Hollow slab beam mechanical property degradation prediction method and system based on vibration frequency

Abstract

The application discloses a hollow slab beam mechanical property degradation prediction method and system based on vibration frequency, relates to the technical field of hollow slab beam mechanical property degradation prediction, and comprises the following steps: collecting vibration data, pre-processing the collected data to obtain standard data; extracting mechanical property parameters from the standard data, establishing a mechanical property degradation model of the hollow slab beam according to the mechanical property parameter extraction result; comprehensively constructing a prediction model between vibration data and mechanical property degradation according to the influence of different degradation states of the mechanical property parameters on the vibration data of the hollow slab beam; and monitoring vibration data of the hollow slab beam in real time, and outputting the mechanical property degradation condition of the hollow slab beam in combination with the prediction model. The application establishes the relationship between vibration frequency and the mechanical property of the cracked hollow slab beam, constructs a prediction model between vibration data and mechanical property degradation, monitors vibration data of the hollow slab beam in real time, analyzes the mechanical property degradation condition of the hollow slab beam, and performs mechanical property degradation early warning on the hollow slab beam.

Description

Technical Field

[0001] This invention relates to the field of mechanical property degradation prediction technology for hollow slab beams, and more specifically to a method and system for predicting mechanical property degradation of hollow slab beams based on vibration frequency. Background Technology

[0002] Bridges are crucial nodes and arteries in highways, railways, and urban roads, playing a pivotal role in economic development. Among these, hollow slab girder bridges account for approximately 50% of small-to-medium span bridges. In-service reinforced concrete (RC) hollow slab girder bridges are subject to long-term cyclic loads and environmental factors, resulting in numerous transverse and longitudinal cracks, hinge joint failure (single-slab stress), and other problems. Further deterioration will lead to serious consequences such as reduced structural durability and bridge failure. To adapt to the needs of economic development, whether the load-bearing capacity of existing highway bridges can meet the load standards after reconstruction, expansion, and load upgrades, and whether the remaining service life of the removed beams and slabs of old bridges can meet the standards required after highway reconstruction and expansion, are critical issues that urgently need to be addressed in highway reconstruction, expansion, and bridge operation management.

[0003] However, due to the complex distribution characteristics of defects in service RC hollow slab beam bridges and the randomness and uncertainty of environmental effects, existing research on the theory and methods of RC beam vibration testing based on dynamic characteristics and the study on the degradation of mechanical properties of cracked RC hollow slab beams are still insufficient in the following aspects: (1) The vertical bending vibration frequency response characteristics of RC hollow slab beams after service degradation are still unclear. Although there are many studies on the cracking mechanism and bearing capacity of hollow slab beams, the research on the vertical bending response frequency of hollow slab beams under different cracking states is not very sufficient, especially the research on RC beams that have been in service for many years is even more lacking. (2) The degradation characteristics and evolution law of the vertical bending vibration frequency of RC hollow slab beams are still unclear. There are relatively few studies on the relationship between the vibration frequency of RC beams and their bearing capacity, stiffness, deflection, etc., and the degradation characteristics and evolution law of the vibration frequency of RC beams are still unclear.

[0004] Therefore, how to propose a method and system for predicting the mechanical property degradation of hollow slab beams based on vibration frequency, establish the relationship between vibration frequency and the mechanical properties of cracked RC hollow slab beams, and perform mechanical property degradation prediction of hollow slab beams are problems that urgently need to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a method and system for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, establishing the relationship between vibration frequency and the mechanical properties of cracked RC hollow slab beams, and predicting the degradation of mechanical properties of hollow slab beams. To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A method for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency includes:

[0007] Sensors are deployed on the hollow slab beams to collect their vibration data in real time, and the collected data is preprocessed to obtain standard data.

[0008] Mechanical performance parameters were extracted from standard data, and a mechanical performance degradation model for hollow slab beams was established based on the extracted mechanical performance parameters.

[0009] Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a prediction model between vibration data and mechanical performance degradation is constructed.

[0010] The vibration data of hollow slab beams are monitored in real time, and the mechanical performance degradation of hollow slab beams is output in combination with the prediction model, and early warning is given.

[0011] Optionally, deploying sensors on the hollow slab beam to collect its vibration data in real time includes: deploying accelerometers and vibration sensors on the hollow slab beam to collect its vibration data in real time, wherein the vibration data includes vibration frequency.

[0012] Optionally, the preprocessing of the collected data to obtain standard data includes:

[0013] Obtain the degraded dataset and the normal dataset, and extract the degraded waveform and the normal waveform;

[0014] Real-time acquisition of vibration data on hollow slab beams; preprocessing of the acquired vibration data.

[0015] The preprocessed signal is subjected to Discrete Fourier Transform to analyze the spectral characteristics of the vibration data.

[0016] Extract positively correlated waveform features based on the degradation type;

[0017] A first similarity threshold and a second similarity threshold are preset, and a similarity calculation based on deep learning is performed. The extracted waveforms are compared with the degraded waveforms and normal waveforms respectively.

[0018] If the similarity between two consecutive extracted waveforms and the normal waveform is lower than the first similarity threshold, and the similarity between two consecutive extracted waveforms and the degraded waveform is higher than the second similarity threshold, then it is considered to meet the corresponding degradation type.

[0019] Optionally, the establishment of the mechanical property degradation model for hollow slab beams includes:

[0020] A mathematical simulation model of hollow slab beams is established. Based on the mathematical simulation model of hollow slab beams, the mechanical performance parameters and performance specificities of hollow slab beams are calculated. The mechanical performance parameters include fatigue damage, crack propagation, and material degradation.

[0021] Positively correlated mechanical performance parameters are selected from the mechanical performance parameters, and performance-specific models corresponding to all positively correlated mechanical performance parameters of hollow slab beams are established. Then, combined with the error model in the mathematical simulation model of hollow slab beams, a mechanical performance degradation model of hollow slab beams is established.

[0022] Optionally, the calculation of the mechanical performance parameters and performance specificity of the hollow slab beam includes:

[0023] After deconstructing the hollow slab beam based on the kinematic model in the mathematical simulation model of the hollow slab beam, the mechanical performance parameters of the hollow slab beam are obtained.

[0024] Based on the error model and probability error model in the mathematical simulation model of hollow slab beam, the mixed error sources of hollow slab beam are decomposed to obtain all positively correlated mechanical performance parameters of hollow slab beam.

[0025] The influence of all positively correlated mechanical property parameters of hollow slab beams on hollow slab beams is calculated to obtain the performance specificity of hollow slab beams.

[0026] Optionally, the mathematical simulation model of the hollow slab beam includes the kinematic model, error model, and probabilistic error model of the hollow slab beam.

[0027] Optionally, the step of constructing a predictive model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams includes:

[0028] Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a first relationship model between vibration data and mechanical performance parameters is established, and the monitored vibration data is correlated with the mechanical performance degradation model. Vertical bending frequency data is extracted from the vibration data, and a second relationship model is constructed based on the extracted vertical bending frequency and mechanical performance parameters. The extracted vertical bending frequency is then correlated with the mechanical performance degradation model. Finally, a prediction model between vibration data and mechanical performance degradation is constructed.

[0029] Optionally, the process of obtaining the degraded dataset and the normal dataset may further include:

[0030] Transform the vibration power spectral density to the amplitude spectrum;

[0031] The amplitude spectrum is padded with zeros according to the sampling frequency range;

[0032] Multiple single-frame pseudo-random signals for determining the vibration power spectral density are obtained by performing inverse Fourier transform on the zero-padded amplitude spectrum and the uniformly distributed phase generated according to the congruence method.

[0033] Multiple single-frame pseudo-random signals are processed by a quadratic B-spline weighting and three-layer superposition merging method to determine a true random vibration signal of a preset duration;

[0034] Degraded and normal datasets are constructed by extracting multiple frames from true random vibration signals of a preset duration, and the degraded and normal waveforms are extracted.

[0035] Optionally, the establishment of the first relationship model between vibration data and mechanical performance parameters includes:

[0036] Establish a multi-objective integrated optimization model for vibration data and mechanical performance parameters;

[0037] Establish a database of vibration data and mechanical performance parameters, denote the vibration frequency as the first correlation function, and denote multiple mechanical performance parameters as multiple characteristic constraints to obtain the database of vibration data and mechanical performance parameters;

[0038] The database is analyzed and compared to understand its operating patterns. The parameters are reweighted to adjust the optimization function. The optimized function, after being adjusted by the database, is fitted to obtain fitted data. The fitted data is then compared with the real-time data to ensure that the monitored real-time vibration data, mechanical performance parameters, and fitted data are close to the set values.

[0039] Optionally, the comprehensive construction of the prediction model between vibration data and mechanical performance degradation includes comparing the prediction results of the first relationship model with the prediction results of the second relationship model, converting the prediction results of the second relationship model and the prediction results of the first relationship model into vector form, calculating the Euclidean distance between the two, iterating the calculation process until the Euclidean distance between the two is less than a preset threshold, and outputting the prediction results of the first relationship model.

[0040] Optionally, a vibration frequency-based system for predicting the mechanical property degradation of hollow slab beams includes:

[0041] Acquisition module: Used to deploy sensors on hollow slab beams to collect their vibration data in real time, and to preprocess the collected data to obtain standard data;

[0042] Mechanical property degradation model construction module: used to extract mechanical property parameters from standard data and establish a mechanical property degradation model for hollow slab beams based on the extracted mechanical property parameters;

[0043] Prediction model building module: used to comprehensively build a prediction model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams;

[0044] Prediction module: Used to monitor the vibration data of hollow slab beams in real time, combine the prediction model to output the mechanical performance degradation of hollow slab beams, and provide early warning.

[0045] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method and system for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, which has the following beneficial effects:

[0046] This invention proposes a method for predicting the mechanical performance degradation of hollow slab beams based on vibration frequency. The method includes: deploying sensors on the hollow slab beam to collect vibration data in real time; preprocessing the collected data to obtain standard data; extracting mechanical performance parameters from the standard data; establishing a mechanical performance degradation model for the hollow slab beam based on the extracted parameters; comprehensively constructing a prediction model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of the hollow slab beam; real-time monitoring of the vibration data of the hollow slab beam; outputting the mechanical performance degradation status of the hollow slab beam based on the prediction model; and providing early warning. This invention addresses the issue that as structural damage develops, structural stiffness changes, and vibration frequency can characterize these changes. Establishing the relationship between vibration frequency and the mechanical properties of cracked RC hollow slab beams can provide a basis for predicting the mechanical performance degradation of cracked RC hollow slab beams based on vibration frequency. With increasing service life, cracking has a more significant impact on the vertical bending response frequency of RC beams. Therefore, it is necessary to conduct research on the influence of cracking on the vertical bending vibration frequency response characteristics of RC hollow slab beams after service degradation. The relationship between vibration frequency and mechanical properties of cracked RC hollow slab beams was established, and the degradation of mechanical properties of hollow slab beams was predicted. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0048] Figure 1 A schematic diagram of the process for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, provided by this invention.

[0049] Figure 2 The structural framework diagram of the mechanical property degradation prediction system for hollow slab beams based on vibration frequency provided by the present invention is shown. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] This invention discloses a method for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, such as... Figure 1 As shown, it includes:

[0052] Sensors are deployed on the hollow slab beams to collect their vibration data in real time, and the collected data is preprocessed to obtain standard data.

[0053] Mechanical performance parameters were extracted from standard data, and a mechanical performance degradation model for hollow slab beams was established based on the extracted mechanical performance parameters.

[0054] Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a prediction model between vibration data and mechanical performance degradation is constructed.

[0055] The vibration data of hollow slab beams are monitored in real time, and the mechanical performance degradation of hollow slab beams is output in combination with the prediction model, and early warning is given.

[0056] Furthermore, the provision of sensors on the hollow slab beam for real-time acquisition of its vibration data includes: deploying accelerometers and vibration sensors on the hollow slab beam to acquire its vibration data in real time, wherein the vibration data includes vibration frequency.

[0057] Furthermore, the preprocessing of the collected data to obtain standard data includes:

[0058] Obtain the degraded dataset and the normal dataset, and extract the degraded waveform and the normal waveform;

[0059] Real-time acquisition of vibration data on hollow slab beams; preprocessing of the acquired vibration data.

[0060] The preprocessed signal is subjected to Discrete Fourier Transform to analyze the spectral characteristics of the vibration data.

[0061] Extract positively correlated waveform features based on the degradation type;

[0062] A first similarity threshold and a second similarity threshold are preset, and a similarity calculation based on deep learning is performed to compare the extracted waveform with the degraded waveform and the normal waveform.

[0063] If the similarity between two consecutive extracted waveforms and the normal waveform is lower than the first similarity threshold, and the similarity between two consecutive extracted waveforms and the degraded waveform is higher than the second similarity threshold, then it is considered to meet the corresponding degradation type.

