Sensor-based mountain highway bridge anti-seismic property detection method and sensor-based mountain highway bridge anti-seismic property detection system

Abstract

The invention relates to the technical field of bridge structure health monitoring, and discloses a sensor-based mountain highway bridge anti-seismic performance detection method and system, and the method comprises the steps: applying a local excitation signal to a connection interface between a sensor and a bridge main body structure during a non-earthquake period, and collecting a local response signal generated by the local excitation signal; extracting characteristic parameters from the local response signal, and comparing the characteristic parameters with a preset reference to generate connection state parameters for quantifying the mechanical connection tightness degree of the connection interface; acquiring an earthquake response signal acquired by the sensor during an earthquake, and calibrating the earthquake response signal according to the connection state parameter to obtain a calibrated earthquake response signal; and extracting seismic performance parameters of the bridge from the calibrated seismic response signals. According to the method, the state of the sensor connection interface is evaluated during the non-earthquake period, and the earthquake response signal is calibrated by using the state parameter, so that the accuracy and the reliability of bridge anti-seismic performance evaluation are improved.

Description

Technical Field

[0001] This application relates to the field of bridge structural health monitoring technology, and more specifically, to a sensor-based method and system for detecting the seismic performance of mountain highway bridges. Background Technology

[0002] In the field of structural health monitoring, seismic performance assessment of mountain highway bridges is a crucial step in ensuring traffic safety. Currently, this assessment primarily relies on accelerometers installed at key locations on the bridge to collect vibration responses. However, the harsh operating environment in mountainous areas, such as long-term impacts from heavy vehicles, severe diurnal and seasonal temperature differences, generates continuous mechanical and thermal stresses on the mechanical connection interface between the sensors and the main bridge structure. This leads to gradual and subtle physical changes in the high-strength bolts, gaskets, and other connectors used for fixing, such as a slight decrease in preload and fretting wear, causing the originally tight connection interface to evolve into a non-ideal vibration transmission path with slight "loosening" or "elasticity." This degradation introduces nonlinear vibration energy loss, causing amplitude attenuation and phase shift in the acquired signal; more seriously, the minute relative movements between connectors (such as impacts and friction) generate high-frequency, transient, non-stationary new types of mechanical noise, directly contaminating the actual bridge vibration signal. Traditional digital filters, designed for stable, frequency-concentrated conventional environmental noise (such as wind and traffic loads), have little effect on filtering out these highly variable connection noises. Meanwhile, routine visual inspections and sensor self-tests are ineffective in detecting the mechanical degradation of this type of connection interface, creating a maintenance blind spot. When an earthquake occurs, the seismic response data collected by the monitoring system is actually a mixed signal severely distorted and contaminated with noise. This makes it difficult to accurately extract key seismic performance parameters such as the bridge's true natural frequency and damping ratio, ultimately leading to inaccurate health assessment conclusions. This could result in misjudging the bridge's safety status, delaying maintenance, or implementing unnecessary reinforcement, posing a potential threat to public safety. Therefore, existing technologies urgently need improvement to address these issues. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this application provides a sensor-based method and system for detecting the seismic performance of mountain highway bridges. By assessing the state of the sensor connection interface during non-earthquake periods and using the state parameters to calibrate the seismic response signal, this method solves the problems of inaccurate measurements and data distortion caused by sensor connection issues in existing technologies, thereby improving the accuracy and reliability of bridge seismic performance assessment.

[0004] In a first aspect, this application provides a sensor-based method for detecting the seismic performance of mountain highway bridges, comprising: during non-earthquake periods, applying a local excitation signal to the connection interface between the sensor and the main structure of the bridge, and collecting the local response signal generated by the local excitation signal; extracting feature parameters from the local response signal, and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection of the connection interface; The seismic response signals collected by sensors during an earthquake are acquired, and the seismic response signals are calibrated according to the connection status parameters to obtain the calibrated seismic response signals. Seismic performance parameters of bridges are extracted from calibrated seismic response signals.

[0005] The above scheme enables the quantification of the mechanical connection tightness of the sensor connection interface by performing local excitation and response analysis on the sensor connection interface during non-earthquake periods. This allows for the calibration of the seismic response signal, effectively solving the signal distortion and noise problems caused by sensor connection degradation, improving the accuracy of the seismic response signal, and ultimately achieving a precise assessment of the seismic performance of bridges. This overcomes the shortcomings of inaccurate assessments in existing technologies.

