CN-120372333-B - Automatic bridge modal parameter identification method and system considering multi-channel information

CN120372333BCN 120372333 BCN120372333 BCN 120372333BCN-120372333-B

Abstract

The invention relates to an automatic recognition method and system for bridge modal parameters considering multichannel information, comprising the following steps of S1, directly analyzing multichannel monitoring signals by using COV-SSI to generate a stable graph, automatically extracting a stable axis by combining a DBSCAN clustering algorithm to realize automatic recognition of modal frequencies, S2, carrying out signal decomposition on multichannel monitoring data by using MvFIF to generate an Intrinsic Mode Function (IMF) group with modal alignment characteristics, S3, calculating the instantaneous frequency and bandwidth of the IMF in the step S2 by using HHT, screening out IMF components containing target modal frequencies and carrying out linear reconstruction, S4, taking the multichannel monitoring data after the reconstruction in the step S3 as the system input of the COV-SSI algorithm, calculating a system matrix and an output matrix, and calculating the modal damping ratio by using eigenvalue decomposition.

Inventors

JIANG YAN
LONG TAO
XIN JINGZHOU
CUI XIAOLEI
ZHANG HONG
WU FENGBO
PENG LIULIU
LI SHAOPENG
TANG QIZHI

Assignees

重庆交通大学

Dates

Publication Date: 20260505
Application Date: 20250513
Priority Date: 20250415

Claims (8)

