Search

CN-121978205-A - Ship structure crack positioning and feature reconstruction method and system

CN121978205ACN 121978205 ACN121978205 ACN 121978205ACN-121978205-A

Abstract

The invention discloses a method and a system for positioning and reconstructing cracks of a ship structure, wherein the method comprises the steps of arranging a sensor array in a region to be detected, transmitting and collecting guided wave signals, carrying out domain transformation, mode screening and matched frequency window filtering on original signals to obtain enhanced signal data, carrying out multi-mode feature extraction on the enhanced signals, constructing multi-dimensional feature vectors containing mode features and coupling relations thereof, realizing preliminary crack positioning by combining sensor position information, further carrying out time inversion and wave field counter propagation on the enhanced signals, constructing an energy field, realizing accurate crack positioning according to an energy field extremum, and extracting crack direction and size information.

Inventors

  • LI YONGZHENG
  • LIU BING

Assignees

  • 江苏科技大学

Dates

Publication Date
20260505
Application Date
20251226

Claims (10)

  1. 1. The method for positioning and reconstructing the crack of the ship structure is characterized by comprising the following steps: Arranging a sensor array in a region to be detected of a detected object, and recording position information of each sensor in the sensor array; Transmitting guided wave signals through the sensor array and collecting response signals to obtain original signal data; performing domain transformation and screening of a plurality of target modes on the original signal data, and performing filtering processing by utilizing a frequency window matched with each screened target mode to obtain enhanced signal data; Carrying out multi-mode feature extraction on the enhanced signal data to construct a multi-dimensional feature vector comprising each target mode feature and inter-mode coupling features, wherein each target mode feature comprises peak amplitude, arrival time and group velocity change rate of each target mode, and each inter-mode coupling feature comprises phase difference and mode conversion energy ratio between each target mode; according to the multidimensional feature vector and the position information of the sensor, solving and obtaining a preliminary positioning result of the crack; Performing time inversion processing and wave field counter propagation calculation on the enhanced signal data, constructing an energy field, obtaining an accurate positioning result of the crack according to an extreme point of the energy field, and performing contrast verification by combining the initial positioning result to confirm a final positioning result; And obtaining the direction information and the size information of the crack according to the spatial distribution characteristics of the energy field around the accurate positioning result, and finishing characteristic reconstruction.
  2. 2. The method for locating and reconstructing a crack of a ship structure according to claim 1, wherein the performing domain transformation and screening of a target pattern on the original signal data, and performing filtering processing by using a frequency window matched with the screened target pattern, to obtain enhanced signal data, comprises: Performing Fourier transformation on the original signal data along the time dimension and the space dimension respectively to obtain frequency-wave number domain signal distribution; Comparing and matching the frequency-wave number domain signal distribution with a theoretical dispersion curve of a measured object, identifying each guided wave mode and energy distribution thereof, and screening out a target mode according to displacement field distribution characteristics, scattering energy ratio and mode separation degree of each guided wave mode; according to the dispersion characteristics of each target mode, respectively determining a concentrated frequency range corresponding to each target mode, and constructing a band-pass frequency window in each concentrated frequency range, wherein the concentrated frequency range is a range in which the proportion of the energy corresponding to the current target mode to the total energy of the whole frequency band is more than 0.7, and the ratio of the distance between the energy corresponding to the current target mode and the wave number of the current target mode is more than 0.15; and filtering the original signal data by utilizing each band-pass frequency window to obtain signal components corresponding to each target mode, wherein the signal components corresponding to each target mode form the enhanced signal data.
  3. 3. The method for locating and reconstructing the crack of the ship structure according to claim 2, wherein the screening the target mode according to the displacement field distribution characteristic, the scattering energy ratio and the mode separation degree of each guided wave mode comprises: In the frequency-wave number domain signal distribution, calculating the ratio of the energy of each guided wave mode in a time window corresponding to the crack reflected wave to the direct wave energy, defining the ratio as a scattering energy ratio, and selecting the guided wave mode with the scattering energy ratio higher than a preset energy ratio threshold as a candidate target mode; and evaluating the mode separation degree between each candidate target mode and other modes in the frequency-wave number domain, wherein the mode separation degree is defined as the ratio of the wave number difference value of the adjacent modes under the same frequency to the average wave number, and selecting the mode with the mode separation degree larger than a preset separation threshold value as the target mode.
  4. 4. The method for locating and reconstructing a crack in a marine structure according to claim 1, wherein the performing multi-mode feature extraction on the enhanced signal data to construct a multi-dimensional feature vector including each target mode feature and inter-mode coupling feature comprises: Envelope extraction is respectively carried out on the signal components of each target mode in the enhanced signal data, and the peak amplitude and the arrival time of each target mode are obtained; Calculating the actual measured group velocity of each target mode according to the signal acquisition point distance and the direct wave arrival time, and comparing the actual measured group velocity with the theoretical group velocity to obtain the group velocity change rate of each target mode; respectively carrying out instantaneous phase extraction on signal components of each target mode, and calculating phase differences among the target modes at crack reflection waves; calculating the ratio of the mode conversion energy to the total reflection energy between the target modes to obtain the mode conversion energy ratio; And integrating the peak amplitude, the arrival time, the group velocity change rate of each target mode, the phase difference between each target mode and the mode conversion energy ratio to construct the multidimensional feature vector.
  5. 5. The method for locating and reconstructing the crack of the ship structure according to claim 1, wherein the step of obtaining the preliminary locating result of the crack by solving according to the multidimensional feature vector and the position information of the sensor comprises the following steps: Correcting the theoretical group velocity of each target mode according to the group velocity change rate in the multidimensional feature vector to obtain corrected group velocity; Calculating the propagation path length of the guided wave from the sensor transmitting signal point to the receiving point through crack reflection according to the arrival time of each target mode in the multidimensional feature vector and the corrected group velocity; Establishing an elliptic equation with crack position coordinates as an unknown number according to the propagation path length and the position information of each sensor; According to the phase difference between each target mode in the multidimensional feature vector, eliminating data with the phase difference exceeding a preset phase difference threshold; selecting mode combinations with the mode conversion energy ratio higher than a preset energy ratio threshold to participate in positioning calculation according to the mode conversion energy ratio in the multidimensional feature vector; And establishing an overdetermined equation set by utilizing a plurality of groups of propagation path lengths, constructing a weight coefficient according to the peak amplitude of each target mode in the multidimensional feature vector, performing normalization calculation by dividing the peak amplitude of each propagation path by the maximum value of the peak amplitudes of all propagation paths, obtaining a value range of 0 to 1, and solving the overdetermined equation set by adopting a weighted optimization algorithm to obtain a preliminary positioning result of the crack.
  6. 6. The method for locating and reconstructing the crack of the ship structure according to claim 1, wherein the performing time reversal processing and wave field counter propagation calculation on the enhanced signal data, constructing an energy field, obtaining an accurate locating result of the crack according to an extreme point of the energy field, performing contrast verification in combination with a preliminary locating result, and confirming a final locating result comprises: The sampling value sequences of all signal channels in the enhanced signal data are arranged in a reverse order according to a time axis, so that a time reversal signal is obtained; inputting the time reversal signal as a boundary condition into a wave equation numerical model, and calculating the back propagation process of the guided wave in the measured object to obtain displacement field distribution at each moment; performing cross-correlation processing on counter-propagating sound fields of all sensor channels, integrating products of sound fields of different channels along a time axis, and constructing a cross energy field as the energy field; extracting coordinates of an energy maximum point in the energy field to obtain an accurate positioning result of the crack; And comparing and verifying the accurate positioning result with the preliminary positioning result, and if the deviation of the accurate positioning result and the preliminary positioning result is within a preset threshold range, confirming that the accurate positioning result is a final positioning result.
  7. 7. The method for locating and reconstructing the crack of the ship structure according to claim 1, wherein the obtaining the direction information and the size information of the crack according to the spatial distribution characteristics of the energy field around the accurate locating result comprises: Taking the accurate positioning result as a center, and extracting an energy contour line of the energy field in a local area of a preset range; performing principal component analysis on the coordinate point set in the energy contour line, and obtaining crack direction information according to the direction of the principal feature vector; And extracting an energy section along the direction corresponding to the crack direction information, and obtaining crack size information according to the widening range of the energy section and a preset calibration relation.
  8. 8. The method for locating and reconstructing a crack of a ship structure according to claim 7, wherein the analyzing the principal component of the coordinate point set in the energy contour line, obtaining crack direction information according to a direction of a principal feature vector, and completing feature reconstruction, comprises: Extracting coordinates of all points in the energy contour line, and calculating the mass center of a coordinate set; performing decentration treatment on the coordinate set by taking the centroid as an origin to construct a covariance matrix; Performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues and corresponding eigenvectors; and determining a feature vector corresponding to the maximum feature value as a main feature vector, and obtaining the crack direction information according to an included angle between the main feature vector and a reference axis.
  9. 9. The method for locating and reconstructing a crack of a ship structure according to claim 7, wherein the extracting an energy profile along a direction corresponding to the crack direction information, and obtaining crack size information according to a widening range of the energy profile and a preset calibration relation, comprises: taking a section line passing through the accurate positioning result in the energy field along the direction corresponding to the crack direction information to obtain a one-dimensional energy section; Determining two position points at which the energy value is reduced to a peak value preset proportion on the one-dimensional energy section, and calculating the distance between the two position points as a widening range; And converting the widening range into crack size information according to a calibration coefficient obtained by calibrating a crack sample with a known size in advance.
  10. 10. A marine vessel structural crack locating and feature reconstruction system, comprising: The sensor array is arranged in a region to be detected of the detected object and is used for transmitting guided wave signals and collecting response signals; the signal acquisition module is used for recording the position information of each sensor in the sensor array and acquiring response signals acquired by the sensor array to form original signal data; The signal enhancement module is used for carrying out Fourier transformation on the original signal data along the time dimension and the space dimension respectively to obtain frequency-wave number domain signal distribution, carrying out contrast matching on the frequency-wave number domain signal distribution and a theoretical dispersion curve of a measured object so as to identify and screen target modes, constructing a band-pass frequency window according to the dispersion characteristics of each target mode, and carrying out filtering processing by utilizing the band-pass frequency window to obtain enhanced signal data; The feature extraction module is used for carrying out multi-mode feature extraction on the enhanced signal data, obtaining peak amplitude, arrival time and group velocity change rate of each target mode and phase difference and mode conversion energy ratio between each target mode, and constructing a multi-dimensional feature vector containing the feature of each target mode and the coupling feature between modes; The preliminary positioning module is used for obtaining a preliminary positioning result of the crack by establishing an elliptic equation set and solving by adopting a weighted optimization algorithm according to the arrival time, the corrected group velocity, the peak amplitude, the phase difference and the mode conversion energy ratio in the multidimensional feature vector and the position information of the sensor; the accurate positioning module is used for carrying out time inversion processing and wave field counter-propagation calculation on the enhanced signal data, constructing an energy field through cross correlation processing, obtaining an accurate positioning result of a crack according to an extreme point of the energy field, carrying out contrast verification by combining with the initial positioning result, and confirming a final positioning result; And the characteristic reconstruction module is used for obtaining the direction information and the size information of the crack through principal component analysis and energy section widening calculation according to the spatial distribution characteristics of the energy field around the accurate positioning result, and finishing characteristic reconstruction.

Description

Ship structure crack positioning and feature reconstruction method and system Technical Field The invention belongs to the technical field of structural health monitoring, and particularly relates to a method and a system for positioning cracks and reconstructing characteristics of a ship structure. Background In the long-term running process of the ship, the steel structure of the ship can be influenced by factors such as wave impact, fatigue load, corrosion and abrasion and the like, so that cracks with different scales are generated. If the crack position and the development trend cannot be found in time, the structural strength is possibly reduced and even serious accidents occur. In the prior art, the common ultrasonic detection method for the ship structure mainly comprises conventional ultrasonic straight probe detection, phased array ultrasonic detection and the like. The method has good effect in local areas such as flat plates, welding lines and the like, but conventional ultrasonic is short-distance spot inspection, and coverage inspection of large ships is difficult. The ship structure comprises components such as reinforcing ribs, cabin partitions and the like, reflection and mode conversion are unavoidable, so that waveform signals are complicated, and accurate positioning is difficult to achieve through a conventional method. The crack length, the inclination angle and the opening degree are different, so that the signal characteristics can be obviously influenced, and the positioning stability is not high. In order to solve the problems of large-scale and complex working conditions, ultrasonic guided wave detection becomes a research hot spot. The guide wave can be spread in the structure in a long distance, which is beneficial to realizing large-scale monitoring. However, the existing guided wave method has the defects of multiple guided wave modes and complex propagation paths, reflected signals are difficult to analyze due to multipath superposition, and the positions and the sizes of cracks are difficult to accurately invert through single reflected wave. Disclosure of Invention The invention aims to provide a ship structure crack positioning and characteristic reconstruction method and system capable of improving crack positioning accuracy. The technical scheme is that the ship structure crack positioning and characteristic reconstruction method comprises the following steps: Arranging a sensor array in a region to be detected of a detected object, and recording position information of each sensor in the sensor array; Transmitting guided wave signals through the sensor array and collecting response signals to obtain original signal data; performing domain transformation and screening of a plurality of target modes on the original signal data, and performing filtering processing by utilizing a frequency window matched with each screened target mode to obtain enhanced signal data; Carrying out multi-mode feature extraction on the enhanced signal data to construct a multi-dimensional feature vector comprising each target mode feature and inter-mode coupling features, wherein each target mode feature comprises peak amplitude, arrival time and group velocity change rate of each target mode, and each inter-mode coupling feature comprises phase difference and mode conversion energy ratio between each target mode; according to the multidimensional feature vector and the position information of the sensor, solving and obtaining a preliminary positioning result of the crack; Performing time inversion processing and wave field counter propagation calculation on the enhanced signal data, constructing an energy field, obtaining an accurate positioning result of the crack according to an extreme point of the energy field, and performing contrast verification by combining the initial positioning result to confirm a final positioning result; And obtaining the direction information and the size information of the crack according to the spatial distribution characteristics of the energy field around the accurate positioning result, and finishing characteristic reconstruction. Further, the filtering processing is performed on the original signal data by using a frequency window matched with the screened target mode to obtain enhanced signal data, including: Performing Fourier transformation on the original signal data along the time dimension and the space dimension respectively to obtain frequency-wave number domain signal distribution; Comparing and matching the frequency-wave number domain signal distribution with a theoretical dispersion curve of a measured object, identifying each guided wave mode and energy distribution thereof, and screening out a target mode according to displacement field distribution characteristics, scattering energy ratio and mode separation degree of each guided wave mode; according to the dispersion characteristics of each target mode, respectively determining a concentrated