Method for detecting anti-seismic performance of anti-seismic support structure based on modal analysis

Abstract

The invention discloses a method for detecting the anti-seismic performance of an anti-seismic support structure based on modal analysis, and relates to the technical field of anti-seismic detection of engineering structures. The method comprises the following steps: S1, synchronously acquiring acceleration time domain signals and real-time surface temperature data of a to-be-tested anti-seismic support under natural environment excitation; s2, decomposing the acceleration signal through a variational mode decomposition algorithm, and screening out a main intrinsic mode component; s3, performing multi-scale permutation entropy analysis on the principal component, and extracting singularity features reflecting internal microscopic damage; s4, correcting the real-time frequency through a temperature influence compensation model to obtain a standard frequency value without temperature interference; and S5, taking the standard frequency offset and the singularity characteristic change rate as macro and micro indexes, and comprehensively judging the anti-seismic performance grade of the support through the dual-index weight matrix. According to the method, the early microscopic damage of the support can be accurately identified, the performance grading judgment can be realized, the detection result is accurately matched with the working condition, and a scientific and reliable technical support is provided for anti-seismic support detection and engineering safety maintenance.

Description

Technical Field

[0001] This invention relates to the field of seismic testing technology for engineering structures, and more specifically, to a method for testing the seismic performance of seismic bearing structures based on modal analysis. Background Technology

[0002] Seismic bearings are core seismic-resistant components in engineering structures such as bridges and high-rise buildings. Their seismic performance directly determines the overall safety of the structure under seismic loads. Therefore, conducting seismic performance testing is a crucial step in the safety maintenance and earthquake disaster prevention of engineering structures. Currently, the industry primarily uses modal analysis to test the dynamic characteristics and seismic performance of seismic bearing structures. The mainstream approach involves obtaining bearing acceleration signals through artificial vibration and then analyzing the signals to extract macroscopic indicators such as frequency, which are then used as the basis for judging the seismic performance of the bearings. This method has been applied to a certain extent in engineering testing.

[0003] However, in practical use, it still has some drawbacks. The existing detection method, which relies on manual vibration, is not only cumbersome in operation and inefficient in on-site testing, but also easily causes mechanical disturbance to the support and auxiliary structures, affecting the authenticity of the test results. At the same time, the detection process does not consider the interference of temperature changes on the stiffness and frequency identification of the support, resulting in deviations in the extracted structural dynamic characteristics and insufficient detection accuracy. Moreover, the traditional method only relies on macroscopic frequency indicators to determine performance, ignoring the identification of early microscopic damage inside the support, which can easily lead to missed damage detection and cannot fully reflect the true seismic performance state of the support, making it difficult to meet the actual detection needs of seismic maintenance in engineering. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for detecting the seismic performance of seismic bearing structures based on modal analysis. The method addresses the problems of artificial vibration disturbance, temperature interference deviation, and failure to detect damage by a single indicator as mentioned in the background art through the following solutions.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for detecting the seismic performance of a seismic bearing structure based on modal analysis, comprising: S1: synchronous data acquisition: acquiring the time-domain acceleration signal of the seismic bearing under test under natural environmental excitation, and synchronously acquiring the real-time surface temperature data of the seismic bearing under test; S2: Feature component extraction: The acceleration time-domain signal is decomposed using the variational mode decomposition algorithm to obtain the intrinsic mode components, and the principal intrinsic mode components reflecting the dynamic characteristics of the seismic support structure under test are selected from them. S3: Microscopic feature identification: Perform multi-scale permutation entropy analysis on the principal intrinsic mode components, and extract singularity features reflecting the microscopic damage inside the seismic bearing under test by calculating the signal complexity of the principal intrinsic mode components at different time scales. S4: Environmental effect compensation: Identify the real-time frequency from the main intrinsic mode components, use a preset temperature influence compensation model, and combine the real-time surface temperature data to correct the real-time frequency to obtain the standard frequency value after eliminating temperature interference. S5: Comprehensive performance assessment: The offset of the standard frequency value is used as a macroscopic performance evaluation index, and the singularity feature is used as a microscopic structure evaluation index. The seismic performance level of the seismic bearing to be tested is comprehensively determined through a dual-index weight matrix.

