CN-117629550-B - Signal excitation and identification method, system and computer equipment

CN117629550BCN 117629550 BCN117629550 BCN 117629550BCN-117629550-B

Abstract

The application provides a signal excitation and identification method, a system and computer equipment, wherein the method comprises the steps of constructing a model, and carrying out modal analysis on the model to obtain natural frequencies and modes; determining excitation period according to natural frequency of a structure to be tested, generating a periodic pulse excitation signal by a vibration exciter, generating vibration by striking a piece to be tested through a push rod, sensing vibration by a sensor to form original vibration signals, decomposing each original vibration signal, obtaining a combined signal according to an I MF component obtained after decomposition, extracting characteristics of the combined signal to obtain corresponding characteristic parameter data, and constructing a data set, so that training of an initial signal recognition model is realized, and recognition of the signal to be tested is completed. The signal excitation and identification method provided by the application can accurately monitor the structural health state in real time so as to discover potential structural faults in advance and improve the running safety and efficiency of equipment.

Inventors

RAO CHUNFANG
XIONG WENTING
MA YIMING
CHEN ZIYING
CHEN SISI
LOU CHAO
HU YAOYANG
YAN XIAOLI
CHEN PENG
YU WENXIN
RAO CHAO
WU SHIJIE
YUAN JIAXIN
WANG YUEXIANG
RUAN YIMING

Assignees

江西师范大学

Dates

Publication Date: 20260505
Application Date: 20231123

Claims (6)

1. A method of signal excitation and identification, the method comprising: Constructing a finite element simulation analysis model of a structure to be tested, carrying out modal analysis on the finite element simulation analysis model of the structure to be tested to obtain the natural frequency and the modal of the structure to be tested, determining the assembly mode between a suspension type fiber bragg grating sensor and the structure to be tested, and obtaining the resonance frequency of the suspension type fiber bragg grating sensor according to the assembly mode between the suspension type fiber bragg grating sensor and the structure to be tested, wherein the method comprises the steps of determining the positions of sticking points and the positions of the fiber bragg gratings according to the modal analysis result, obtaining the positions of the sticking points, the distances from the sticking points to the fiber bragg gratings and the distances from the fiber bragg gratings to a tail fiber according to the positions of the sticking points to the fiber bragg gratings, the distances from the fiber bragg gratings to the tail fiber and the grating length of the fiber bragg gratings, and calculating the total length L according to the L to obtain the resonance frequency of the suspension type fiber bragg grating sensor; Determining an excitation period according to the natural frequency, generating a periodic pulse excitation signal by a vibration exciter according to the excitation period, and generating an original vibration signal by striking a structural member to be tested with a known state signal through a push rod according to the periodic pulse excitation signal, wherein the known state signal at least comprises a torque signal, a crack signal, a friction coefficient signal and a material purity signal, and the excitation signal is obtained according to the following formula: , wherein, The excitation signal is represented by a signal representing the excitation signal, Representing the unit impulse function, N is an integer, A is impulse intensity, T is time, T is excitation period, and the step of calculating the resonance frequency of the suspended fiber Bragg grating sensor according to the total length L comprises the following steps of calculating the resonance frequency of the suspended fiber Bragg grating sensor according to the following formula: Wherein f 0 represents the resonance frequency of the suspended fiber Bragg grating sensor, and C represents the speed of sound waves in the optical fiber; Decomposing and extracting the original vibration signal by adopting an empirical mode decomposition algorithm to obtain a multi-order IMF component, wherein the method comprises the steps of decomposing the original vibration signal into two parts of local smoothing and local oscillation, extracting extreme points of the local smoothing part to obtain a group of local extreme point sequences, carrying out interpolation fitting on the local extreme point sequences to obtain a group of local smoothing functions, obtaining a group of local oscillation functions according to the local smoothing functions and the original vibration signal, and obtaining a first-order IMF component based on the local oscillation functions; And calculating cross-correlation coefficients between the time domain signals of the IMF components of each order and the original vibration signals respectively, comprising obtaining the cross-correlation coefficients according to the following formula: the mean value of the time domain signal of all sampled kth order IMF components is obtained according to the following formula: the mean value of the time domain signals of all sampled original vibration signals is obtained according to the following formula: Where r represents a cross-correlation coefficient between the time domain signal of the kth order IMF component and the time domain signal of the original vibration signal, i represents the ith sampling point, m represents the total sampling point, A time domain signal representing the ith sample point of the kth order IMF component, The mean value of the time domain signal representing the k-th order IMF component of all samples, The mean value of the time domain signal representing all sampled raw vibration signals, A time domain signal representing the original vibration signal of the i-th sampling point; Screening at least one target IMF component from all IMF components according to the cross-correlation coefficient, denoising the target IMF component with suspended fiber Bragg grating resonance noise according to the resonance frequency of the suspended fiber Bragg grating sensor, and superposing all the target IMF components after denoising to obtain a combined signal; extracting dimensional parameters and dimensionless parameters from the combined signals respectively, and screening the dimensional parameters and the dimensionless parameters to obtain characteristic parameter data so as to construct a data set according to the characteristic parameter data; And inputting the data set into an initial signal recognition model for training to obtain a final signal recognition model, and inputting the signal to be tested into the final signal recognition model to obtain a health state classification result.
2. The signal excitation and identification method according to claim 1, wherein the step of obtaining a set of local oscillation functions from the local smoothing function and the original vibration signal, and obtaining the first-order IMF component based on the local oscillation functions comprises: The local oscillation function is obtained according to the following formula: Wherein, the The local oscillation function is represented by a function of local oscillation, Representing the original vibration signal(s), Representing a local smoothing function; Judging whether the local oscillation function meets a preset IMF component condition or not, wherein the preset IMF component condition comprises the fact that the number of extreme points contained in the local oscillation function is equal to the number of zero crossing points and the average value of the envelope function is zero; If the local oscillation function does not meet the preset IMF component condition, taking the obtained local oscillation function as a new vibration signal, and calculating again according to the new vibration signal to obtain a new local oscillation function until the new local oscillation function meets the preset IMF component condition, and outputting the local oscillation function meeting the preset IMF component condition as a first-order IMF component; if the local oscillation function meets the preset IMF component condition, the local oscillation function is performed As a first order IMF component; The kth order IMF component is obtained according to the following formula: Wherein n represents the number of decomposition times, Representing the IMF component of the kth order, Representing the remaining components after the nth decomposition.
3. The signal excitation and identification method according to claim 1, wherein the steps of screening at least one target IMF component from all IMF components according to the cross-correlation coefficient, denoising the target IMF component having suspended fiber bragg grating resonance noise according to the suspended fiber bragg grating sensor resonance frequency, and superimposing all the target IMF components after denoising to obtain a combined signal include: judging whether the cross-correlation coefficient corresponding to any IMF component is larger than a first preset threshold value or not, and taking all IMF components larger than the first preset threshold value as target IMF components; And determining a filtering frequency range according to the resonance frequency of the suspended fiber Bragg grating sensor, so as to denoise a target IMF component with suspended fiber Bragg grating resonance noise according to the filtering frequency range.
4. The signal excitation and identification method according to claim 1, wherein the steps of extracting the dimensional parameter and the non-dimensional parameter from the combined signal, and filtering the dimensional parameter and the non-dimensional parameter to obtain the characteristic parameter data include: the dimensionless parameters comprise average value, standard deviation, maximum value, minimum value, residual error, peak-to-peak value and energy, and the dimensionless parameters comprise skewness, kurtosis, waveform factors, amplitude factors, impact factors and margin factors; and adopting a principal component analysis algorithm to perform dimension reduction treatment on all the dimensional parameters and the dimensionless parameters so as to remove redundant variables and obtain characteristic parameter data.
5. A signal excitation and identification system, wherein the system is applied to a signal excitation and identification method as claimed in claim 1, the system comprising: The finite element modal analysis module is used for constructing a finite element simulation analysis model of the structure to be tested, carrying out modal analysis on the finite element simulation analysis model of the structure to be tested to obtain the natural frequency and the mode of the structural member to be tested, determining the assembly mode between the suspended fiber Bragg grating sensor and the structural member to be tested, and obtaining the resonance frequency of the suspended fiber Bragg grating sensor according to the assembly mode between the suspended fiber Bragg grating sensor and the structural member to be tested; The excitation signal generation module is used for determining an excitation period according to the natural frequency, generating a periodic pulse excitation signal according to the excitation period by the vibration exciter, and generating an original vibration signal by striking a structural member to be tested with a known state signal according to the periodic pulse excitation signal through the ejector rod, wherein the known state signal at least comprises a torque signal, a crack signal, a friction coefficient signal and a material purity signal; the vibration signal decomposition module is used for decomposing and extracting the original vibration signal by adopting an empirical mode decomposition algorithm to obtain multi-order IMF components, and calculating cross-correlation coefficients between time domain signals of each order IMF component and the original vibration signal respectively; The signal denoising module is used for screening at least one target IMF component from all IMF components according to the cross correlation coefficient, denoising the target IMF component with suspension type fiber Bragg grating resonance noise according to the resonance frequency of the suspension type fiber Bragg grating sensor, and superposing all the target IMF components after denoising to obtain a combined signal; The data set construction module is used for respectively extracting dimensional parameters and dimensionless parameters from the combined signals, screening the dimensional parameters and the dimensionless parameters to obtain characteristic parameter data, and constructing a data set according to the characteristic parameter data; The recognition result output module is used for inputting the data set into the initial signal recognition model for training to obtain a final signal recognition model, and inputting the signal to be tested into the final signal recognition model to obtain a health state classification result.
6. A computer device comprising a memory and a processor, wherein: The memory is used for storing a computer program; the processor is adapted to implement the signal excitation and identification method of any of claims 1-4 when executing a computer program stored on the memory.