[0064] Specifically, the waveform similarity comparison includes: performing correlation calculations between the real-time waveform and the normal waveform and the degraded waveform, respectively, using the following formula:

[0065] Rxy(Φ)=x(t)y(t+Φ)dt; where x(t) is the waveform function of the normal dataset and the degraded dataset, y(t) is the real-time waveform function calculated by real-time sampling, Φ is the time offset, and Rxy(Φ) is the cross-correlation function; the preprocessing of the collected vibration data includes filtering, noise removal, normalization, and baseline drift removal.

[0066] Specifically, obtaining the degraded dataset and the normal dataset includes:

[0067] Data is collected and systematically sampled to select discrete vibration signals at fixed periods and time intervals. The discrete vibration signals are then subjected to a Discrete Fourier Transform, as shown in the formula:

[0068] In the formula: G(r) represents the transformed spectral signal, r represents the frequency component, u is the time-domain point index, and i is the imaginary unit;

[0069] Perform spectral analysis on the discrete Fourier transform results, including amplitude and phase, and save the spectral analysis results as normal datasets and degraded datasets, respectively.

[0070] Furthermore, the establishment of the mechanical property degradation model for hollow slab beams includes:

[0071] A mathematical simulation model of hollow slab beams is established. Based on the mathematical simulation model of hollow slab beams, the mechanical performance parameters and performance specificities of hollow slab beams are calculated. The mechanical performance parameters include fatigue damage, crack propagation, and material degradation.

[0072] Positively correlated mechanical performance parameters are selected from the mechanical performance parameters, and performance-specific models corresponding to all positively correlated mechanical performance parameters of hollow slab beams are established. Then, combined with the error model in the mathematical simulation model of hollow slab beams, a mechanical performance degradation model of hollow slab beams is established.

[0073] Furthermore, the calculation of the mechanical performance parameters and performance specificity of the hollow slab beam includes:

[0074] After deconstructing the hollow slab beam based on the kinematic model in the mathematical simulation model of the hollow slab beam, the mechanical performance parameters of the hollow slab beam are obtained.

[0075] Based on the error model and probability error model in the mathematical simulation model of hollow slab beam, the mixed error sources of hollow slab beam are decomposed to obtain all positively correlated mechanical performance parameters of hollow slab beam.

[0076] The influence of all positively correlated mechanical property parameters of hollow slab beams on hollow slab beams is calculated to obtain the performance specificity of hollow slab beams.

[0077] Furthermore, a performance-specific model is established for all positively correlated mechanical performance parameters of hollow slab beams. Then, combined with the error model in the mathematical simulation model of hollow slab beams, a mechanical performance degradation model of hollow slab beams is established.

[0078] The formula for the performance-specific model of each positively correlated mechanical performance parameter is:

[0079] D g =u(t,λ) i ), where u(t,λ i ) is a performance-specific function of each positively correlated mechanical performance parameter with respect to time t and the number of data collections i, where t is the time since the hollow slab beam was put into operation, and λ is the number of data collections i. i For the performance-specific factor that has undergone i acquisitions.

[0080] Furthermore, the Weibull distribution parameter estimation method is used to obtain the performance specificity model of the positively correlated mechanical performance parameters: the historical performance specificity model of the positively correlated mechanical performance parameters is analyzed, and the actual performance specificity is fitted by the Weibull distribution parameter estimation method to obtain the probability error distribution, thereby constructing the performance specificity model.

[0081] The formula for the mechanical property degradation model of hollow slab beams is:

[0082] D h =J error ·y(D g1 D g1 ,...,D gn ), where y(D g1 D g1 ,...,D gn ) is a function of the performance-specific model Epartdeg for each positively correlated mechanical performance parameter.

[0083] Furthermore, the prediction model that comprehensively constructs the relationship between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams includes:

[0084] Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a first relationship model between vibration data and mechanical performance parameters is established, and the monitored vibration data is correlated with the mechanical performance degradation model. Vertical bending frequency data is extracted from the vibration data, and a second relationship model is constructed based on the extracted vertical bending frequency and mechanical performance parameters. The extracted vertical bending frequency is then correlated with the mechanical performance degradation model. Finally, a prediction model between vibration data and mechanical performance degradation is constructed.

[0085] Furthermore, the process of obtaining the degraded dataset and the normal dataset includes:

[0086] Transform the vibration power spectral density to the amplitude spectrum;

[0087] The amplitude spectrum is padded with zeros according to the sampling frequency range;

[0088] Multiple single-frame pseudo-random signals for determining the vibration power spectral density are obtained by performing inverse Fourier transform on the zero-padded amplitude spectrum and the uniformly distributed phase generated according to the congruence method.