[0006] The technical effects and advantages of this invention are as follows: 1. This invention relies on natural environmental excitation to complete the acquisition of acceleration signals without the need to apply artificial excitation force. This simplifies the on-site testing process, improves testing efficiency, and avoids mechanical disturbance to the support and auxiliary structures caused by artificial excitation. It ensures that the test results are consistent with the actual working conditions of the support and is suitable for seismic support testing operations in various engineering sites. 2. This invention introduces a temperature influence compensation model, which, combined with synchronously acquired surface temperature data, corrects the real-time frequency identification, effectively eliminating the interference of temperature environment changes on support stiffness and frequency identification, and improving the accuracy of structural dynamic feature extraction. At the same time, it constructs a macro-micro dual-index evaluation system, using standard frequency offset to reflect the macro performance of overall support stiffness degradation, and singularity features to reflect the micro-damage state inside the support, avoiding the limitations of single index judgment and realizing comprehensive detection of support performance. 3. This invention achieves accurate identification of early microscopic damage inside seismic bearings through multi-scale permutation entropy analysis, overcoming the shortcomings of traditional methods that only focus on macroscopic indicators and miss early damage. At the same time, a four-level performance judgment system is constructed based on a dual-index weight matrix. According to the comprehensive judgment index and the source of index contribution, it accurately distinguishes four states of bearing performance: good performance, early microscopic damage, certain performance degradation, and severe performance degradation. This provides a graded and targeted decision-making basis for the seismic maintenance of engineering structures and effectively reduces the safety hazards of engineering structures caused by bearing performance failure. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a schematic diagram of the comprehensive performance evaluation process of the present invention. Detailed Implementation

[0008] 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.

[0009] As attached Figure 1 and attached Figure 2 The seismic performance testing method for seismic bearing structures based on modal analysis, as shown, includes: S1: Data Synchronous Acquisition: Acquire the acceleration time-domain signal of the seismic bearing under test under natural environmental excitation, and synchronously acquire the real-time surface temperature data of the seismic bearing under test. It should be noted that the specific process of data synchronization and collection is as follows: Acceleration time-domain signal acquisition: A high-precision triaxial accelerometer with an accuracy of no less than ±0.001g was selected as the acquisition device. This sensor was rigidly fixed at the corresponding connection positions on the lower surface of the cap beam and the upper surface of the pier of the seismic support to be tested. During sensor installation, it was ensured that the sensing axis precisely coincided with the main shear deformation direction of the support. The installation contact surfaces were ground and cleaned, and secured with dedicated fasteners to avoid signal acquisition distortion caused by gaps in the contact surfaces or loose installation. After sensor installation, power-on testing was performed to confirm that the sensor had no signal drift, poor contact, or other abnormalities. Based on the detection... The system measures micro-vibrations in the field, as well as natural environmental micro-vibrations caused by surrounding traffic loads and natural wind loads. Without applying artificial excitation force, the system starts the acceleration signal acquisition program and continuously acquires the support acceleration signal at a sampling frequency of 1000Hz to generate the original acceleration time-domain signal. The acquired original acceleration time-domain signal is then subjected to mean reduction processing frame by frame. That is, the mean value within each frame of the original acceleration signal is calculated, and then the mean value within each frame is subtracted from each sampling point within the frame. This eliminates the DC deviation caused by the zero-point drift of the sensor, and finally obtains the preprocessed acceleration time-domain signal x(t).

[0010] Real-time surface temperature data acquisition: An infrared temperature sensor with a measurement accuracy of no less than ±0.5℃ was selected as the temperature acquisition device. The sensor was mounted directly in front of the seismic support under test using a fixed bracket. The vertical distance between the sensor's detection end and the support surface was controlled within the range of 0.5m-1.0m. The detection angle was adjusted so that the infrared detection area of ​​the sensor completely covered the effective surface of the rubber body of the support. After the sensor was installed, calibration and debugging were performed to confirm that there were no temperature jumps or blind spots. At the same time as starting the acceleration signal acquisition program, the infrared temperature sensor acquisition program was started to achieve time synchronization with the acceleration time domain signal acquisition. The temperature sampling frequency was kept consistent with the acceleration signal sampling frequency. During the acquisition process, it was ensured that the support surface was unobstructed, free from strong convective airflow, and free from close-range radiation from external heat sources. The support surface temperature value was continuously acquired, and finally, real-time surface temperature data T, which corresponded one-to-one with the timestamp of the acceleration time domain signal x(t) and met the acquisition accuracy standard, was obtained. real .