[0006] Furthermore, this application also proposes that, prior to the step of applying a local excitation signal to the connection interface between the sensor and the main bridge structure, the following steps are included: An initial excitation signal is applied to the connection interface, and an initial response signal is acquired; Perform spectral analysis on the initial response signal to identify initial abnormal frequency bands; A local excitation signal is generated based on the initial abnormal frequency band.

[0007] This technical solution enables the identification of abnormal frequency bands at the connection interface through preliminary excitation and spectrum analysis, and generates targeted local excitation signals accordingly. This makes subsequent local response signal acquisition more efficient and accurate, thereby enabling a more precise assessment of the connection interface's state.

[0008] In some preferred embodiments, the step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface includes: Extract the nonlinear damage features of the local response signal. The nonlinear damage features include at least one of the following: harmonic distortion ratio, amplitude-frequency response nonlinear index, and instantaneous frequency modulation features. The nonlinear damage characteristics are compared with a preset benchmark to update the connection state parameters and diagnose the damage type.

[0009] Furthermore, the steps for diagnosing the type of injury include: A damage pattern feature library is pre-established, which contains nonlinear damage features corresponding to different types of microscopic damage. The nonlinear damage features extracted in real time are matched with a damage pattern feature library to diagnose specific damage types.

[0010] Based on the above, this application further proposes that, after the step of comparing nonlinear damage characteristics with a preset benchmark to update connection state parameters and diagnose damage type, the following steps are included: Based on the connection status parameters, determine whether the preset calibration threshold has been reached to initiate calibration and determine the basic calibration amount; The seismic response signal is adaptively calibrated based on the basic calibration values ​​and the diagnosed damage types. The adaptive calibration process includes at least one of amplitude compensation, phase compensation, and nonlinear noise suppression.

[0011] In some implementations, the step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface further includes: When ice crystals are present, identify specific frequency band drops or abnormal high-frequency peaks caused by ice crystals. An adaptive frequency focusing excitation and dynamic amplitude scanning are applied to extract nonlinear ice crystal features, which include at least one of the following: the proportion of subharmonic or superharmonic energy, the area of ​​the hysteresis loop in the amplitude-frequency response, and the response amplitude jump threshold. Based on the characteristics of nonlinear ice crystals, the state of ice crystals is determined, and the connection state parameters are updated according to a predetermined strategy based on the state of ice crystals.

[0012] Preferably, after determining the ice crystal state based on the nonlinear ice crystal characteristics and updating the connection state parameters according to the ice crystal state using a predetermined strategy, the following steps are included: The seismic response signal is calibrated based on the updated connection status parameters and ice crystal status; When the condition is determined to be ice crystal, frequency-selective gain compensation or narrowband filtering is applied to the seismic response signal for specific frequency band collapses or abnormal high-frequency peaks caused by ice crystals. When the connection status parameters indicate an abnormal mechanical connection, overall amplitude compensation and phase compensation are performed on the seismic response signal.

[0013] In one embodiment, the step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection of the connection interface includes: extracting at least one of the following feature parameters from the local response signal: attenuation rate of the local response signal, energy dissipation characteristics, high-frequency harmonic components, and instantaneous frequency change characteristics. Based on the deviation of the feature parameters from the preset benchmark, multiple sub-health scores are calculated; The connection state parameters are obtained by weighted averaging of multiple sub-health scores.

[0014] As a technological improvement, the steps for extracting the seismic performance parameters of bridges from calibrated seismic response signals include: Modal decomposition is performed on the calibrated seismic response signal to obtain the modal response signal; Time-frequency analysis was performed on the modal response signal to obtain the natural frequency and damping ratio of the modal response signal as seismic performance parameters of the bridge.

[0015] Secondly, this application proposes a sensor-based seismic performance testing system for mountain highway bridges, used to execute the aforementioned gate hoist opening stroke data processing method. The system includes: The local response signal acquisition module is used to apply a local excitation signal to the connection interface between the sensor and the main structure of the bridge during non-earthquake periods, and to acquire the local response signal generated by the local excitation signal. The connection status parameter generation module is used to extract feature parameters from the local response signal and compare the feature parameters with a preset benchmark to generate connection status parameters that quantify the tightness of the mechanical connection of the connection interface. The seismic response signal calibration module is used to acquire the seismic response signals collected by sensors during an earthquake, and to calibrate the seismic response signals according to the connection status parameters to obtain the calibrated seismic response signals. The seismic performance parameter extraction module is used to extract the seismic performance parameters of bridges from the calibrated seismic response signal.

[0016] This technical solution enables the effective execution of the aforementioned detection methods through modular design, providing an integrated solution that improves the automation and efficiency of the detection process.

[0017] In summary, this application provides a sensor-based method and system for detecting the seismic performance of mountain highway bridges. The method applies a local excitation signal to the connection interface between the sensor and the main bridge structure and collects the local response signal, effectively identifying and quantifying the mechanical connection tightness of the interface and generating connection status parameters. This innovative step directly addresses the problems of signal distortion and the introduction of new mechanical noise caused by sensor connection degradation in existing technologies. By actively probing the microscopic characteristics of the connection interface, it overcomes the limitations of traditional methods in accurately assessing the connection status. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart illustrating a sensor-based method for detecting the seismic performance of mountain highway bridges, as provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of a sensor-based seismic performance testing system for mountain highway bridges, provided as an embodiment of this application.

[0020] Labeling Explanation: 210, Local Response Signal Acquisition Module; 220, Connection Status Parameter Generation Module; 230, Seismic Response Signal Calibration Module; 240, Seismic Performance Parameter Extraction Module. Detailed Implementation

[0021] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Traditional methods for testing the seismic performance of mountain highway bridges can introduce measurement inaccuracies when the interface between the sensor and the bridge structure undergoes slight degradation. This inaccuracy manifests as nonlinear losses in vibration energy transfer and the generation of new types of mechanical noise. These problems distort the seismic response data acquired by the sensors, failing to accurately reflect the bridge's true seismic performance. Consequently, this affects the accuracy of bridge health assessments, potentially leading to incorrect maintenance decisions and posing a potential risk to the safe operation of the bridge.

[0024] Firstly, please refer to Figure 1 This application proposes a sensor-based method for detecting the seismic performance of mountain highway bridges, including: S1. During non-earthquake periods, apply a local excitation signal to the connection interface between the sensor and the main structure of the bridge, and collect the local response signal generated by the local excitation signal. S2. Extract feature parameters from the local response signal and compare the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface. S3. Acquire the seismic response signal collected by the sensor during the earthquake, and calibrate the seismic response signal according to the connection status parameters to obtain the calibrated seismic response signal; S4. Extract the seismic performance parameters of the bridge from the calibrated seismic response signal.

[0025] To better understand the technical solution proposed in this application, some key terms are explained first. A sensor refers to a device used to sense and measure the vibration response of a bridge, such as an accelerometer or displacement sensor. These sensors are typically mechanically fixed to the main bridge structure, forming a connection interface. A local excitation signal refers to a specific signal actively applied to the connection interface between the sensor and the main bridge structure during non-earthquake periods, with the aim of stimulating a local vibration response at the connection interface. A local response signal refers to the vibration signal collected by the sensor after the application of the local excitation signal, reflecting the mechanical characteristics of the connection interface. Connection state parameters are indicators that quantify the tightness of the mechanical connection at the connection interface; they reflect the health condition of the connection interface, such as the presence of loosening, wear, or corrosion. A seismic response signal refers to the vibration signal of the bridge under seismic loading collected by the sensor during an earthquake. Seismic performance parameters are key indicators for evaluating the seismic resistance of a bridge, such as natural frequency, damping ratio, and mode shape.

[0026] Compared to the closest existing technology, this application has the advantage that existing technologies typically rely on filtering the raw seismic response signal to remove noise, but this filtering method is difficult to effectively address novel, nonlinear mechanical noise generated by interface degradation. This application, by applying local excitation during non-seismic periods and extracting characteristic parameters of the interface, can more precisely quantify the tightness of the mechanical connection at the interface, generating connection state parameters. These parameters directly reflect the distortion and noise sources that may exist during signal transmission. Calibrating the seismic response signal based on these connection state parameters enables more targeted and adaptive signal correction, such as amplitude compensation, phase compensation, or nonlinear noise suppression. This calibration method is far more effective than traditional general filtering methods because it directly addresses the root cause of the problem. Therefore, this application can obtain a calibrated seismic response signal that is closer to the actual signal, resulting in more accurate and reliable subsequently extracted bridge seismic performance parameters (such as natural frequency and damping ratio). This method not only improves the accuracy of bridge health assessments, but also provides a more solid scientific basis for bridge management departments to formulate maintenance and reinforcement strategies. It effectively avoids misjudgments and resource waste caused by inaccurate data, and significantly improves the operational safety and resilience of mountain highway bridges.

[0027] This application proposes applying a local excitation signal to the connection interface between the sensor and the main bridge structure during non-earthquake periods and acquiring the local response signal to generate connection state parameters. However, during implementation, if the applied local excitation signal fails to sufficiently excite potential damage or specific dynamic characteristics of the connection interface, the acquired local response signal may not fully reflect the actual mechanical connection tightness of the connection interface, thus affecting the accuracy of the connection state parameters. To address this, this application further proposes optimizing the generation method of the local excitation signal to improve the targeting and effectiveness of the excitation.