[0089] Multiple single-frame pseudo-random signals are processed by a quadratic B-spline weighting and three-layer superposition merging method to determine a true random vibration signal of a preset duration;

[0090] Degraded and normal datasets are constructed by extracting multiple frames from true random vibration signals of a preset duration, and the degraded and normal waveforms are extracted.

[0091] Furthermore, the establishment of the first relationship model between vibration data and mechanical performance parameters includes:

[0092] Establish a multi-objective integrated optimization model for vibration data and mechanical performance parameters;

[0093] Establish a database of vibration data and mechanical performance parameters, denote the vibration frequency as the first correlation function, and denote multiple mechanical performance parameters as multiple characteristic constraints to obtain the database of vibration data and mechanical performance parameters;

[0094] The database is analyzed and compared to understand its operating patterns. The parameters are reweighted to adjust the optimization function. The optimized function, after being adjusted by the database, is fitted to obtain fitted data. The fitted data is then compared with the real-time data to ensure that the monitored real-time vibration data, mechanical performance parameters, and fitted data are close to the set values. This results in a multi-objective comprehensive optimization model for vibration data and mechanical performance parameters.

[0095] Furthermore, a second relationship model between vertical bending frequency and mechanical performance parameters is established, including:

[0096] Establish a multi-objective integrated optimization model for vertical bending frequency and mechanical performance parameters;

[0097] A database of vertical bending frequency and mechanical performance parameters is established. The vibration frequency is denoted as the second correlation function, and multiple mechanical performance parameters are denoted as multiple characteristic constraints to obtain the database of vertical bending frequency and mechanical performance parameters.

[0098] The database is analyzed and compared to understand its operating rules. The parameters are reweighted to adjust the optimization function. The optimized function, after being adjusted by the database, is fitted to obtain fitted data. The fitted data is then compared with the real-time data to ensure that the monitored real-time vertical bending frequency, mechanical performance parameters and the fitted data are close to the set values. This results in a multi-objective comprehensive optimization model for vertical bending frequency and mechanical performance parameters.

[0099] Furthermore, the comprehensive construction of the prediction model between vibration data and mechanical performance degradation includes comparing the prediction results of the first relationship model with the prediction results of the second relationship model, converting the prediction results of the second relationship model and the prediction results of the first relationship model into vector form, calculating the Euclidean distance between the two, iterating the calculation process until the Euclidean distance between the two is less than a preset threshold, and outputting the prediction result of the first relationship model.

[0100] In a specific implementation, the method further includes: evaluating the correctness of the prediction results using a third-party relational model, including:

[0101] The horizontal vibration acceleration signal is collected. The obtained vibration acceleration signal includes the vibration acceleration value and its corresponding time. It is assumed that the vibration acceleration value constitutes the signal time domain amplitude sequence and the corresponding time constitutes the signal time sequence.

[0102] Impact signals are signals with short durations and extremely high instantaneous physical quantities. Therefore, the maximum value in the monitored vibration acceleration signal must be the transient impact response. Thus, the maximum value in the vibration acceleration signal is chosen as the starting point. The transient impact data sequence is extracted from the signal time-domain amplitude sequence and the signal time sequence. That is, the maximum value point of the signal time-domain amplitude sequence is found, and the N data points following the maximum value point are extracted to form the transient impact signal time-domain amplitude sequence. The corresponding signal time sequence at the same position is extracted, and the starting point is taken as the time zero point to obtain the N data points that form the transient impact signal time sequence.

[0103] Based on the free decay vibration dynamics model, nonlinear data fitting is performed on transient impact data to obtain the optimal impact response parameters; based on the damped free decay vibration dynamics model of a single degree of freedom system, the maximum likelihood method is used to perform nonlinear data fitting to obtain the optimal impact response parameters.

[0104] The predicted results are broken down to obtain amplitude and damping coefficient. The degradation of mechanical performance parameters is evaluated based on the magnitude of the optimal impact response parameters. Evaluation thresholds are set based on historical data, including setting an amplitude threshold and setting a coefficient threshold. When the amplitude threshold is exceeded, it indicates performance degradation. When the coefficient threshold is exceeded, it indicates performance degradation.

[0105] In a specific implementation, a hollow slab beam mechanical performance degradation prediction system based on vibration frequency, such as... Figure 2 As shown, it includes:

[0106] Acquisition module: Used to deploy sensors on hollow slab beams to collect their vibration data in real time, and to preprocess the collected data to obtain standard data;

[0107] Mechanical property degradation model construction module: used to extract mechanical property parameters from standard data and establish a mechanical property degradation model for hollow slab beams based on the extracted mechanical property parameters;

[0108] Prediction model building module: used to comprehensively build a prediction model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams;

[0109] Prediction module: Used to monitor the vibration data of hollow slab beams in real time, combine the prediction model to output the mechanical performance degradation of hollow slab beams, provide early warnings, promptly identify potential problems and take corresponding maintenance and remedial measures.

[0110] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0111] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, characterized in that, include: Sensors are deployed on hollow slab beams to collect their vibration data in real time, and the collected data is preprocessed to obtain standard data. The preprocessing of the collected data to obtain standard data includes: Obtain the degraded dataset and the normal dataset, and extract the degraded waveform and the normal waveform; Real-time acquisition of vibration data on hollow slab beams; preprocessing of the acquired vibration data. The preprocessed signal is subjected to Discrete Fourier Transform to analyze the spectral characteristics of the vibration data. Extract positively correlated waveform features based on the degradation type; A first similarity threshold and a second similarity threshold are preset, and a similarity calculation based on deep learning is performed. The extracted waveforms are compared with the degraded waveforms and normal waveforms respectively. If the similarity between two consecutive extracted waveforms and the normal waveform is lower than the first similarity threshold, and the similarity between two consecutive extracted waveforms and the degraded waveform is higher than the second similarity threshold, then it is considered to meet the corresponding degradation type. Mechanical performance parameters were extracted from standard data, and a mechanical performance degradation model for hollow slab beams was established based on the extracted mechanical performance parameters. The establishment of the mechanical property degradation model for hollow slab beams includes: A mathematical simulation model of hollow slab beams is established. Based on the mathematical simulation model of hollow slab beams, the mechanical performance parameters and performance specificities of hollow slab beams are calculated. The mechanical performance parameters include fatigue damage, crack propagation, and material degradation. Positively correlated mechanical performance parameters are selected from the mechanical performance parameters, and performance-specific models corresponding to all positively correlated mechanical performance parameters of hollow slab beams are established. Then, combined with the error model in the mathematical simulation model of hollow slab beams, a mechanical performance degradation model of hollow slab beams is established. Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a prediction model between vibration data and mechanical performance degradation is constructed. The method for constructing a predictive model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams includes: Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a first relationship model between vibration data and mechanical performance parameters is established, and the monitored vibration data is correlated with the mechanical performance degradation model. Vertical bending frequency data is extracted from the vibration data, and a second relationship model is constructed based on the extracted vertical bending frequency and mechanical performance parameters. The extracted vertical bending frequency is then correlated with the mechanical performance degradation model. Finally, a comprehensive prediction model between vibration data and mechanical performance degradation is constructed. The vibration data of hollow slab beams are monitored in real time, and the mechanical performance degradation of hollow slab beams is output in combination with the prediction model, and early warning is given.

2. The method for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency according to claim 1, characterized in that, The method of deploying sensors on the hollow slab beam for real-time acquisition of its vibration data includes: deploying accelerometers and vibration sensors on the hollow slab beam to acquire its vibration data in real time, wherein the vibration data includes vibration frequency.

3. The method for predicting the mechanical property degradation of hollow slab beams based on vibration frequency according to claim 1, characterized in that, The calculated mechanical performance parameters and performance specificities of the hollow slab beam include: After deconstructing the hollow slab beam based on the kinematic model in the mathematical simulation model of the hollow slab beam, the mechanical performance parameters of the hollow slab beam are obtained. Based on the error model and probability error model in the mathematical simulation model of hollow slab beam, the mixed error sources of hollow slab beam are decomposed to obtain all positively correlated mechanical performance parameters of hollow slab beam. The influence of all positively correlated mechanical property parameters of hollow slab beams on hollow slab beams is calculated to obtain the performance specificity of hollow slab beams.