[0011] S2: Feature component extraction: The acceleration time-domain signal is decomposed using the variational mode decomposition algorithm to obtain the intrinsic mode components, and the principal intrinsic mode components reflecting the dynamic characteristics of the seismic support structure under test are selected from them. It should be specifically noted that the intrinsic mode components are obtained in the following ways: The Variational Mode Decomposition (VMD) algorithm is invoked, with the preprocessed acceleration time-domain signal x(t) used as the sole input signal. Before starting the decomposition operation, the core parameters of the algorithm are preset to ensure decomposition accuracy and avoid mode aliasing. The number of decomposition layers K is determined based on the model and specifications of the seismic bearing under test, the structural dimensions, and the actual signal complexity of the acceleration time-domain signal x(t). The value range is 3-8 layers. For natural rubber bearings (low stiffness, signal frequency 0.5~5Hz), 3-5 layers are initially selected; for neoprene rubber bearings (high stiffness, signal frequency 3~10Hz), 5-8 layers are initially selected. Through multiple sets of trial calculations, the decomposition is verified to ensure that the center frequencies of each component do not overlap and the correlation coefficient ρ of the principal components is consistent. max ≥0.8 is the criterion for judgment, and the optimal number of decomposition layers K that meets the criterion is selected; The penalty factor α is set to 1000-5000 and dynamically adjusted according to the signal amplitude of x(t). When the signal amplitude is <0.01g, it is set to 3000~5000, and when the signal amplitude is ≥0.01g, it is set to 1000~3000. The main component frequency domain energy ratio is ≥70% as the criterion to ensure the sparsity and independence of each intrinsic mode component after decomposition. The iterative convergence criterion is set to the intrinsic mode component error obtained between two consecutive iterations being less than 10. -6 The maximum number of iterations is set to 1000 to avoid decomposition failure or insufficient decomposition accuracy due to non-convergence of iterations.

[0012] The VMD algorithm is initiated to decompose the signal. The core of this decomposition process is to solve a pre-defined constrained variational problem, the expression of which is as follows: , where {u k} represents the set of K eigenmode components obtained after decomposition, u k (t) represents the k-th intrinsic mode component (k=1,2,...,K), {ω k} represents the set of center frequencies corresponding to each eigenmode component, ω k Let ∂ be the center frequency of the k-th intrinsic mode component (in rad / s). t δ(t) represents the partial derivative with respect to time t, δ(t) is the Dirac function, j is the imaginary unit, and * represents the convolution operation. The square of the L2 norm is represented by st, which indicates the solution to {u k} and {ω k In the process of minimizing the objective function, it is necessary to satisfy the following conditions: This constraint condition; by solving the constraint variational problem, the goal is to minimize the sum of the estimated bandwidths of each eigenmode component, thereby decomposing the input acceleration time-domain signal x(t) into K independent narrowband eigenmode components u with non-overlapping center frequencies. k (t), after the decomposition operation is completed, all intrinsic mode components u are output synchronously. k (t) and its corresponding center frequency ω k This means acquiring the intrinsic modal components.

[0013] It should be further explained that the screening process for the principal intrinsic mode components is as follows: Calculation of correlation coefficient: The Pearson correlation coefficient calculation method is used to calculate the u of each intrinsic mode component obtained from the decomposition. k The correlation coefficient ρ between x(t) and the preprocessed acceleration time-domain signal x(t) of the original input k (k=1,2,...,K), during the calculation, retain 4 significant figures after the decimal point to obtain K corresponding correlation coefficients ρ. k Correlation coefficient ρ k The value range of is [-1, 1]. The closer its absolute value is to 1, the stronger the correlation between the intrinsic mode component and the original acceleration time-domain signal, and the better it can reflect the core characteristics of the original signal.