[0028] In this regard, this application further proposes that, prior to the step of applying a local excitation signal to the connection interface between the sensor and the main bridge structure, the following steps are included: An initial excitation signal is applied to the connection interface, and an initial response signal is acquired; Spectral analysis is performed on the preliminary response signal to identify preliminary abnormal frequency bands; The local excitation signal is generated based on the initial abnormal frequency band.

[0029] Specifically, the preliminary excitation signal can be understood as a broadband or swept-frequency excitation, the purpose of which is to comprehensively excite the inherent vibration characteristics and potential damage response of the connection interface. For example, the preliminary excitation signal can take the form of white noise, swept-frequency sine wave, or impact excitation, and be applied to the connection interface through devices such as piezoelectric actuators, electromagnetic vibrators, or impact hammers. The preliminary response signal refers to the vibration response data collected by sensors at the connection interface after the preliminary excitation signal is applied. Spectral analysis of the preliminary response signal refers to converting the time-domain preliminary response signal into a frequency-domain spectrum using signal processing techniques such as Fourier transform. Through spectral analysis, preliminary anomalous frequency bands can be identified. These preliminary anomalous frequency bands are those exhibiting abnormal energy concentration, frequency shift, harmonic distortion, or increased damping in the spectrum. These anomalous characteristics may be related to microscopic damage, loosening, or material degradation of the connection interface. In practical applications, generating the local excitation signal based on the preliminary anomalous frequency bands means designing and generating a local excitation signal with specific frequency components and amplitude characteristics based on the identified preliminary anomalous frequency bands. For example, the frequency of the local excitation signal can be focused on the initial abnormal frequency band to more effectively excite and amplify the damage-related nonlinear response; its amplitude and duration can also be adjusted according to the characteristics of the abnormal frequency band to ensure that the local response signal can be fully acquired.

[0030] In some preferred embodiments, when detecting the connection interface between the sensor and the main structure of a mountain highway bridge, a wideband swept frequency signal is first applied to the connection interface as an initial excitation signal. After the sensor acquires the initial response signal, a Fast Fourier Transform (FFT) is performed on the signal for spectral analysis. In the spectrum, an abnormal energy peak is observed in the 500Hz to 600Hz frequency band, and the harmonic distortion ratio in this band is significantly higher than in other bands, which is identified as the initial abnormal frequency band. Based on this, the system generates a narrowband sinusoidal excitation signal with a center frequency of 550Hz and a bandwidth of 50Hz as a local excitation signal, and applies it to the connection interface with an appropriate amplitude. In this way, the nonlinear response of the connection interface in this specific frequency band can be more effectively excited, thereby obtaining a local response signal containing richer damage information, providing high-quality data for the subsequent generation of connection state parameters.

[0031] In some embodiments described above, this application proposes a method for extracting feature parameters from local response signals and comparing them with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface. However, in practical applications, generating only a comprehensive connection state parameter may not be sufficient to accurately identify the types of microscopic damage or early signs of degradation within the connection interface, thus limiting the ability to conduct refined assessments and early warnings of bridge structural health. Failure to address these issues could lead to misjudgments or delayed responses to potential structural risks in bridges. Therefore, this application further proposes a more refined method for feature parameter extraction and comparison, focusing on nonlinear damage characteristics to more accurately update connection state parameters and diagnose specific damage types.

[0032] The steps described above for extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface include: Extract the nonlinear damage features of the local response signal, wherein the nonlinear damage features include at least one of the following: harmonic distortion ratio, amplitude-frequency response nonlinearity index, and instantaneous frequency modulation features; The nonlinear damage characteristics are compared with the preset benchmark to update the connection state parameters and diagnose the damage type.

[0033] Specifically, nonlinear damage characteristics refer to the nonlinear phenomena that appear in the response signal when a structure is excited, caused by factors such as material damage, poor interfacial contact, or microcracks. These nonlinear phenomena are usually more sensitive to minor damage and can provide richer damage information than linear characteristics. Among them, the harmonic distortion ratio refers to the proportion of energy in the response signal that is an integer multiple of the excitation frequency (harmonic) when a single frequency excitation is applied. An increase in this ratio usually indicates the presence of nonlinear sources within the structure, such as cracks or loosening. Amplitude-frequency response nonlinearity indicators refer to the nonlinear changes exhibited by the structure's amplitude-frequency response curve under different excitation amplitudes, such as drift or jumps in resonant frequencies. These changes can be used to assess the severity of damage. Instantaneous frequency modulation characteristics refer to the properties of the instantaneous frequency of the response signal changing with time or amplitude. When damage exists in the structure, its instantaneous frequency may exhibit abnormal modulation phenomena, reflecting changes in local stiffness or damping.

