Power transmission tower modal parameter automatic identification method, device and system

Abstract

The invention discloses a power transmission tower modal parameter automatic identification method, device and system, and relates to the field of power transmission tower structure monitoring. According to the specific implementation scheme, a plurality of measurement points are arranged on a power transmission tower body in the height direction, and response data a are obtained; based on the response data a, a random subspace method is used for solving and drawing a first stability graph, through first modal data between each first stability point and a nearest point of an adjacent order, fuzzy C-means clustering (Fuzzy C-Means Clustering, abbreviated as FCM) is used for clustering the first modal data, false points are deleted, and a second stability graph is obtained; secondly, extracting stable points, meeting conditions, of each order by adopting two-time hierarchical clustering, and averaging to obtain the frequency, damping and vibration mode of each order; according to the method, the modal parameters of the power transmission tower are indirectly solved by using the data collected by the sensors arranged at the multiple measuring points, and the problems that the modal parameter identification efficiency of the power transmission tower is low and the reliability is poor are solved.

Description

Technical Field

[0001] The present invention relates to the field of transmission tower structure monitoring, and more specifically, to a method, device and system for automatically identifying modal parameters of a transmission tower. Background Art

[0002] A transmission tower is a widely used structural form in the power industry. Due to being exposed to the environment of external wind vibration and harmful substance erosion for a long time, its structural safety and durability will gradually degrade with the increase of service life, and sudden accidents may occur under extreme conditions. Therefore, the structural health monitoring of transmission towers is very important for ensuring the safety of the whole life cycle of transmission towers.

[0003] As the core characterization parameters of structural dynamic characteristics, modal parameters can sensitively reflect damage characteristics such as structural stiffness degradation and connection looseness of transmission towers, and thus become important quantitative indicators for evaluating the health status of transmission towers. Existing modal parameter identification methods include traditional modal parameter identification and modal parameter identification under ambient excitation. Among them, traditional modal parameter identification methods use frequency response functions or impulse response functions to identify the modal parameters of structures based on known input and output data. However, due to the difficulty of measuring data of large and complex structures, traditional modal parameter identification methods have many difficulties in practical applications. In contrast, modal parameter identification methods under ambient excitation do not require input data and only use the output data of the structure for identification, which has significant advantages such as no need for artificial excitation, saving time and cost, not damaging the structure and not affecting the normal use of the structure. Therefore, it is more widely used in practical engineering.

[0004] However, existing modal parameter identification methods still require manual intervention during the process of modal parameter identification, and require operators to have certain professional knowledge of modal analysis. However, manual intervention prolongs the data analysis cycle, making it difficult to meet the real-time warning requirements under extreme weather, and the subjective experience differences reduce the reliability of damage identification. At the same time, with the increasing demand for long-term continuous monitoring of transmission towers, the cost required for manual analysis in operational modal analysis is relatively large. Therefore, in the field of health monitoring of transmission towers, there is an urgent need to realize the automatic identification of modal parameters in the analysis of vibration test signals of large and complex structures. Summary of the Invention

[0005] The present invention provides a method, device and system for automatically identifying modal parameters of a transmission tower, which uses the data collected from multiple measurement points set on the transmission tower to automatically identify the modal parameters of the main body of the transmission tower. Through the method for automatically identifying modal parameters of a transmission tower of the present invention, the influence of manual intervention on the identification result is reduced, and the efficiency and reliability of modal parameter identification are improved.

[0006] According to a first aspect of the present invention, there is provided a method for automatically identifying modal parameters of a transmission tower, comprising the following steps:

[0007] Step S1, a plurality of measurement points are arranged on the main body of the transmission tower along the height direction, and response data a of each measurement point is obtained;

[0008] Step S2, based on the response data a, the SSI method (Stochastic Subspace Identification, SSI) is used for solution to obtain a first stability diagram, and the first stability diagram includes first stability points at multiple different orders, and the first stability points include true points and false points;

[0009] Step S3, based on the first stability diagram and the first stability points, the first modal data between each first stability point and its nearest point of the adjacent order are respectively calculated, and the first modal data includes Δf, Δζ, 1-MAC, ΔMPC, ΔMPD;

[0010] Step S4, based on the first modal data, through Fuzzy C-Means Clustering (FCM for short), a second stability diagram and second stability points are obtained;

[0011] Step S5, based on the second stability diagram and the second stability points, through the first hierarchical clustering and the second hierarchical clustering, a third stability diagram and third stability points are obtained;

[0012] Step S6, based on the third stability diagram and the third stability points, a final modal parameter matrix is obtained.