[0014] Determine the screening frequency range: Based on the factory design parameters, model specifications, and relevant bridge engineering seismic testing standards of the seismic bearing to be tested, determine the design fundamental frequency range of the seismic bearing to be tested, preset to f. design ±0.5Hz (where f design(Design fundamental frequency for the support, in Hz); simultaneously, the intrinsic mode components u k (t) corresponds to the center frequency ω k (Unit: rad / s) Convert to corresponding frequency f k (Unit: Hz), conversion formula is f k =ω k / (2π)

[0015] Screening of principal intrinsic mode components: Select the unique intrinsic mode component that simultaneously satisfies two core conditions. Condition 1: The correlation coefficient ρ of this intrinsic mode component. k For all K correlation coefficients ρ k The maximum value in (denoted as ρ) max Condition 1: Ensure its strongest correlation with the original acceleration signal; Condition 2: The frequency f corresponding to this intrinsic mode component. k The frequency falls within the preset design fundamental frequency range of the seismic bearing under test, ensuring that it can truly reflect the structural dynamic characteristics of the bearing itself, and eliminating irrelevant interference components such as environmental noise and overall structural vibration. The intrinsic modal components that simultaneously satisfy the above two conditions are determined as the principal intrinsic modal components reflecting the structural dynamic characteristics of the seismic bearing under test, denoted as u. main (t), synchronously record its corresponding center frequency ω main and correlation coefficient ρ max This means completing the screening of the principal intrinsic mode components.

[0016] S3: Microscopic feature identification: Perform multi-scale permutation entropy analysis on the principal intrinsic mode components, and extract singularity features reflecting the microscopic damage inside the seismic bearing under test by calculating the signal complexity of the principal intrinsic mode components at different time scales. It should be noted that the specific process of microscopic feature recognition is as follows: Multi-scale coarse-grained processing is performed on the principal intrinsic mode components to achieve signal complexity analysis at different time scales: the preset scale factor τ has a value range of 1≤τ≤10, and the maximum scale factor τ is determined. max =10, scale factor τ ranges from 1 to τ max The values ​​are taken sequentially, with each scaling factor corresponding to an analytical scale; for each preset scaling factor τ, the principal intrinsic mode components u are... main (t) Divide the time series evenly into several non-overlapping time windows of length τ, calculate the mean of all signal sampling points within each time window, and use this mean as a sampling point of the coarse-grained time series. Repeat this process for all time windows to obtain the coarse-grained time series u corresponding to each scale factor τ. τ (m), where m is the sampling point number of the coarse-grained time series, m=1,2,...,M, and M is the length of the coarse-grained time series. (N is the original principal eigenmode component u) main The total number of sampling points (t), (For rounding down), and M decreases as the scale factor τ increases, ultimately obtaining 10 sets of coarse-grained time series at different scales, corresponding to analysis scales from τ=1 to τ=10.

[0017] Calculate the permutation entropy of the coarse-grained time series at each scale to quantify the signal complexity at different time scales: for each scale factor τ, the coarse-grained time series u τ (m) represents the core parameter for calculating the permutation entropy, with the embedding dimension d ranging from 3 to 5, and the delay time τ. d The value ranges from 1 to 3, based on the principal eigenmode components u. main The signal length and frequency characteristics of (t) were used as the core criterion for trial calculation (the embedding dimension was selected based on the minimum value of entropy that tends to be stable, and the delay time was selected based on the maximum value of entropy). The optimal embedding dimension d=4 and delay time τ were determined through trial calculation. d =2; Using the permutation entropy calculation method, each coarse-grained time series u τ (m) By embedding dimension d and delay time τ d Reconstruct the phase space to obtain several phase space vectors. Sort each phase space vector in ascending order according to the numerical value of its internal elements and assign a corresponding arrangement number. Calculate the probability of different arrangement numbers. Calculate the arrangement entropy value H(τ) at this scale according to the information entropy formula. Keep 4 significant digits after the decimal point during the calculation. Complete the arrangement entropy calculation for all scale factors τ (1≤τ≤10) in sequence. Finally, obtain 10 arrangement entropy values ​​and form the arrangement entropy sequence {H(1), H(2), ..., H(10)}.

[0018] By introducing health benchmark data and calculating comprehensive singularity features, the feature extraction of microscopic damage inside the seismic bearing under test is achieved: A health benchmark permutation entropy sequence {H} of the same model and batch as the seismic bearing under test is introduced. base (1), H base (2),...,H base (10)}, the health baseline permutation entropy sequence H base(τ) Obtain data in a priority order: Prioritize obtaining the reference data of the same batch of bearings from the manufacturer's calibration report; if there is no manufacturer data, select ≥3 healthy bearings from the same batch that have not been in service or have been in service for ≤1 year, and take the arithmetic mean of each point after testing them using the same method; if there is no measured data, generate a health status signal by simulating the finite element model of the same size and material, and then calculate it using the same method; this sequence represents the result of the bearings in a healthy state without damage when they are manufactured or initially installed, and corresponds one-to-one with the scale factor τ of the current test, and the coefficient of variation of the entropy value of each scale must be ≤8% to verify the validity, which can truly reflect the benchmark level of signal complexity when the bearings have no microscopic damage.

[0019] The comprehensive singularity feature S is calculated using the following formula. This feature quantifies the degree of nonlinear distortion of the principal eigenmode component signals, thereby reflecting the microscopic damage state inside the seismic bearing under test. The formula is as follows: Where, τ max =10 is the maximum scale factor, H(τ) is the permutation entropy value corresponding to the scale factor τ in the current detection, H base (τ) represents the permutation entropy value corresponding to the scaling factor τ under the healthy baseline state. This represents the summation of the absolute values ​​of all differences between τ from 1 to 10; After the calculation is completed, the accuracy of the comprehensive singularity feature S is checked to ensure that the error of the calculation result is less than 10. -4 Finally, a comprehensive singularity feature S with satisfactory accuracy is obtained. The magnitude of this feature value is negatively correlated with the continuity of the internal structure of the seismic bearing under test. That is, the larger the S value, the worse the continuity of the internal structure of the bearing and the more obvious the micro-damage. Thus, the extraction of singular features reflecting the micro-damage inside the seismic bearing under test is completed.

[0020] S4: Environmental effect compensation: Identify the real-time frequency from the main intrinsic mode components, use a preset temperature influence compensation model, and combine the real-time surface temperature data to correct the real-time frequency to obtain the standard frequency value after eliminating temperature interference. It should be noted that the specific process for compensating for the environmental effects is as follows: Identify the real-time frequency f from the principal eigenmode components. real The Fast Fourier Transform (FFT) algorithm is used to analyze the principal eigenmode components u. main (t) Perform frequency domain analysis, preset the core parameters of the FFT algorithm, and set the number of sampling points to 2. 14 (i.e., 16384 points), the calculated frequency resolution is f. res =1000 / 16384≈0.061Hz, ensuring frequency identification accuracy; for u main (t) Fill with zeros up to 2 14After processing the 16384 points, an FFT operation is initiated to convert the time-domain signal into a frequency-domain signal, obtaining the corresponding power spectral density function. By searching for peak points in the power spectral density function, interference peaks with power spectral density values ​​less than 10% of the maximum peak value are removed (to avoid frequency identification errors caused by environmental noise). The frequency corresponding to the maximum peak value is determined as the real-time frequency f of the seismic support under test. real The calculation process retains three significant figures after the decimal point, ultimately obtaining the real-time surface temperature data T. real Real-time frequency sequence {f} synchronized in time dimension real,1 ,f real,2 ,...,f real,n}, where n is the total number of data points collected.

[0021] Determine the preset temperature influence compensation model and core parameters: Based on the properties of the rubber main material (natural rubber or neoprene rubber) of the seismic bearing to be tested, a preset temperature influence compensation model is established. This model is based on the linear relationship between the bearing rubber stiffness and temperature change, and its specific expression is as follows: f std =f real ×[1+β×(T ref -T real )] where f std The standard frequency value (in Hz) after eliminating temperature interference, f real To identify the obtained real-time frequency (unit: Hz), β is the temperature coefficient of the bearing rubber material (unit: 1 / ℃), with a value range of 0.002~0.005 / ℃. Based on the rubber material model and factory test parameters of the seismic bearing under test, the optimal value was determined through multiple sets of temperature-frequency calibration tests. The calibration test temperature range covered -10℃ to 60℃, consistent with the actual engineering environment temperature range; T ref The preset reference temperature (unit: °C) is uniformly set to 20 °C (the standard indoor reference temperature, which meets the requirements of the seismic testing code for bridge engineering). real This is the real-time surface temperature data collected (unit: °C).