1. The automatic bridge modal parameter identification method considering the multi-channel information is characterized by comprising the following steps of: S1, directly analyzing a multichannel monitoring signal by using COV-SSI to generate a stability diagram, and automatically extracting a stability axis by combining a DBSCAN clustering algorithm to realize automatic identification of modal frequency; S2, carrying out signal decomposition on the multichannel monitoring data by adopting MvFIF to generate an intrinsic mode function set with mode alignment characteristics; the specific steps of the step S2 are as follows: s21, determining the filter length, namely determining the unique filter length by utilizing the rotation angle of the vector changing along with time; For the number of measured points Bridge multi-channel monitoring data: T represents the transpose, use the vector Rotation angle over time To determine a unique filter length Wherein The definition is as follows: (15) for filter length The following formula is used: (16) Wherein, the Representing sequences The number of data points included in the data set, Is that The number of the middle extreme points, Is a weight parameter that is used to determine the weight of the object, Rounding a number to the nearest integer, by which the average scale of the highest frequency rotation in the embedded signal can be estimated; S22, iterating the decomposition signal according to the following formula until the difference between adjacent iterations meets a stopping criterion; (17) In the formula, The number of iterations is indicated and, Represent the first The first channel is at The IMF of the step is performed, Is that The corresponding fourier transform is used to determine, Representing the identity matrix of the cell, Representing a diagonal matrix, the DFT representing a discrete fourier transform; The formula of the stopping criterion is: (18) Wherein, the Representing Euclidean norms when And is also provided with , When the natural number is set, the formula (18) is established; Obtaining a first IMF according to the following equation (19), subtracting the newly generated IMF from the current signal, and repeating the above calculation steps until Obtaining a final IMF group according to a formula (21); (19) (20) (21) Wherein, the initial value of the IMF is 0, Representing an inverse fourier transform; To this end, original The dimension monitoring signal is simultaneously decomposed into a plurality of IMF groups: (22) (23) (24) Wherein, the Representing each channel signal I.e. the number of decomposition levels, Is the index of the number of layers of the IMF, Represent the first IMF array of layer S3, calculating the instantaneous frequency and bandwidth of the IMF in the step S2 through HHT, screening out IMF components containing the target modal frequency and carrying out linear reconstruction; S4, taking the multi-channel monitoring data reconstructed in the step S3 as the system input of the COV-SSI algorithm, calculating a system matrix and an output matrix, and calculating a modal damping ratio by utilizing eigenvalue decomposition: s151, screening poles in the stable graph to ensure the authenticity of data, wherein the modal parameter meets the following criteria that 1) the modal damping ratio is always positive value, and 2) the modal damping ratio is not more than 15%; S152, selecting the initial point in the step S151, calculating the point number in the neighborhood, namely randomly selecting a frequency point which is not accessed from all poles of the stable graph, and recording the frequency point as a point Calculate the point Radius of (2) All points within the range are noted as If (3) The number of points is less than the density parameter MinPts, and the points are Marked as noise points if The number of points is not less than MinPts, the point As core point, take the point Starting a new cluster for the start point; S153, judging whether the point is a noise point or a core point according to the points in the neighborhood in the step S152, and expanding the clusters based on the noise point or the core point And All points in (a) are initialized as a new cluster, and marked as a cluster From the cluster Starting from the core point in the cluster, checking the neighborhood of the core point in turn, and for the cluster Each point core point in (a) Checking its neighborhood If (3) The number of points in the sequence is not less than MinPts, then All points in (a) are added to the cluster In, for each core point in the neighborhood, repeatedly checking the neighborhood until the cluster Cannot be extended any more; s154, repeating the steps of the steps S151-S153 until all poles are accessed; s155, taking the frequency average value and the damping ratio average value of poles in each cluster as representative values of data in the clusters.
2. The automatic recognition method of bridge modal parameters considering multichannel information according to claim 1, wherein step S1 specifically comprises the following steps: S11, constructing a discrete state model: The random discrete time state space model of the freedom degree system has the expression: (1) Wherein, the As a vector of discrete-time states, In order to monitor the data, For the number of the measuring points, In order for the process to be noisy, In order to measure the noise of the light, In the form of a matrix of discrete systems, Is a discrete output matrix; based on the multi-channel monitoring signal, discrete is utilized Time of day monitoring data Construction of Hankel matrix The expression is as follows: (2) Wherein, the Represent the first Sequence composed of output signals of all measuring points at moment, hankel matrix sharing Individual block rows and Columns, divided into And Two matrices, each matrix having a respective "past" and "future" representation Individual block lines, parameters Representing the number of samples used to calculate the output covariance matrix, theoretically ; S12, calculating an output covariance matrix, namely estimating the output covariance matrix by using the limited data quantity: (3); S13, constructing a Toeplitz matrix, namely constructing the Toeplitz matrix based on the output covariance matrix in the step S12, and further finishing to obtain a state-output covariance matrix; multichannel monitoring data according to step S11 Constructing a Toeplitz matrix by using the output covariance matrix, wherein the expression is as follows: (4) further finishing the Toeplitz matrix to obtain: (5) Wherein, the Representing a state-output covariance matrix, equation (5) shows that the Toeplitz matrix can be decomposed into a generalized observable matrix Expanding a controllable matrix Wherein For the order of the system, For the number of measurements, a generalized observable matrix is observed Comprising a system matrix And output matrix ; S14, singular value decomposition, namely performing singular value decomposition on the Toeplitz matrix in the step S13 to obtain a system matrix And output matrix ; The Toeplitz matrix is subjected to singular value decomposition by combining the formula (5) in the step S13 to obtain: (6) (7) deriving the output matrix from equation (7) Front equal to generalized observable matrix A row; Definition of the definition Is a matrix of two sub-matrices of (a) And : (8); Obtaining a system matrix from the formula (8) in the step S14 : (9) Wherein, the Representing a pseudo-inverse of the matrix; s15, constructing a stability diagram, namely obtaining a system matrix through matrix operation, decomposing characteristic values of the system matrix, calculating modal frequency identification modal parameters, and further primarily screening poles to construct the stability diagram; Matrix of system And (3) performing eigenvalue decomposition: (10) Wherein, the Is a matrix of discrete-time feature vectors of the system, Is the first Order characteristic value A diagonal matrix formed; The frequency of the system is given by: (11) Wherein, the Represents the conjugation of the polymer and the polymer, Is the modal frequency; and (3) observing the changes of the modal frequency, the damping ratio and the modal shape under different model orders by combining a stability diagram, and identifying the real physical mode, wherein the calculation formula is as follows: (12) (13) (14) In the formula, The order of the model is represented and, 、、 Respectively representing the modal frequency, damping ratio and modal shape of each order; 、、 The method comprises the steps of respectively obtaining characteristic frequency tolerance, damping ratio tolerance and vibration mode tolerance, obtaining MAC as a mode guarantee criterion, and automatically extracting a stable axis from a generated stable graph by using a DBSCAN clustering algorithm.
3. The method for automatically identifying bridge modal parameters in consideration of multi-channel information according to claim 1, wherein step S2 uses pre-defined parameters including weight coefficients, stopping criteria and maximum number of iterations when decomposing signals of the multi-channel monitoring data using MvFIF.
4. The automatic recognition method of bridge modal parameters considering multichannel information according to claim 1, wherein step S3 specifically comprises the following steps: s31, calculating instantaneous frequency and bandwidth, namely calculating instantaneous average frequency and bandwidth of each Intrinsic Mode Function (IMF) by using Hilbert-Huang transform (HHT); s32, screening IMF components, namely associating the identified structural modal frequency with the instantaneous frequency bandwidth of the IMF components, and reserving the IMF components with the bandwidths fully containing the target modal frequency; S33, performing linear reconstruction on the selected IMF, and extracting basic information for capturing the dynamic characteristics of the structural mode.
5. The method for automatically identifying bridge modal parameters in consideration of multi-channel information as set forth in claim 2, wherein step S4 is specifically implemented by taking the multi-channel monitoring data reconstructed in step S3 as system input of COV-SSI algorithm, and calculating according to formulas (1) - (10) in step S1 to obtain a system matrix And output matrix Wherein the damping ratio is Calculated from the following formula: (25) Wherein, the The real part is represented by a real part, And (3) obtaining poles corresponding to each order of modes of the stable graph by utilizing formulas (12) - (14) in the step S1, and taking the average damping ratio as a final recognition result.
6. An identification system for executing the automatic identification method of bridge modal parameters taking into account multi-channel information according to any one of claims 1-5, comprising a frequency identification module, a multi-element signal decomposition module, a frequency-based signal screening and reconstruction module and a damping ratio identification module; The frequency identification module is used for automatically identifying the modal frequency of the bridge structure by analyzing the multichannel monitoring signals and combining a DBSCAN clustering algorithm, and providing a basis for the subsequent modal parameter identification; the multi-element signal decomposition module is characterized in that MvFIF is utilized to decompose multi-channel monitoring data, and an intrinsic mode function group with excellent mode alignment characteristic is extracted so as to better process complex signals; The frequency-based signal screening and reconstructing module is characterized in that basic information for capturing dynamic characteristics of structural modes is extracted by screening IMF components related to the frequency of the structural modes and performing linear reconstruction, so that the signal-to-noise ratio and the content of useful information of the signals are improved; The damping ratio identification module specifically takes the reconstructed data as input, calculates the modal damping ratio, and finally realizes accurate identification of the bridge structural modal parameters, thereby providing key data support for bridge damage diagnosis and state evaluation.
7. A computer device comprising a processor and a memory, the memory being configured to store a computer executable program, the processor reading part or all of the computer executable program from the memory and executing the computer executable program, the processor executing part or all of the computer executable program, being configured to implement the method for automatically identifying bridge modal parameters taking into account multi-channel information according to any one of claims 1 to 5.
8. A computer readable storage medium having a computer program stored therein, which when executed by a processor, is capable of implementing the method for automatically identifying bridge modal parameters taking into account multi-channel information according to any one of claims 1-5.