[0013] According to a second aspect of the present invention, there is provided a processing device for automatically identifying modal parameters of a transmission tower, which includes a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the method described in any item of the first aspect is implemented.

[0014] According to a third aspect of the present invention, there is provided a processing system for automatically identifying modal parameters of a transmission tower, which includes a distributed file system and the processing device described in the second aspect; wherein, the distributed file system has a storage server, and the storage server has a memory and a disk.

[0015] According to the technology of the present invention, by installing a plurality of measurement points on the main body of the transmission tower and arranging sensors at the measurement points, it is possible to automatically identify the modal parameters of the transmission tower under ambient excitation. This method is simple and convenient to calculate, and has strong operability.

[0016] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. Description of the Drawings

[0017] Figure 1 It is a schematic flow chart of a method for automatically identifying modal parameters of a transmission tower in an embodiment of the present invention.

[0018] Figure 2 It is a layout diagram of measurement points of a transmission tower in an embodiment of the present invention, where "○" represents measurement points.

[0019] Figure 3 It is the first stability diagram in an embodiment of the present invention.

[0020] Figure 4 It is the second stability diagram in an embodiment of the present invention.

[0021] Figure 5 It is a probability density function curve in an embodiment of the present invention.

[0022] Figure 6 It is the third stability diagram in an embodiment of the present invention. Detailed Description of the Embodiments

[0023] To further understand the content of the present invention, the present invention will be described in detail in combination with embodiments. It should be understood that the embodiments are only for explaining the present invention rather than limiting it.

[0024] In order to solve the above problems, the present invention provides a method for automatically identifying modal parameters of a transmission tower to achieve automatic identification of modal parameters of a transmission tower without manual intervention.

[0025] Figure 1 It is a schematic flow chart of a method for automatically identifying modal parameters of a transmission tower in an embodiment of the present invention, as Figure 1 shown.

[0026] Step S1, set a plurality of measurement points on the main body of the transmission tower along the height direction, and obtain the response data a of each measurement point.

[0027] In this step, as Figure 2 shown, it is assumed that D measurement points are set on the main structure of the transmission tower along the height direction, and sensors including velocity sensors, acceleration sensors, and displacement sensors are set at the measurement points. The sensors set can be one or more of them to obtain the response data a. Among them, D is not less than 3, and the measurement points are evenly distributed. For accurate calculation, at least 2000 steps of response are extracted at each measurement point. Assuming that 5 measurement points are set, 6000 steps of response are extracted at each measurement point, and an input matrix b of 6000×5 is obtained.

[0028] Step S2: Based on the response data a, solve it by the SSI method to obtain the first stability diagram. The first stability diagram includes first stability points at multiple different orders. The first stability points include true points and false points.

[0029] Use the SSI method (Stochastic Subspace Identification, SSI) to solve the modal parameters for the obtained input matrix b, and draw a stability diagram according to the results. The information contained in each stability point in the stability diagram includes frequency, damping, and mode shape. Assume that the maximum calculation order is set to 100. As Figure 3 shown, obtain the first stability diagram. The first stability diagram contains multiple first stability points. The first stability points include true points and false points. Among them, only retain the stability points with a damping ratio from zero to twenty percent, a frequency from zero to fifty hertz, and the points with complex conjugates as the first stability points:

[0030] 0% < ξ i < 20%

[0031] 0 ≤ f i ≤ 50

[0032]

[0033] In the formula, i represents the i-th stability point, ξ represents the damping ratio, and f represents the frequency. represents having complex conjugates and the imaginary part of the eigenvalue is not zero.

[0034] Step S3: Based on the first stability diagram and the first stability points, calculate and obtain the first modal data between each first stability point and its nearest point of the adjacent order. The first modal data includes Δf, Δζ, 1 - MAC, ΔMPC, and ΔMPD.

[0035] The first modal data can be obtained by the following formula:

[0036]

[0037] In the formula, MAC, respectively represent the frequency difference, damping difference, mode shape similarity, modal phase collinearity change amount, and average phase deviation change amount between the n-th order mode i and the (n + 2)-th order mode m.

[0038]

[0039] Among them, represents transpose

[0040]

[0041] where φ ij is the j-th element of the complex modal vector; w j is the weight of the j-th element, and w j can be taken as φ ij ; Re(·) and Im(·) are respectively the real part and the imaginary part of a complex number; arccos is the inverse cosine function; v 11 and v 21 are respectively the parameters of the V matrix after singular value decomposition of the complex modal vector [Re(φ i ), Im(φ j )].

[0042] Step S4: Based on the first modal data, obtain the second stability diagram and the second stable point through Fuzzy C-Means Clustering (abbreviated as FCM).

[0043] Specifically, this step can be implemented in the following way:

[0044] Perform data preprocessing on the first modal data to obtain the second modal data, including:

[0045] Step S401: Perform BOX-COX transformation on the first modal data:

[0046]

[0047] where h(γ) is the new variable after transformation, p is the original continuous dependent variable, and γ is the transformation parameter.

[0048] Step S402: Standardize the transformed first modal data:

[0049]

[0050] where is the average value, and σ(h T,i (m)) is the standard deviation.

[0051] Step S403: Combine the second modal data in pairs, and through the FCM clustering method (fuzzy C-means clustering), take the point with the lower left corner as the clustering center for each combination of clustering results to obtain multiple first clustering results.

[0052] Specifically, use FCM clustering to classify the modes with high similarity under adjacent model orders in the first stability diagram into one category, and the modes with low similarity into another category. Therefore, the number of clusters C = 2. For any calculated mode p, define its feature as m p , and perform FCM clustering on the feature, that is, minimize the following objective function:

[0053]

[0054] In the formula, i represents the category; ||·|| represents the Euclidean norm of the vector; t represents the fuzzy factor; μ i is the clustering center of the i-th category; η represents the membership function, and the sum of the membership degrees of any clustering feature m p is equal to 1.

[0055] The minimization of the objective function is achieved by iteratively updating the clustering centers and their membership degrees of each category. Among them, the clustering center μ i and the feature m p and the membership function η ip are updated by the following formulas:

[0056]

[0057] For each clustering result, take the point with the lower left corner as the clustering center, and finally take the common point of each clustering result as the real point. As Figure 4 shown, a stable graph after further cleaning is obtained, that is, the second stable graph.

[0058] Step S5, based on the second stable graph and the second stable points, through the first hierarchical clustering and the second hierarchical clustering, obtain the third stable graph and the third stable points.

[0059] Specifically, this step specifically includes:

[0060] Step S510, through the first hierarchical clustering, extract the second stable points with similar frequencies in the second stable graph to obtain the second clustering result.

[0061] Specifically, this step can be specifically implemented in the following way:

[0062] Step S511, through calculate the distance Δf1 between each second stable point in the second stable graph and the distance Δf2 between each second stable point and its adjacent order nearest point.

[0063] Step S512, perform a probability density function curve fitting on Δf2, and select the abscissa value corresponding to the first wave valley on the probability density function curve as the truncation threshold for the first hierarchical clustering of Δf1.

[0064] To achieve the automatic setting of the distance threshold for the frequency index, perform a probability density function curve fitting on Δf2. As Figure 5 shown, since there will be a large number of points with close distances near each stable axis in the stable graph, these points will form peaks and valleys on the probability density function curve. Select the abscissa value corresponding to the first wave valley on the probability density function curve as the truncation threshold of the distance.

[0065] Step S513, by setting the minimum number of elements N that make up the clustering result.

[0066] Among them, α cl is the percentage of the maximum number of elements that can be arranged in a cluster, ord max is the maximum calculation order, ord min is the minimum calculation order, Δord is the change in the calculation order. According to experience, set α cl to 0.25.

[0067] Step S514, delete the second stable points with the number of intra-class elements less than N in all clustering results to obtain the second clustering result.

[0068] Step S520, based on the second clustering result, remove the modal data with a mode shape deviation greater than the truncation threshold from the second clustering result through the second hierarchical clustering to obtain the third stable diagram and the third stable points.

[0069] Specifically, this step can be implemented in the following way:

[0070] Step S521, calculate the modal irrelevance d between each order of modal data in the second clustering result through d = 1 - MAC i,j where MAC i,j is the similarity of the mode shapes of the i-th order mode and the j-th order mode.

[0071] Step S522, based on the modal irrelevance d between each order of modal data, arrange the modal irrelevance d in ascending order, and select the modal irrelevance d at the 90th percentile as the truncation threshold.

[0072] If the MAC value between two modes is large, it indicates that these two modes are relatively close. At this time, d is small. Therefore, select the value at the 90% position of the ascending order of d for each order of mode as the threshold for the second hierarchical clustering.

[0073] Step S523, perform the second hierarchical clustering on the second clustering result based on the truncation threshold, select the stable points with the number of intra-class elements greater than N in the result of the second hierarchical clustering as the true points, retain the true points, and obtain the third stable diagram and the third stable points. Among them, the third stable points in the third stable diagram form two columns of stable axes.

[0074] Select the points with more than N elements in the result of the second hierarchical clustering. These points are the true modes. After screening according to the criterion, the stable diagram has two columns of stable axes. As Figure 6 shown, it is the third stable diagram after screening. The third stable points in the figure form two columns of stable axes

[0075] Step S6: Obtain the final modal parameter matrix based on the third stability diagram and the third stable points.

[0076] Specifically, calculate the average values of the frequencies, damping ratios, and mode shapes corresponding to the third stable points of each order in the finally obtained third stability diagram. The obtained average values represent the final solutions of each order. The frequencies and damping ratios of the final solutions are 1×2 matrices, and the mode shapes are 5×2 matrices.

[0077] In the existing modal identification of transmission towers, due to excessive interference from human subjective factors, the efficiency of modal identification is low, and the reliability of the finally obtained modal parameter identification results is insufficient. In the embodiments of the present disclosure, first, a plurality of sensors are arranged along the height direction of the tower body to obtain response data a, and then the stochastic subspace method is used for solution to obtain the initial stability diagram. For the initial stability diagram, only the stable points with damping ratios from zero to twenty percent and frequencies from zero to fifty hertz, as well as the points with complex conjugates, are retained; then calculate Δf, Δζ, 1 - MAC, ΔMPC, and ΔMPD between each point and the nearest point of the adjacent order; perform BOX - COX transformation and standardization on the obtained data, and use FCM clustering to delete the false modes; then adopt hierarchical clustering to automatically classify the points with similar frequencies in the stability diagram, and only retain the points with more than N points; the second hierarchical clustering automatically re - classifies the stable points with similar mode shapes in each category of the first clustering result, and again retains the points of the categories with more than N points; obtain the stability diagram after cleaning; finally, calculate the average values of the frequencies, damping ratios, and mode shapes corresponding to the stable points of each order to represent the final solutions of each order.

[0078] Through the technology of the present invention, by installing a plurality of sensors on the transmission tower, it is possible to accurately and automatically identify the modal parameters of the transmission tower under ambient excitation, thereby solving the problems of excessive interference from human subjective factors in the modal identification of the transmission tower, low efficiency of modal identification, and poor reliability of the finally obtained modal parameter identification results.

[0079] The second aspect of the present invention provides a processing device for automatically identifying the modal parameters of a transmission tower, which includes a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps of the lattice tower structure wind load identification method described above are implemented.

[0080] The third aspect of the present invention provides a processing system for automatically identifying the modal parameters of a transmission tower, which includes a distributed file system and the processing device described in the second aspect; wherein, the distributed file system has a storage server, and the storage server has a memory and a disk.

[0081] It is easily understandable that those skilled in the art can combine, split, reorganize, etc. the embodiments of the present invention based on one or several embodiments provided by the present invention to obtain other embodiments, and these embodiments do not exceed the protection scope of the present invention.

[0082] The above has schematically described the present invention and its implementation manners. This description is not restrictive. What is shown in the embodiments is only part of the implementation manners of the present invention, and the actual structure is not limited thereto. Therefore, if those of ordinary skill in the art are inspired by it and design, without creative efforts, structural manners and embodiments similar to the technical solution without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. An automatic identification method for modal parameters of a transmission tower, comprising: Setting a plurality of measurement points along the height direction on the main body of the transmission tower to obtain the response data a of each measurement point; Based on the response data a, solving through the SSI method to obtain a first stability diagram, the first stability diagram including a plurality of first stable points at different orders, and the first stable points including real points and false points; Based on the first stability diagram and the first stable points, respectively calculating to obtain first modal data between each first stable point and its nearest point of the adjacent order, the first modal data including Δf, Δζ, 1-MAC, ΔMPC, ΔMPD; Based on the first modal data, obtaining a second stability diagram and second stable points through fuzzy C-means clustering; Based on the second stability diagram and the second stable points, obtaining a third stability diagram and third stable points through the first hierarchical clustering and the second hierarchical clustering; Based on the third stability diagram and the third stable points, obtaining a final modal parameter matrix.

2. The automatic identification method for modal parameters of a transmission tower according to claim 1, wherein the step of, based on the first stability diagram and the first stable points, respectively calculating to obtain first modal data between each first stable point and its nearest point of the adjacent order, and the first modal data including Δf, Δζ, 1-MAC, ΔMPC, ΔMPD, comprises: Calculating the first modal data through the following formula:

3. The automatic identification method for modal parameters of a transmission tower according to claim 1, wherein the step of, based on the first modal data, obtaining a second stability diagram and second stable points through the fuzzy C-means clustering method, comprises: Preprocess the first modal data to obtain the second modal data, including, by Perform a BOX-COX transformation on the data, and then by Standardize the data; Combining the second modal data in pairs, and through the FCM clustering method, taking the point with the lower left corner as the clustering center for the clustering result obtained each time to obtain a plurality of first clustering results; Based on all the first clustering results, taking the common points of each first clustering result as real points, retaining the real points, and obtaining a second stability diagram and second stable points.

4. The automatic identification method for modal parameters of a transmission tower according to claim 3, wherein the step of combining the second modal data in pairs, and through the FCM clustering method, taking the point with the lower left corner as the clustering center for the clustering result obtained each time to obtain a plurality of first clustering results, comprises: Through the FCM clustering method, classifying the modal data of adjacent orders with high similarity in the first stability diagram into the first type of modal data, and classifying the modal data of adjacent orders with low similarity into the second type of modal data, to obtain a plurality of first clustering results, including: For each modality p, define its feature as m p , and perform FCM clustering on the features to minimize the objective function Among them, through and the cluster center μ i , the feature m p and the membership function η ip are updated. By iteratively updating the cluster center μ i , the feature m p and the membership function η ip of each type of modal data, the objective function is minimized.

5. The automatic identification method for modal parameters of a transmission tower according to claim 1, wherein the step of, based on the second stability diagram and the second stable points, obtaining a third stability diagram and third stable points through the first hierarchical clustering and the second hierarchical clustering, comprises: Through the first hierarchical clustering, extracting the second stable points with similar frequencies in the second stability diagram to obtain a second clustering result; Based on the second clustering result, removing the modal data with a mode deviation greater than the truncation threshold in the second clustering result through the second hierarchical clustering to obtain a third stability diagram and third stable points.

6. A method for automatically identifying modal parameters of a transmission tower according to claim 5, wherein the second stable points with similar frequencies in the second stable diagram are extracted through the first hierarchical clustering to obtain a second clustering result, including: By calculating the distance Δf1 between each second stable point in the second stable graph and the distance Δf2 between each second stable point and its adjacent order nearest point; Performing a probability density function curve fitting on Δf2, and selecting the abscissa value corresponding to the first wave trough on the probability density function curve as the truncation threshold for the first hierarchical clustering of Δf1; By setting the minimum number of elements N that make up the clustering result, where α cl is 0.25; Deleting the second stable points with the number of elements within the category less than N in all clustering results to obtain a second clustering result.

7. A method for automatically identifying modal parameters of a transmission tower according to claim 6, wherein based on the second clustering result, the modal data with a mode deviation greater than the truncation threshold in the second clustering result is removed through the second hierarchical clustering to obtain a third stable diagram and third stable points, including: Calculate the modal irrelevance d between the modal data of each order in the second clustering result through d = 1 - MAC i,j where MAC i,j is the similarity of the mode shapes of the i-th order mode and the j-th order mode; Based on the modal irrelevance d between each order of modal data, arranging the modal irrelevance d in ascending order, and selecting the modal irrelevance d at the 90th percentile as the truncation threshold; Performing a second hierarchical clustering on the second clustering result based on the truncation threshold, and selecting the stable points with the number of elements within the category greater than N in the result of the second hierarchical clustering as real points, retaining the real points to obtain a third stable diagram and third stable points, wherein the third stable points in the third stable diagram form two columns of stable axes.

8. A method for automatically identifying modal parameters of a transmission tower according to claim 1, wherein based on the third stable diagram and third stable points, a final modal parameter matrix is obtained, including: Respectively calculating the average values of the frequencies, damping ratios, and mode shapes corresponding to each order of third stable points to obtain a frequency matrix, a damping matrix, and a mode shape matrix for each order.

9. A processing device for automatic identification of modal parameters of a transmission tower, comprising a memory and a processor, wherein a computer program is stored in the memory, and is characterized in that: When the processor executes the computer program, the steps of the processing method according to any one of claims 1 to 8 are implemented.

10. A processing system for automatic identification of modal parameters of transmission towers, characterized in that: Including a distributed file system and the processing device in claim 9; wherein the distributed file system has a storage server, and the storage server has a memory and a disk.

Images