Search

CN-122014534-A - Fan tower natural frequency abnormality identification method, storage medium and electronic equipment

CN122014534ACN 122014534 ACN122014534 ACN 122014534ACN-122014534-A

Abstract

The invention provides a method for identifying natural frequency abnormality of a fan tower, a storage medium and electronic equipment, belonging to the technical field of state monitoring and fault diagnosis of wind power generation equipment, wherein the method comprises the following steps of S1, collecting acceleration signals of the fan tower, and preprocessing the acceleration signals to eliminate direct current components; the method comprises the steps of S2, carrying out grouping processing on preprocessed acceleration signals, sequentially carrying out frequency spectrum analysis on each group of acceleration signals, sequentially identifying alternative frequencies of each group in a preset frequency band, S3, storing the continuously obtained alternative frequencies by adopting a sliding window mechanism, removing abnormal values in a window based on a statistical abnormality detection method, and calculating to obtain natural frequencies, S4, comparing the natural frequencies with a reference natural frequency, judging the health state of a tower structure according to whether the relative error exceeds a preset threshold value, and triggering corresponding alarms. The invention overcomes the defects of the traditional peak value picking method, and the reliability of the result obtained by real-time monitoring is higher.

Inventors

  • CHEN JINSONG
  • MING YOULIN
  • YU QIANYU
  • LIU ZIYU
  • FENG WENYUE
  • CHANG KUN

Assignees

  • 武汉智原科技有限责任公司

Dates

Publication Date
20260512
Application Date
20260302

Claims (10)

  1. 1. The method for calculating the natural frequency of the fan tower and identifying the abnormality is characterized by comprising the following steps of: S1, acquiring acceleration signals of a fan tower, and preprocessing the acceleration signals to eliminate direct current components; S2, grouping the preprocessed acceleration signals, sequentially performing spectrum analysis on each group of acceleration signals, and sequentially identifying alternative frequencies of each group in a preset frequency band; S3, storing the continuously obtained alternative frequencies by adopting a sliding window mechanism, and calculating to obtain the natural frequency after eliminating abnormal values in the window based on a statistical abnormality detection method; S4, comparing the natural frequency with a reference natural frequency, judging the health state of the tower structure according to whether the relative error exceeds a preset threshold value, and triggering a corresponding alarm.
  2. 2. The method according to claim 1, wherein the acquiring of the acceleration signal in step S1 specifically comprises: And a triaxial acceleration sensor is arranged at the top of the fan tower, and the sampling frequency is set to be 10Hz to collect acceleration signals.
  3. 3. The method according to claim 1, wherein step S2 specifically comprises: S21, filtering a direct current component in the acceleration signal by adopting a high-pass filter with the cutoff frequency of 0.1 Hz; S22, carrying out fast Fourier transform calculation on each group of data by taking every 2048 continuous data as a group on the acceleration signal with the direct current component filtered; S23, selecting the frequency with the largest peak value in the frequency range of 0-0.5 Hz as the alternative frequency of each group based on the frequency spectrum information corresponding to each group of data.
  4. 4. The method according to claim 1, wherein step S3 specifically comprises: S31, establishing a fixed-length sliding window for storing the continuously acquired alternative frequency sequences; S32, detecting abnormal values of alternative frequencies in the window by adopting a quarter bit distance method, and identifying and eliminating non-natural frequency components generated by fan working condition interference; S33, calculating an arithmetic average value of the alternative frequencies passing through the abnormality detection to obtain the natural frequency reflecting the structural state of the tower.
  5. 5. The method of claim 4, wherein the sliding window has a length of 421.
  6. 6. The method of claim 4, wherein the outlier detection of the candidate frequencies within the window using a quarter-bit method specifically comprises: ordering the alternative frequencies in the sliding window according to the values to obtain an ordered sequence ; Calculating the median of the ordered sequences And by For reference, calculate slave To the point of The median of (2) is the first quartile From the slave To the point of The median of (2) is the third quartile ; According to the formula Calculating a quartile range; Setting the lower bound of abnormal value detection as The upper boundary is ; And judging the alternative frequencies with the values exceeding the lower and upper ranges as abnormal values and eliminating the abnormal values.
  7. 7. The method according to claim 1, wherein determining the health status of the tower structure in step S4 specifically comprises: S41, setting the natural frequency calculated in the mode of the step S3 as a reference natural frequency when the sliding window stores the full candidate frequency for the first time ; S42, after each new alternative frequency is obtained, updating the window and recalculating the corresponding natural frequency in the mode of step S3 to obtain the real-time natural frequency ; S43, according to the formula And calculating a relative error, and when the relative error exceeds a preset safety threshold, judging that the tower structure is abnormal and triggering an inspection alarm.
  8. 8. The method according to claim 7, further comprising a real-time update mechanism, comprising in particular: At the completion of the reference natural frequency After the setting of (2), starting a real-time update flow: after each group of acceleration signals are collected and a new alternative frequency is calculated, updating operation is carried out on the sliding window: firstly removing the earliest entering alternative frequency in the window; then, moving all the remaining alternative frequencies to the left by one bit according to the time sequence; Finally, adding the alternative frequency obtained by the current latest calculation into the tail end of the window; Based on the updated alternative frequencies in the window, calculating the natural frequency again according to the mode of S3 to obtain the natural frequency as the real-time natural frequency 。
  9. 9. A computer readable medium, characterized in that it has stored thereon a computer program which, when executed by a processor, implements the steps of the method according to any of claims 1-8.
  10. 10. An electronic device, comprising: one or more processors; A memory for storing one or more programs; when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-8.