Description

Automatic bridge modal parameter identification method and system considering multi-channel information Technical Field The invention belongs to the technical field of bridge structure health monitoring and signal processing, relates to an automatic bridge modal parameter identification method considering multichannel information, and particularly relates to an automatic bridge modal parameter identification method and system integrating multi-element rapid iterative filtering (MvFIF) and DBSCAN clustering, computer equipment and a computer readable storage medium. Background The traffic flow of the modern society is rapidly increased, the types of vehicles are increasingly diversified, and the proportion of heavy-duty vehicles is continuously increased, so that the load working condition born by the bridge is more complicated and severe. In recent years, natural disasters such as earthquakes, floods, strong winds and the like, and extreme events such as ship collision, fire disasters and the like frequently occur, and the changes of traffic load working conditions are interwoven with various adverse events, so that the problems of accelerated degradation of durability, great reduction of bearing capacity, gradual attenuation of structural resistance and the like of the bridge are caused. The changes of the modal parameters (such as frequency, damping ratio, vibration mode and the like) of the bridge structure can reflect the problems of damage, fatigue or aging and the like, so that the accurate identification of the modal parameters has important significance for the damage diagnosis, state evaluation and disaster prevention of the bridge. The working state modal analysis (OMA) is a structural dynamics testing method based on environmental excitation, and can be used for testing when the structure is in a working state, without manual excitation and without affecting the normal use of the structure, so that the method is mainly applied to modal analysis of large-scale structures. OMA is mainly classified into a frequency domain method, a time domain method and a time-frequency domain method according to the difference of identification domains. The random subspace identification (SSI) is the most commonly used time domain modal parameter identification method at present, and is widely applied to the operation modal analysis of large civil infrastructures. However, in an actual bridge monitoring scenario, the input data of SSI is inevitably contaminated by noise due to environmental factors and equipment limitations. Covariance random subspace identification, COV-SSI, is a systematic identification method for modal parameter identification (such as frequency, damping ratio and vibration mode), which is commonly used for structural health monitoring, vibration analysis of bridges and buildings. Although the COV-SSI is theoretically incorporated into a noise model, when the measured noise variance is large, noise components in singular value decomposition of a covariance matrix can mask real modal information, and deviation occurs in modal parameter estimation, so that the performance of the SSI, particularly the recognition accuracy of a modal damping ratio, is directly influenced. Meanwhile, as the characteristic of the covariance matrix can be changed due to the existence of noise, false modes (namely noise points) exist in the stable graph generated based on the SSI besides the real modes of the structure, and the noise points do not accord with the distribution rule of the normal mode characteristics, so that the manual selection and extraction of the physical modes on the stable graph are seriously influenced. Therefore, how to accurately and automatically identify the modal parameters of the bridge structure in a complex multi-channel strong noise environment is a key problem to be solved by those skilled in the art. Disclosure of Invention In view of the above, the invention provides a bridge modal parameter automatic identification method and system considering multi-channel information, which can consider the space-time correlation among the multi-channel data to realize synchronous decomposition of the multi-channel monitoring data, and can accurately identify structural modal parameters under different noise levels by effectively filtering and extracting useful information related to structural modal parameters based on frequency, and simultaneously, realize automation of modal parameter extraction, thereby avoiding subjective errors caused by manual intervention. In order to achieve the above purpose, the present invention provides the following technical solutions: A bridge modal parameter automatic identification method considering multichannel information comprises the following steps: S1, directly analyzing a multichannel monitoring signal by using COV-SSI to generate a stability diagram, and automatically extracting a stability axis by combining a DBSCAN clustering algorithm to realize auto