4. The method for predicting the mechanical property degradation of hollow slab beams based on vibration frequency according to claim 1, characterized in that, The process of obtaining the degraded and normal datasets includes: Transform the vibration power spectral density to the amplitude spectrum; The amplitude spectrum is padded with zeros according to the sampling frequency range; Multiple single-frame pseudo-random signals for determining the vibration power spectral density are obtained by performing inverse Fourier transform on the zero-padded amplitude spectrum and the uniformly distributed phase generated according to the congruence method. Multiple single-frame pseudo-random signals are processed by a quadratic B-spline weighting and three-layer superposition merging method to determine a true random vibration signal of a preset duration; Degraded and normal datasets are constructed by extracting multiple frames from true random vibration signals of a preset duration, and the degraded and normal waveforms are extracted.

5. The method for predicting the mechanical property degradation of hollow slab beams based on vibration frequency according to claim 1, characterized in that, The first relationship model between vibration data and mechanical performance parameters includes: Establish a multi-objective integrated optimization model for vibration data and mechanical performance parameters; Establish a database of vibration data and mechanical performance parameters, denote the vibration frequency as the first correlation function, and denote multiple mechanical performance parameters as multiple characteristic constraints to obtain the database of vibration data and mechanical performance parameters; The database is analyzed and compared to understand its operating patterns. The parameters are reweighted to adjust the optimization function. The optimized function, after being adjusted by the database, is fitted to obtain fitted data. The fitted data is then compared with the real-time data to ensure that the monitored real-time vibration data, mechanical performance parameters, and fitted data are close to the set values.

6. The method for predicting the mechanical property degradation of hollow slab beams based on vibration frequency according to claim 1, characterized in that, The comprehensive construction of the prediction model between vibration data and mechanical performance degradation includes comparing the prediction results of the first relationship model with the prediction results of the second relationship model, converting the prediction results of the second relationship model and the prediction results of the first relationship model into vector form, calculating the Euclidean distance between the two, iterating the calculation process until the Euclidean distance between the two is less than a preset threshold, and outputting the prediction result of the first relationship model.

7. A system for predicting the degradation of mechanical properties of hollow slab beams based on vibration frequency, characterized in that, include: Acquisition module: Used to deploy sensors on hollow slab beams to collect their vibration data in real time, and to preprocess the collected data to obtain standard data; The preprocessing of the collected data to obtain standard data includes: Obtain the degraded dataset and the normal dataset, and extract the degraded waveform and the normal waveform; Real-time acquisition of vibration data on hollow slab beams; preprocessing of the acquired vibration data. The preprocessed signal is subjected to Discrete Fourier Transform to analyze the spectral characteristics of the vibration data. Extract positively correlated waveform features based on the degradation type; A first similarity threshold and a second similarity threshold are preset, and a similarity calculation based on deep learning is performed. The extracted waveforms are compared with the degraded waveforms and normal waveforms respectively. If the similarity between two consecutive extracted waveforms and the normal waveform is lower than the first similarity threshold, and the similarity between two consecutive extracted waveforms and the degraded waveform is higher than the second similarity threshold, then it is considered to meet the corresponding degradation type. Mechanical property degradation model construction module: used to extract mechanical property parameters from standard data and establish a mechanical property degradation model for hollow slab beams based on the extracted mechanical property parameters; The establishment of the mechanical property degradation model for hollow slab beams includes: A mathematical simulation model of hollow slab beams is established. Based on the mathematical simulation model of hollow slab beams, the mechanical performance parameters and performance specificities of hollow slab beams are calculated. The mechanical performance parameters include fatigue damage, crack propagation, and material degradation. Positively correlated mechanical performance parameters are selected from the mechanical performance parameters, and performance-specific models corresponding to all positively correlated mechanical performance parameters of hollow slab beams are established. Then, combined with the error model in the mathematical simulation model of hollow slab beams, a mechanical performance degradation model of hollow slab beams is established. Prediction model building module: used to comprehensively build a prediction model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams; The method for constructing a predictive model between vibration data and mechanical performance degradation based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams includes: Based on the influence of different degradation states of mechanical performance parameters on the vibration data of hollow slab beams, a first relationship model between vibration data and mechanical performance parameters is established, and the monitored vibration data is correlated with the mechanical performance degradation model. Vertical bending frequency data is extracted from the vibration data, and a second relationship model is constructed based on the extracted vertical bending frequency and mechanical performance parameters. The extracted vertical bending frequency is then correlated with the mechanical performance degradation model. Finally, a comprehensive prediction model between vibration data and mechanical performance degradation is constructed. Prediction module: Used to monitor the vibration data of hollow slab beams in real time, combine the prediction model to output the mechanical performance degradation of hollow slab beams, and provide early warning.

Images