[0022] Substitute real-time surface temperature data into standard frequency calculation: Calculate the identified real-time frequency f real Real-time surface temperature data T collected real And the determined temperature coefficient β, reference temperature T ref The calculations were performed by substituting each point into the temperature effect compensation model, strictly adhering to the accuracy requirements of the four arithmetic operations, retaining four significant decimal places for each calculation step to avoid cumulative errors; for the real-time frequency sequence {f real,1 ,f real,2 ,...,f real,n} and the corresponding real-time surface temperature sequence {Treal,1 ,T real,2 ,...,T real,n The compensation calculation is performed point by point to obtain the corresponding standard frequency sequence {f}. std,1 ,f std,2 ,...,f std,n}

[0023] Accuracy verification of standard frequency values: The calculated standard frequency sequence is smoothed using a moving average filtering method with a filter window length of 5 points. Abnormal fluctuations in standard frequency values ​​are eliminated (when the deviation of a standard frequency at a certain point from the average value of its 5-point moving average window is greater than 0.1Hz, it is considered an outlier and replaced using linear interpolation); after verification, it is ensured that the calculation error of all standard frequency values ​​is less than 10. -4 The arithmetic mean of the standard frequency sequence is calculated to obtain the standard frequency value f that eliminates temperature interference and meets the accuracy requirements. std The standard frequency value can accurately reflect the inherent frequency characteristics of the seismic bearing under test at the reference temperature, eliminating the interference of temperature changes on the bearing stiffness and frequency identification, thus completing the environmental effect compensation process.

[0024] S5: Comprehensive performance assessment: The offset of the standard frequency value is used as a macroscopic performance evaluation index, and the singularity feature is used as a microscopic structure evaluation index. The seismic performance level of the seismic bearing to be tested is comprehensively determined through a dual-index weighting function.

[0025] It should be noted that the specific process for the comprehensive performance assessment is as follows: Calculate the relative change of the standard frequency to provide macroscopic calculation parameters for the construction of comprehensive judgment indicators: using the support health benchmark frequency f init For reference, calculate the standard frequency value f. std The relative change is calculated using the following formula: The calculation process retains 4 significant figures after the decimal point. The value reflects the degree of change of the current macroscopic natural frequency of the seismic bearing under test relative to its healthy state, and indirectly reflects the degradation of the overall stiffness of the bearing.

[0026] A comprehensive judgment index G is constructed and calculated to achieve a weighted fusion evaluation of macroscopic and microscopic indicators: Based on the relative change of macroscopic frequency and the comprehensive singularity characteristic S of microscopic indicators, a dimensionless comprehensive judgment index G is constructed. This index, through the mathematical logic of weighted square root extraction, achieves a comprehensive quantification of the changes in the macroscopic performance of the support and the microscopic damage state, taking into account the influence weights of both indicators. The specific calculation formula is as follows: Wherein, λ is the weighting coefficient, ranging from 0.5 to 0.7. Considering the engineering characteristics of seismic bearings, the preset optimal value is 0.6. This value takes into account both the core influence of macroscopic frequency characteristics on the overall seismic performance of the bearing and the potential risk of damage to the internal structure of the bearing due to microscopic singularity characteristics; f std To eliminate temperature interference, the standard frequency value, f init The healthy frequency of the bearing of the same model and batch as the bearing under test at the time of manufacture or initial installation is used. S is the comprehensive singularity characteristic reflecting the internal micro-damage of the bearing. Each step of the calculation retains 4 significant figures after the decimal point, and finally obtains the comprehensive judgment index G that meets the accuracy standard. The larger the G value, the more significant the degradation of the bearing's seismic performance and the more serious the combined effect of micro-damage and macro-stiffness changes.

[0027] To provide a quantitative basis for performance grading, preset thresholds and frequency term change criteria were established. Based on factory testing standards for seismic bearings, seismic maintenance specifications for bridge engineering, and extensive engineering measurement data, a dimensionless first preset threshold γ1=0.05 and a second preset threshold γ2=0.2 were established. This threshold division, combined with the actual impact of bearing performance degradation, can accurately distinguish different performance states of the bearings. Simultaneously, the criterion for "small frequency term change" was clarified, i.e., when the absolute value of the relative frequency change... When the frequency term changes little, the increase in the comprehensive judgment index G is mainly contributed by the microscopic singularity characteristic S term, reflecting that the macroscopic performance of the support has not deteriorated significantly, and the performance change mainly stems from early internal microscopic damage.