[0034] In practical applications, the preset benchmark can be understood as a reference value of nonlinear damage characteristics obtained through experiments or simulations when the bridge structure is in a healthy state or a known damaged state. Comparing the real-time extracted nonlinear damage characteristics with the preset benchmark can quantify the degree of damage at the current connection interface and update the connection status parameters based on the comparison results, making them more accurately reflect the tightness of the mechanical connection. Furthermore, by analyzing the specific manifestations of the nonlinear damage characteristics, the type of damage can also be diagnosed, such as whether it is caused by loose connectors, material fatigue cracks, interface peeling, or other microscopic defects.

[0035] Specifically, the above-mentioned steps for diagnosing damage types include: pre-establishing a damage pattern feature library, which contains nonlinear damage features corresponding to different types of microscopic damage; and matching the nonlinear damage features extracted in real time with the damage pattern feature library to diagnose the specific damage type.

[0036] The damage mode feature library is a pre-built database that stores various known microscopic damage types, such as crack initiation, material fatigue, connection loosening, or corrosion, and their corresponding typical nonlinear damage features. These nonlinear damage features may include harmonic distortion ratios, amplitude-frequency response nonlinear indices, or instantaneous frequency modulation characteristics. They are obtained through experimental testing, numerical simulation, or historical data analysis and can uniquely or significantly characterize a specific damage mode. Specifically, during actual detection, when nonlinear damage features are extracted in real time from the local response signal, these real-time features are input into a matching algorithm. The matching algorithm compares the real-time extracted nonlinear damage features with the features stored in the damage mode feature library. By calculating similarity, distance metrics, or other pattern recognition techniques, it can be determined which damage mode in the library the real-time feature is most similar to, thereby diagnosing the specific damage type.

[0037] In some of the embodiments described above in this application, although it is possible to update the connection state parameters and diagnose the damage type by extracting nonlinear damage features, in the implementation process, relying solely on connection state parameters for calibration may not be sufficient to address the complex effects of different damage types on seismic response signals. For example, different types of damage (such as microcracks, interface debonding, or bolt loosening) may cause different modes of amplitude attenuation, phase shift, or nonlinear noise in the seismic response signal. If a single or non-adaptive calibration strategy is used, it may not be able to accurately restore the true state of the signal, thereby affecting the accuracy of the final seismic performance assessment.

[0038] In response, this application further proposes that after comparing the nonlinear damage characteristics with the preset benchmark to update the connection state parameters and diagnose the damage type, the method further includes: Based on the connection status parameters, determine whether a preset calibration threshold has been reached to initiate calibration and determine the basic calibration amount; The seismic response signal is adaptively calibrated based on the basic calibration values ​​and the diagnosed damage type. The adaptive calibration process includes at least one of amplitude compensation, phase compensation, and nonlinear noise suppression.

[0039] Specifically, after acquiring the connection status parameters indicating the tightness of the mechanical connection at the quantification interface, the system first determines whether these parameters have reached a preset calibration threshold. This preset calibration threshold can be an empirical value or a critical value trained using historical data, used to indicate whether the connection interface needs calibration and the urgency of the calibration. If the connection status parameters exceed or fall below this threshold, the calibration process is initiated, and a base calibration amount is determined based on the degree of deviation of the connection status parameters. This base calibration amount serves as the initial adjustment benchmark for the adaptive calibration process.

[0040] Furthermore, adaptive calibration is a process of fine-tuning the seismic response signal based on the determined baseline calibration values ​​and previously diagnosed damage types. For example, if microcracks are diagnosed at the interface and the connection status parameters indicate slight loosening, adaptive calibration might focus on specific amplitude compensation and nonlinear noise suppression of the seismic response signal. Amplitude compensation aims to correct signal energy attenuation or gain anomalies caused by interface loosening or damage; phase compensation corrects time delays or phase distortions caused by changes in the signal transmission path; and nonlinear noise suppression filters out nonlinear noise components introduced by damage or nonlinear behavior to improve signal purity. These compensation and suppression measures can be used individually or in combination to maximize the fidelity of the seismic response signal.

[0041] In some preferred embodiments, assuming that during non-earthquake periods, slight bolt loosening is detected at the connection interface between a sensor and the main bridge structure through local excitation signal acquisition and analysis, and the damage type is diagnosed as "interface fretting wear" through nonlinear damage feature analysis, the connection status parameters may indicate that the tightness is below a preset calibration threshold, thereby initiating the calibration process and determining a basic calibration amount, for example, setting it to compensate for the signal amplitude by 5%. During an earthquake, the seismic response signal acquired by the sensor is obtained. Based on the previously diagnosed "interface fretting wear" damage type and the determined basic calibration amount, the system performs adaptive calibration processing. Specifically, since fretting wear may cause signal energy dissipation and increased high-frequency noise, adaptive calibration processing may include applying amplitude compensation to the seismic response signal to recover energy, while simultaneously performing nonlinear noise suppression to filter out nonlinear noise components generated by wear. Furthermore, if fretting wear also causes slight phase lag, phase compensation can be further applied. Through this targeted adaptive calibration, the calibrated seismic response signal can more accurately reflect the true dynamic behavior of the bridge under seismic loading, providing high-quality input data for subsequent extraction of seismic performance parameters.

[0042] In some embodiments described above, nonlinear damage characteristics of local response signals are extracted and compared with a preset benchmark to update connection status parameters and diagnose damage types. However, in the actual operating environment of mountainous highway bridges, especially in cold seasons, the connection interface between the sensor and the main bridge structure may be affected by ice crystals. The formation of ice crystals alters the local mechanical properties of the connection interface, potentially leading to non-damaging anomalies in the local response signal, thus affecting the accuracy of connection status parameters and reducing the reliability of subsequent seismic response signal calibration and seismic performance parameter extraction. If these problems are not addressed, the health status of the connection interface may be misjudged in the presence of ice crystals, or deviations may be introduced into the calibration of the seismic response signal. Therefore, this application further proposes a scheme to identify specific signal characteristics caused by ice crystals and apply adaptive excitation to extract nonlinear ice crystal characteristics when ice crystals are present, thereby determining the ice crystal state and updating the connection status parameters, to improve the accuracy of the connection status parameters and the environmental adaptability of the detection method.

[0043] The steps described above, which involve extracting feature parameters from local response signals and comparing these feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface, further include: When ice crystals are present, identify specific frequency band drops or abnormal high-frequency peaks caused by ice crystals. An adaptive frequency focusing excitation and dynamic amplitude scanning are applied to extract nonlinear ice crystal features, which include at least one of the following: subharmonic or superharmonic energy ratio, amplitude-frequency response hysteresis loop area, and response amplitude jump threshold. Based on the nonlinear ice crystal characteristics, the ice crystal state is determined, and the connection state parameters are updated according to a predetermined strategy based on the ice crystal state.

[0044] Specifically, when ice crystals form at the interface between the sensor and the main bridge structure, they alter the local stiffness and damping characteristics of this interface, causing specific frequency response changes in the local response signal generated by the local excitation signal. These changes may manifest as specific frequency band dips caused by ice crystals, i.e., a significant reduction in response energy within certain frequency ranges, or as abnormal high-frequency peaks, i.e., energy concentration in normally inactive high-frequency regions. These characteristics can be identified through spectral analysis or time-frequency analysis of the local response signal.

[0045] To more accurately extract nonlinear features related to ice crystals, adaptive frequency focusing excitation and dynamic amplitude scanning can be applied. Adaptive frequency focusing excitation involves adjusting the frequency of the excitation signal based on the initially identified anomalous frequency bands caused by ice crystals, concentrating its energy in these specific frequency bands to enhance the response to the ice crystal effect. Dynamic amplitude scanning involves gradually changing the amplitude of the excitation signal within a certain frequency range to observe the nonlinear behavior of the response signal. In this way, nonlinear ice crystal features can be extracted, such as the proportion of subharmonic or superharmonic energy, reflecting the frequency components generated when the signal propagates in a nonlinear medium; the area of ​​the amplitude-frequency response hysteresis loop, which quantifies the degree of nonlinear hysteresis between excitation and response; and the response amplitude jump threshold, indicating the critical point at which the response signal amplitude abruptly changes under a specific excitation amplitude. These nonlinear features can more sensitively and accurately characterize the influence of ice crystals on the mechanical properties of the interface.

[0046] In practical applications, the ice crystal state can be determined based on the extracted nonlinear ice crystal characteristics. For example, by setting a threshold, when the proportion of subharmonic or superharmonic energy, the area of ​​the amplitude-frequency response hysteresis loop, or the response amplitude jump threshold exceeds a preset value, it can be determined that ice crystals exist at the connection interface. Based on the determined ice crystal state, the connection state parameters are updated using a predetermined strategy. The predetermined strategy may include correcting the connection state parameters, such as reducing or compensating for the connection state parameters calculated through nonlinear damage characteristics when ice crystals are present, to eliminate false anomalies caused by ice crystals and ensure that the connection state parameters can truly reflect the tightness of the mechanical connection.