Description

Signal excitation and identification method, system and computer equipment Technical Field The application relates to the technical field of structural health monitoring, in particular to a signal excitation and identification method, a system and computer equipment. Background Structural health monitoring and diagnosis is extremely important in the fields of aerospace, medical and structural engineering. It generally requires high-precision sensing technology to detect possible minor changes, such as loosening of artificial bone structures in the medical field, which may have a serious impact on the diagnosis and treatment effect. Vibration devices for monitoring whether a structure is loose are generally active excitation, and from the perspective of the excited tool, can be classified into hammering excitation and exciter excitation. In the hammering excitation, an excitation signal with extremely small time duration is applied to the object to be detected only once, and the excitation signal is approximately an impulse function with larger energy from the signal angle. Because the excitation energy is only supplied when excited once, the energy is needed to be large, the to-be-detected piece can be damaged, and the excitation mode is not suitable for being used as an excitation mode in long-term structural health monitoring for small-scale and low-mass structural members. In the excitation mode of the vibration exciter, a push rod for outputting force in the vibration exciter is generally connected with a piece to be detected through a connector, and in the arrangement, the push rod of the vibration exciter is connected with the piece to be detected through the connector. The excitation signal can be a determination signal and a random signal, wherein the determination signal generally has two modes of sinusoidal sweep excitation and sinusoidal fast sweep excitation, and the random excitation signal comprises the forms of pure random excitation, windowed random excitation, pseudo-random excitation, periodic random excitation, burst random excitation and the like. The excitation of the above signals to the structure to be measured is based on the physical structure of the exciter arrangement. However, for small mass, small scale structures, the mass or size of the connector itself is of the same order of magnitude or greater than the structure to be tested. Thus, the arrangement of such a typical exciter vibrator will directly influence the dynamics of the structure to be measured, and therefore is not suitable for monitoring small-mass, small-scale structural members. Disclosure of Invention Based on the above, the application aims to provide a signal excitation and identification method, a system and computer equipment, which are based on the use of a suspended FBG sensing system to solve the problem that the traditional monitoring method is difficult to be applied to small-mass and small-scale structural members. In one aspect, the application provides a signal excitation and identification method, which comprises the following steps: building a finite element simulation analysis model of a structure to be tested, and carrying out modal analysis on the finite element simulation analysis model of the structure to be tested to obtain the natural frequency and the mode of a structural member to be tested, thereby determining the assembly mode between the suspended fiber Bragg grating sensor and the structural member to be tested, and obtaining the resonance frequency of the suspended fiber Bragg grating sensor according to the assembly mode between the suspended fiber Bragg grating sensor and the structural member to be tested; Determining an excitation period according to the natural frequency, generating a periodic pulse excitation signal by a vibration exciter according to the excitation period, and generating an original vibration signal by striking a structural member to be tested with a known state signal by a push rod according to the periodic pulse excitation signal, wherein the known state signal at least comprises a torque signal, a crack signal, a friction coefficient signal and a material purity signal; decomposing and extracting the original vibration signal by adopting an empirical mode decomposition algorithm to obtain multi-order IMF components, and calculating cross-correlation coefficients between time domain signals of each-order IMF component and the original vibration signal respectively; Screening at least one target IMF component from all IMF components according to the cross-correlation coefficient, denoising the target IMF component with suspended fiber Bragg grating resonance noise according to the resonance frequency of the suspended fiber Bragg grating sensor, and superposing all the target IMF components after denoising to obtain a combined signal; extracting dimensional parameters and dimensionless parameters from the combined signals respectively, and screening the dimensio