[0028] A comprehensive assessment process is implemented to determine seismic performance based on the G value and the source of contribution from the indicators. Using the comprehensive assessment indicator G, the first preset threshold γ1, the second preset threshold γ2, and the frequency term change assessment criteria as the basis, the seismic performance of the tested seismic bearings is assessed in four categories. The assessment process strictly adheres to the principle of matching the source of indicator contribution with the degree of performance degradation to ensure the accuracy and guidance of the assessment results. If the comprehensive judgment index G is less than the first preset threshold γ1, i.e. G<γ1, then the seismic performance of the seismic bearing to be tested is judged to be good. At this time, the macroscopic frequency characteristics of the bearing are basically consistent with the healthy state, there is no obvious microscopic damage inside, the overall seismic performance meets the requirements of engineering use, and it can be put into normal use and tested according to the regular cycle. If the comprehensive judgment index G exceeds the first preset threshold γ1, but the absolute value of the relative change in frequency... , that is, G > γ1 and the frequency term changes little. It is comprehensively determined that the increase in the comprehensive determination index G is mainly caused by the microscopic singularity feature S term. Then it is determined that there is early microscopic damage in the anti-seismic bearing to be tested. At this time, the macroscopic stiffness of the bearing has not shown obvious degradation, and the overall anti-seismic performance has not been significantly affected for the time, but microscopic damage signs have appeared inside. It is necessary to conduct a local inspection of the bearing, focus on detecting the damage conditions of key parts such as the rubber main body and the internal stiffening layer, and shorten the inspection cycle to track the development trend of microscopic damage; If the comprehensive determination index G is between the first preset threshold γ1 and the second preset threshold γ2, and the absolute value of the relative change in frequency , that is, γ1 < G < γ2 and the frequency term shows obvious changes. Then it is determined that the anti-seismic performance of the anti-seismic bearing to be tested has degraded to a certain extent. At this time, the macroscopic stiffness of the bearing has undergone detectable degradation, and there is accompanying internal microscopic damage. The two together lead to a decline in performance. It is necessary to conduct a comprehensive inspection of the bearing, evaluate the irreversibility of stiffness degradation, and formulate maintenance or reinforcement measures according to the inspection results; If the comprehensive determination index G exceeds the second preset threshold γ2, that is, G > γ2, then it is determined that the anti-seismic performance of the anti-seismic bearing to be tested has seriously degraded. At this time, the macroscopic frequency characteristics of the bearing have changed significantly, and the internal microscopic damage has reached a relatively serious level. The combined effect of the two causes the anti-seismic performance of the bearing to not meet the engineering use requirements, and there are serious structural safety hazards. It is recommended to immediately take measures to replace the bearing to avoid anti-seismic accidents of the bridge structure caused by the failure of the bearing performance.

[0029] Secondly: In the drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments of the present disclosure are involved. For other structures, the usual design can be referred to. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other; Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting the seismic performance of seismic bearing structures based on modal analysis, characterized in that, include: S1: Data Synchronous Acquisition: Acquire the acceleration time-domain signal of the seismic bearing under test under natural environmental excitation, and synchronously acquire the real-time surface temperature data of the seismic bearing under test. S2: Feature component extraction: The acceleration time-domain signal is decomposed using the variational mode decomposition algorithm to obtain the intrinsic mode components, and the principal intrinsic mode components reflecting the dynamic characteristics of the seismic support structure under test are selected from them. S3: Microscopic feature identification: Perform multi-scale permutation entropy analysis on the principal intrinsic mode components, and extract singularity features reflecting the microscopic damage inside the seismic bearing under test by calculating the signal complexity of the principal intrinsic mode components at different time scales. S4: Environmental effect compensation: Identify the real-time frequency from the main intrinsic mode components, use a preset temperature influence compensation model, and combine the real-time surface temperature data to correct the real-time frequency to obtain the standard frequency value after eliminating temperature interference. S5: Comprehensive performance assessment: The offset of the standard frequency value is used as a macroscopic performance evaluation index, and the singularity feature is used as a microscopic structure evaluation index. The seismic performance level of the seismic bearing to be tested is comprehensively determined through a dual-index weight matrix.

2. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The specific method for synchronous data collection is as follows: Acceleration time domain signal: Acquired by a high-precision triaxial accelerometer with an accuracy of not less than ±0.001g. The sensor is rigidly fixed to the corresponding connection position between the lower surface of the cap beam and the upper surface of the pier of the seismic support to be tested. After power-on debugging, the original signal is continuously acquired at a sampling frequency of 1000Hz. The original signal is then subjected to mean reduction processing to obtain the preprocessed acceleration time domain signal. Real-time surface temperature data: collected by an infrared temperature sensor with an accuracy of no less than ±0.5℃. The sensor is set up in front of the seismic support to be tested, and the vertical distance between the probe end and the support surface is controlled within the range of 0.5m-1.0m. After calibration and debugging, it is collected synchronously with the acceleration time domain signal at the same frequency to obtain real-time surface temperature data corresponding one-to-one with the timestamp of the acceleration time domain signal.

3. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The intrinsic mode components are obtained as follows: the preprocessed acceleration time-domain signal is used as the input signal of the variational mode decomposition algorithm. The preset algorithm decomposition layer is 3-8 layers, the penalty factor is 1000-5000, and the iterative convergence criterion is that the error of the intrinsic mode components obtained in two adjacent iterations is less than 10. -6 The upper limit of the number of iterations is 1000. The algorithm is started to solve the constrained variational problem, decomposes the input acceleration time-domain signal into narrowband intrinsic mode components that are independent of each other and whose center frequencies do not overlap, and outputs all intrinsic mode components and their corresponding center frequencies simultaneously.

4. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The selection method for the principal intrinsic modal components is as follows: using the Pearson correlation coefficient calculation method, the correlation coefficient between each intrinsic modal component and the preprocessed acceleration time-domain signal is calculated, the center frequency corresponding to each intrinsic modal component is converted into the corresponding frequency, and combined with the preset selection frequency range of the design fundamental frequency range of the seismic bearing to be tested, the intrinsic modal components with the maximum correlation coefficient and the corresponding frequency falling within the selection frequency range are selected as the principal intrinsic modal components reflecting the structural dynamic characteristics of the seismic bearing to be tested.

5. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The process of multi-scale permutation entropy analysis is as follows: The principal intrinsic mode components are subjected to multi-scale coarse-graining processing. The preset scale factor range is 1≤τ≤10. The principal intrinsic mode components are uniformly divided into several non-overlapping time windows of length τ according to the time series. The mean value of the signal sampling points within each time window is calculated to obtain the coarse-grained time series corresponding to each scale factor. The preset embedding dimension for permutation entropy calculation is 3-5, and the delay time is 1-3. The phase space of each coarse-grained time series is reconstructed according to the embedding dimension and delay time. The elements within the phase space vector are arranged in ascending order and assigned permutation numbers. The probability of different permutation numbers is statistically analyzed. The permutation entropy value at each scale is calculated according to the information entropy formula to form a permutation entropy sequence.

6. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The singularity feature is extracted as follows: A healthy baseline permutation entropy sequence of the same model and batch as the seismic bearing to be tested is introduced. This sequence corresponds one-to-one with the scale factor of the detected permutation entropy sequence. Based on the standard that the maximum scale factor is 10, the absolute values ​​of the differences between the detected permutation entropy sequence and the healthy baseline permutation entropy sequence are summed and averaged to obtain the comprehensive singularity feature. The accuracy of this feature is verified to ensure that the error of the calculation result is less than 10. -4 This comprehensive singularity feature is the singularity feature that reflects the microscopic damage inside the seismic bearing to be tested.

7. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The temperature influence compensation model is based on the linear relationship between the stiffness of the bearing rubber and temperature. The reference temperature in the model is uniformly set to 20℃, and the temperature coefficient of the bearing rubber material ranges from 0.002 to 0.005 / ℃. The real-time frequency and the collected real-time surface temperature data are substituted into the model for point-by-point calculation. The calculation results are processed by moving average filtering, and after removing abnormal fluctuation values, the arithmetic mean is calculated to obtain the standard frequency value after eliminating temperature interference.

8. The method for detecting the seismic performance of seismic support structures based on modal analysis according to claim 1, characterized in that: The method for determining the seismic performance level is as follows: Using the healthy reference frequency of the same model and batch of healthy bearings at the time of manufacture or initial installation as a reference, the relative change in the standard frequency value is calculated. The weight coefficient of the preset dual-index weighting function is set to a range of 0.5~0.7, with an optimal value of 0.

6. A comprehensive judgment index is constructed by combining the relative change in the standard frequency with singularity characteristics. The first preset threshold for the comprehensive judgment index is 0.05, and the second preset threshold is 0.

2. Simultaneously, the absolute value of the relative change in frequency is set to ≤0.02 as the criterion for minimal change in the frequency term. Based on the comparison results between the comprehensive judgment index and the threshold, and the change in the frequency term, the seismic performance level is determined into four categories: good performance, early microscopic damage, some degree of performance degradation, and severe performance degradation.

Images