[0047] To further confirm and quantify the impact of ice crystals, the system applies adaptive frequency focusing excitation, concentrating the energy of the excitation signal into the identified anomalous high-frequency band and performing dynamic amplitude scanning. During this process, the system monitors the nonlinear characteristics of the response signal. For example, when the excitation amplitude reaches a certain threshold, the amplitude of the response signal exhibits a significant jump, while a significant increase in the proportion of subharmonic energy is detected, and a distinct hysteresis loop is observed on the amplitude-frequency response curve. These nonlinear ice crystal characteristics are extracted.

[0048] Based on these extracted nonlinear ice crystal features, the system determines that the current connection interface is in an ice crystal state. For example, if the proportion of subharmonic energy exceeds a preset threshold of 0.05 and the area of ​​the amplitude-frequency response hysteresis loop is greater than a preset 0.1 unit, ice crystals are confirmed to be present. According to a predetermined strategy, the system corrects the connection state parameters previously calculated using nonlinear damage features (such as harmonic distortion ratio). Specifically, if ice crystals cause the connection interface to exhibit characteristics similar to loosening, but are not actually structural damage, the system compensates for the connection state parameters based on the quantitative characteristics of the ice crystals. For example, it adjusts them to be closer to the healthy state value without ice crystals, or introduces an ice crystal influence factor to weight the connection state parameters. In this way, the final generated connection state parameters can accurately reflect the mechanical connection tightness of the connection interface, eliminating interference from ice crystals and providing a more reliable basis for subsequent seismic response signal calibration.

[0049] Traditional sensor-based methods for detecting the seismic performance of mountain highway bridges involve applying local excitation signals and acquiring local response signals during non-earthquake periods. Feature parameters are then extracted and compared with a preset benchmark to generate connectivity parameters, which are used to calibrate the seismic response signal. While these methods can identify ice crystal formation and update connectivity parameters, in practical applications, the presence of ice crystals or abnormal mechanical connections can have complex and specific effects on the accuracy of the seismic response signal. Relying solely on a single, general connectivity parameter for calibration may not adequately compensate for specific signal distortions caused by ice crystals or mechanical connection anomalies, resulting in deviations in the calibrated seismic response signal and consequently affecting the accuracy of the extracted bridge seismic performance parameters.

[0050] In response, this application proposes a more refined calibration strategy, which aims to perform targeted calibration of the seismic response signal based on the updated connectivity parameters and ice crystal state, so as to improve the accuracy of calibration and the reliability of seismic performance assessment.

[0051] The steps described above, including determining the ice crystal state based on the nonlinear ice crystal characteristics and updating the connection state parameters according to the ice crystal state using a predetermined strategy, include: The seismic response signal is calibrated based on the updated connection status parameters and the ice crystal state; When the state is determined to be ice crystal, frequency-selective gain compensation or narrowband filtering is applied to the seismic response signal for specific frequency band collapses or abnormal high-frequency peaks caused by the ice crystals. When the connection status parameter indicates an abnormal mechanical connection, the overall amplitude and phase compensation are performed on the seismic response signal.

[0052] Specifically, after obtaining the updated connection status parameters and ice crystal status, the seismic response signal will be calibrated based on this information. When the system determines that it is currently in an ice crystal state, since ice crystals may cause specific frequency band collapses or abnormally high-frequency peaks in the seismic response signal, frequency-selective gain compensation or narrowband filtering will be applied to the seismic response signal to accurately eliminate these interferences introduced by ice crystals. Frequency-selective gain compensation refers to adjusting the signal amplitude within a specific frequency range to compensate for energy loss caused by ice crystals or suppress abnormal gain; narrowband filtering refers to precisely filtering out or enhancing specific frequency band signals caused by ice crystals by designing filters with specific bandwidths, thereby eliminating their interference with the seismic response signal.

[0053] Furthermore, when connection status parameters indicate anomalies in the mechanical connection between the sensor and the main bridge structure, such as loosening, wear, or local failure, these anomalies typically cause shifts in the overall amplitude and phase of the seismic response signal. To correct these systematic deviations, overall amplitude compensation and phase compensation are performed on the seismic response signal. Overall amplitude compensation aims to adjust the overall signal strength to restore it to normal levels; phase compensation is used to correct for time lag or lead in the signal, ensuring signal accuracy.

