine blade, and wind turbine |
| CN106837709A (en) * | 2017-04-20 | 2017-06-13 | 北京金风科创风电设备有限公司 | Monitoring method and monitoring system for wind generating set blade |
| US20190010923A1 (en) * | 2017-07-10 | 2019-01-10 | WindESCo, Inc. | System and method for augmenting control of a wind turbine assembly |
| CN108087210A (en) * | 2017-12-19 | 2018-05-29 | 北京金风科创风电设备有限公司 | Wind generating set blade abnormity identification method and device |
| CN116498496A (en) * | 2023-03-27 | 2023-07-28 | 国家电投集团江苏海上风力发电有限公司 | Wind generating set monitoring management system and method based on multi-sensor signals |
Non-Patent Citations (1)
| Title |
|---|
| 陶佳丽;王敏;周游游;陈冰;: "风电叶片结构健康监测系统", 电气自动化, no. 04, 30 July 2016 (2016-07-30), pages 62 - 64 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10072635B2 (en) | Wind turbine and method for evaluating health state of blade thereof | |
| CN117196353B (en) | Environmental pollution assessment and monitoring method and system based on big data | |
| CN117195137B (en) | Rotor die casting error detecting system based on data analysis | |
| EP4100646B1 (en) | Wake monitoring, wake management and sensory arrangements to such | |
| CN112418306B (en) | Gas turbine compressor fault early warning method based on LSTM-SVM | |
| CN117407734B (en) | Cylinder tightness detection method and system | |
| CN116950917A (en) | Large centrifugal compressor blade crack online diagnosis method based on blade tip timing | |
| CN111980871A (en) | Wind driven generator sensor state diagnosis system based on PS0-ANFIS | |
| CN117609739B (en) | Structure on-line monitoring method based on multi-point deformation data joint analysis | |
| CN117905651A (en) | Structure monitoring method based on deformation analysis | |
| WO2022064038A1 (en) | Method and system for wind speed determination using vibration data | |
| CN110578659A (en) | System and method for processing SCADA data of wind turbine generator | |
| CN112067289A (en) | Motor shaft and transmission shaft abnormal vibration early warning algorithm based on neural network | |
| US11339763B2 (en) | Method for windmill farm monitoring | |
| CN113807693B (en) | Wind turbine generator generating capacity optimization effect evaluation method based on airborne radar | |
| Gu et al. | A prediction method of operation trend for large axial-flow fan based on vibration-electric information fusion | |
| CN114033736A (en) | Fan fault monitoring system based on air pressure pulse signals | |
| CN113933012B (en) | Propeller rotating speed measuring method based on K-means clustering | |
| JP7125518B2 (en) | Abnormality diagnosis support method and abnormality diagnosis support device for multivariate data | |
| CN112001234B (en) | Rotary mechanical sampling data clustering method based on affine invariant measurement | |
| CN116341993B (en) | State monitoring method and system for polystyrene production process | |
| CN117874687B (en) | Data interaction method of industrial tablet personal computer | |
| CN109946057B (en) | Wind turbine blade damage diagnosis method based on natural frequency | |
| CN115492727A (en) | Method and system for monitoring icing state of blades of wind driven generator | |
| CN116680659A (en) | Rotary mechanical variable rotation speed state monitoring method based on probability distribution slicing measurement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |