CN-121976917-A - Fan tower drum shaking mode real-time detection method, storage medium and program product

CN121976917ACN 121976917 ACN121976917 ACN 121976917ACN-121976917-A

Abstract

The invention discloses a method for detecting shaking modes of a fan tower in real time, a storage medium and a program product. The method comprises the steps of obtaining triaxial motion data of multiple monitoring layers of the tower in real time, calculating theoretical displacement based on the height and the shaking angle of a sensor, dynamically fusing the theoretical displacement with the actually measured displacement, calculating the maximum value of each layer in a sliding time window, generating shaking strength sequences representing spatial distribution characteristics of the tower in a cylindrical mode through normalization processing, matching the sequences with a pre-stored reference mode database to identify dominant mode orders, and triggering an alarm when the duration of a high-order mode exceeds a threshold value. According to the invention, the detection reliability is improved through displacement fusion, the noise interference is suppressed by the self-adaptive time window, the transverse comparison of the cross-specification tower is realized through normalization processing, the mode order is automatically identified through mode matching, and the structural damage is effectively prevented.

Inventors

LUO WENYAN
HE XIAOYUN
LI LINBO
YANG QI
GUO FU
Ma Kuichao
CHEN CHONG
XU JUNFENG
CHEN YINPENG
ZHANG CHENGJIANG
YANG JINGXIAN
Cui Zeqian
TAN LIANGHUI

Assignees

华电(云南)新能源发电有限公司
华电电力科学研究院有限公司
武汉智原科技有限责任公司

Dates

Publication Date: 20260505
Application Date: 20251203

Claims (10)

1. A real-time detection method for a shaking mode of a fan tower is characterized by comprising the following steps: s1, three-axis motion data of a plurality of sensors distributed on different monitoring layers of a tower are obtained in real time, wherein the data comprise a shaking angle, shaking acceleration and shaking displacement; s2, calculating theoretical shaking displacement based on the height of each sensor and the shaking angle of each sensor, and dynamically fusing the theoretical displacement with the actually measured displacement of the sensor at the corresponding position; S3, calculating the maximum value of the displacement of each monitoring layer after fusion in a sliding time window containing a plurality of complete shaking periods; S4, dividing the displacement maximum value of each monitoring layer by the total height of the tower barrel to generate a normalized shaking strength sequence, wherein the sequence represents the spatial distribution characteristic of the deformation of the tower barrel; s5, performing pattern matching on the normalized shaking intensity sequence and a pre-stored reference mode database, and identifying the order of the current dominant shaking mode; and S6, when the continuous duration of the high-order mode is recognized to exceed the preset threshold, generating an alarm signal for triggering the fan operation parameter adjustment.
2. The method for detecting the shaking mode of the tower drum of the fan in real time according to claim 1, wherein at least three monitoring layers are arranged along the height direction of the tower drum, wherein the highest monitoring layer is arranged adjacent to the tower top, each monitoring layer comprises a plurality of sensors uniformly distributed in the circumferential direction, and the distance between the highest monitoring layer and the tower top is smaller than a preset proportion threshold value of the total height Yu Datong.
3. The method for detecting the shaking mode of the tower drum of the fan in real time according to claim 1, wherein the dynamic fusion process is realized by a state estimation algorithm, and the formula is as follows: ; the optimal estimation value of the fusion displacement at the kth moment; Is a displacement predicted value calculated based on the sensor height and the shaking angle, namely theoretical shaking displacement, Is an observation vector formed by actually measuring displacement by a sensor; Representing an observation relation matrix between the displacement predicted value and the actual measurement value; The value of the fusion weight coefficient is determined by the noise characteristic of the system.
4. The method for detecting the shaking mode of the tower drum of the fan in real time according to claim 1, wherein the length range of the sliding time window is 2-5 shaking periods.
5. The method for detecting the shaking mode of the tower drum of the fan in real time according to claim 1, wherein calculating the maximum value of the fused displacement of each monitoring layer comprises: according to the current dominant shaking frequency characteristic, the time window T is adaptively divided into a plurality of continuous analysis time periods; In each analysis period, calculating local maxima of fusion displacement of each monitoring layer in parallel; Taking the maximum value of the local maximum value of each monitoring layer in all sub-windows as the displacement maximum value of the layer in the time window; And if the displacement maximum value of a certain monitoring layer is smaller than the noise threshold value set based on the displacement fluctuation characteristic of the monitoring layer, setting the maximum value to zero, wherein the dynamic window segmentation is satisfied that the duration of each analysis period does not exceed the preset proportion upper limit of the basic shaking period.
6. The method of claim 1, wherein dividing the displacement maxima of each monitoring layer by the total tower height to generate a normalized shake intensity sequence comprises: Acquiring the actual installation height of each monitoring layer to the bottom of the tower barrel to form a height vector h= [ h 1 ,h 2 ,…,h N ] T ; inputting displacement maximum value of each monitoring layer For each monitoring layer: Wherein the method comprises the steps of For the total height of the tower drum, Is a stiffness correction factor that is highly correlated; And outputting a normalized shaking intensity sequence S= [ S 1 ,S 2 ,…,S N ] T ].
7. The method for detecting the vibration mode of the tower drum of the fan according to claim 1, wherein the reference mode database is obtained through finite element simulation and actual measurement of health state and comprises reference transfer functions of modes of each order, and reference transfer function vectors are expressed as , And the normalized reference shaking intensity of the nth order mode at the ith monitoring layer is represented.
8. The method for detecting the vibration mode of the tower drum of the fan according to claim 1, wherein the step of performing pattern matching on the normalized vibration intensity sequence and a pre-stored reference mode database to identify the order of the currently dominant vibration mode comprises the steps of: Extracting a reference transfer function vector R (n) corresponding to the current identification order from the pre-stored reference mode database; Where n represents the modal order index, Is a column vector of dimension N x 1, Representing the normalized reference shaking intensity of the nth order mode at the ith monitoring layer; normalized shaking intensity sequence generated in S4 Performing spatial alignment according to the monitoring layers and R (n) , wherein S i represents a real-time normalized shaking intensity value of an ith monitoring layer, and N represents the number of the monitoring layers arranged along the height direction of the tower; Calculating a matching degree index M of S and R (n) through a vector space similarity algorithm; When the matching degree index M exceeds the mode confirmation threshold, the current dominant mode is judged to be the nth order.
9. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and the computer program when executed by a computer causes the computer to perform the fan tower vibration mode real-time detection method according to any one of claims 1 to 8.
10. A computer program product comprising computer readable code, or a non-transitory computer readable storage medium carrying computer readable code, which when run in a processor of an electronic device, performs the fan tower sway mode real-time detection method of any one of claims 1 to 8.

Description

Fan tower drum shaking mode real-time detection method, storage medium and program product Technical Field The invention relates to the technical field of wind power generation equipment state monitoring, in particular to a real-time detection method, a storage medium and a program product for a shaking mode of a fan tower. Background The shaking mode is a natural motion mode which is presented when the fan is periodically shaken under the action of external excitation, the shaking mode of the high-order mode is harmful shaking, and the fan tower barrel is damaged when the fan tower barrel is continuously shaken in the high-order mode for a long time. In the normal power generation process, the fan can be excited by various external excitation, and the fan tower can swing in various different modes. Disclosure of Invention In view of the technical defects and technical drawbacks existing in the prior art, embodiments of the present invention provide a method for detecting a shaking mode of a tower drum of a fan in real time, a storage medium, and a program product, which overcome or at least partially solve the above problems, and specifically, the following schemes are provided: As a first aspect of the invention, a method for detecting shaking modes of a tower drum of a fan in real time is provided, which comprises the following steps: s1, three-axis motion data of a plurality of sensors distributed on different monitoring layers of a tower are obtained in real time, wherein the data comprise a shaking angle, shaking acceleration and shaking displacement; s2, calculating theoretical shaking displacement based on the height of each sensor and the shaking angle of each sensor, and dynamically fusing the theoretical displacement with the actually measured displacement of the sensor at the corresponding position; S3, calculating the maximum value of the displacement of each monitoring layer after fusion in a sliding time window containing a plurality of complete shaking periods; S4, dividing the displacement maximum value of each monitoring layer by the total height of the tower barrel to generate a normalized shaking strength sequence, wherein the sequence represents the spatial distribution characteristic of the deformation of the tower barrel; s5, performing pattern matching on the normalized shaking intensity sequence and a pre-stored reference mode database, and identifying the order of the current dominant shaking mode; and S6, when the continuous duration of the high-order mode is recognized to exceed the preset threshold, generating an alarm signal for triggering the fan operation parameter adjustment. Further, at least three monitoring layers are arranged along the height direction of the tower, wherein the highest monitoring layer is arranged adjacent to the tower top, each monitoring layer comprises a plurality of sensors which are uniformly distributed in the circumferential direction, and the distance between the highest monitoring layer and the tower top is smaller than a preset proportion threshold value of Yu Datong total height. Further, the dynamic fusion process is implemented by a state estimation algorithm, and the formula is as follows: ; the optimal estimation value of the fusion displacement at the kth moment; Is a displacement predicted value calculated based on the sensor height and the shaking angle, namely theoretical shaking displacement, Is an observation vector formed by actually measuring displacement by a sensor; Representing an observation relation matrix between the displacement predicted value and the actual measurement value; The value of the fusion weight coefficient is determined by the noise characteristic of the system. Further, the sliding time window is 2-5 shaking periods in length. Further, calculating the maximum value of the fused displacement of each monitoring layer comprises: according to the current dominant shaking frequency characteristic, the time window T is adaptively divided into a plurality of continuous analysis time periods; In each analysis period, calculating local maxima of fusion displacement of each monitoring layer in parallel; Taking the maximum value of the local maximum value of each monitoring layer in all sub-windows as the displacement maximum value of the layer in the time window; And if the displacement maximum value of a certain monitoring layer is smaller than the noise threshold value set based on the displacement fluctuation characteristic of the monitoring layer, setting the maximum value to zero, wherein the dynamic window segmentation is satisfied that the duration of each analysis period does not exceed the preset proportion upper limit of the basic shaking period. Further, dividing the displacement maxima of each monitoring layer by the total tower height to generate a normalized sloshing intensity sequence includes: Acquiring the actual installation height of each monitoring layer to the bottom of the tower barrel to form a