Description

Fan tower natural frequency abnormality identification method, storage medium and electronic equipment Technical Field The invention relates to the technical field of wind power generation equipment state monitoring and fault diagnosis, in particular to a method for calculating natural frequency of a tower drum of a fan and identifying abnormality, a storage medium and electronic equipment. Background With the shift of global energy structures to green low carbon, wind power generation is an important component of clean energy, and the installed capacity and the single machine scale of the wind power generation continue to rapidly increase. The fan tower is used as a key bearing structure for supporting the wind wheel and the engine room, and the structural health state of the fan tower is directly related to the operation safety and service life of the whole machine. The natural frequency of the tower is a key dynamic characteristic parameter of the tower, not only affects the power coupling between the whole machine and the pneumatic and control system, but also is a sensitive index for evaluating the structural integrity of the tower (such as foundation settlement, weld cracking, barrel corrosion, bolt pretightening force loss and the like). Currently, the peak picking method is mainly used for acquiring and analyzing the natural frequency of a fan tower in the industry. In order to obtain the actual frequency, a high-precision acceleration sensor is usually installed on the tower, and the dominant frequency is identified by collecting vibration response signals under environmental excitation (such as wind load and unit start-stop) or manual excitation and performing spectrum analysis (such as Fourier transformation). The method is relatively direct, but the natural frequency error of real-time monitoring and identification is large, the wind wheel rotating frequency of the general flexible tower is relatively close to the natural frequency of the tower, when the fan is in a working state, the rotor rotates 1P and 3P frequencies and the gear box operates to cause the vibration of the tower, so that signals with amplitude higher than the natural frequency can appear on the frequency spectrum nearby the natural frequency, if the frequency with the largest amplitude is picked up simply by the peak value to be selected as the natural frequency of the wind power tower, and when the fan operates under full load, other frequencies are taken as the natural frequency with probability. 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 identifying abnormal natural frequencies of a fan tower, a storage medium, and an electronic device, which overcome or at least partially solve the above problems, so as to reduce errors generated by different calculation natural frequencies of fan working conditions, and realize evaluation of health status of the fan tower based on the natural frequencies obtained by real-time calculation, and the specific scheme is as follows; As a first aspect of the present invention, there is provided a method for calculating a natural frequency of a fan tower and identifying an abnormality, comprising the steps of: S1, acquiring acceleration signals of a fan tower, and preprocessing the acceleration signals to eliminate direct current components; S2, grouping the preprocessed acceleration signals, sequentially performing spectrum analysis on each group of acceleration signals, and sequentially identifying alternative frequencies of each group in a preset frequency band; S3, storing the continuously obtained alternative frequencies by adopting a sliding window mechanism, and calculating to obtain the natural frequency after eliminating abnormal values in the window based on a statistical abnormality detection method; S4, comparing the natural frequency with a reference natural frequency, judging the health state of the tower structure according to whether the relative error exceeds a preset threshold value, and triggering a corresponding alarm. In some embodiments, the acquiring the acceleration signal in step S1 specifically includes: And a triaxial acceleration sensor is arranged at the top of the fan tower, and the sampling frequency is set to be 10Hz to collect acceleration signals. In some embodiments, step S2 specifically includes: S21, filtering a direct current component in the acceleration signal by adopting a high-pass filter with the cutoff frequency of 0.1 Hz; S22, carrying out fast Fourier transform calculation on each group of data by taking every 2048 continuous data as a group on the acceleration signal with the direct current component filtered; S23, selecting the frequency with the largest peak value in the frequency range of 0-0.5 Hz as the alternative frequency of each group based on the frequency spectrum information corresponding to each group of data. In some