[0054] When a subsequent earthquake occurs, the sensors acquire seismic response signals. At this point, the system calibrates the seismic response signals based on the previously determined ice crystal state. Specifically, for the 200Hz-300Hz frequency band collapse, the system applies frequency-selective gain compensation, increasing the signal amplitude in this band to compensate for energy loss; simultaneously, for anomalous high-frequency peaks in the 1kHz-1.2kHz range, the system employs narrowband filtering to precisely filter out anomalous signals in this frequency band.

[0055] Specifically, in the aforementioned sensor-based method for detecting the seismic performance of mountain highway bridges, the step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface can be further refined.

[0056] The step of extracting feature parameters from the local response signal and comparing the feature parameters with the preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface includes: Extract at least one of the following characteristic parameters from the local response signal: attenuation rate, energy dissipation characteristics, high-frequency harmonic components, and instantaneous frequency change characteristics of the local response signal; Based on the deviation between the feature parameters and the preset benchmark, multiple sub-health scores are calculated; The connection status parameter is obtained by weighted averaging of the multiple sub-health scores.

[0057] The decay rate of the local response signal refers to the rate at which the amplitude of the local response signal decays over time after the application of a local excitation signal. This parameter reflects the energy dissipation capacity and structural damping characteristics at the interface. For example, it can be obtained by calculating the descent slope or half-life of the response signal envelope. The energy dissipation characteristics refer to the ability of the interface to convert mechanical energy into other forms of energy under excitation. This characteristic is closely related to the friction, damping, and micro-damage of the interface. Specifically, it can be characterized by calculating the ratio of the energy integral of the response signal to the excitation energy over one excitation cycle. The high-frequency harmonic components refer to the nonlinear high-frequency components that appear in the local response signal when a local excitation signal of a specific frequency is applied. These high-frequency harmonics are usually caused by nonlinear contact, friction, or micro-cracks at the interface, and their amplitude and frequency distribution can sensitively indicate the tightness and damage status of the interface. For example, they can be extracted by performing a Fourier transform on the local response signal and analyzing the energy or amplitude at integer multiples of the fundamental frequency. The instantaneous frequency change characteristics refer to the frequency fluctuations or drifts of the local response signal over a short timescale. Loosening or damage to the connection interface can lead to changes in local stiffness, which in turn causes a change in the instantaneous frequency of the response signal. This characteristic can be extracted using time-frequency analysis methods such as Hilbert-Huang transform (HHT) or wavelet transform.

[0058] Further, based on the deviation between the feature parameters and the preset benchmark, multiple sub-health scores are calculated. The preset benchmark is a reference value or range established in advance using test data of the connection interface in a healthy or known state. Each extracted feature parameter (such as decay rate, energy dissipation characteristics, etc.) is compared with the corresponding preset benchmark, and an independent sub-health score is calculated according to its degree of deviation. For example, the larger the deviation, the lower the sub-health score, and vice versa. Subsequently, the multiple sub-health scores are weighted and averaged to obtain the connection state parameter. The purpose of the weighted average is to comprehensively consider the contribution and sensitivity of different feature parameters to the tightness of the mechanical connection of the connection interface. For example, the weight of each sub-health score can be determined based on experience, expert knowledge, or machine learning methods, thereby more comprehensively and accurately quantifying the overall health status of the connection interface.

[0059] Specifically, the steps for extracting the seismic performance parameters of bridges from the calibrated seismic response signal include: Modal decomposition is performed on the calibrated seismic response signal to obtain the modal response signal; time-frequency analysis is performed on the modal response signal to obtain the natural frequency and damping ratio of the modal response signal as the seismic performance parameters of the bridge.

[0060] Modal decomposition is a method that breaks down complex vibration responses into a series of simple modal vibrations, each with a specific natural frequency and damping ratio. Through modal decomposition, the complex overall seismic response signal of a bridge can be effectively separated into individual vibration modes, thus more clearly identifying the dynamic characteristics of the bridge under seismic loading. The modal response signal refers to the response signal obtained after modal decomposition, representing each vibration mode of the bridge.

[0061] Furthermore, time-frequency analysis is an analytical method that can simultaneously reveal the changes of a signal in the time and frequency domains. Performing time-frequency analysis on modal response signals can accurately capture the changes in the natural frequencies and damping ratios of a bridge over time during an earthquake. Natural frequencies are the inherent vibration frequencies of a structure during free vibration, reflecting its stiffness characteristics; damping ratios quantify the structure's energy dissipation capacity during vibration, reflecting its energy dissipation characteristics. Using these natural frequencies and damping ratios as seismic performance parameters of a bridge allows for a comprehensive and accurate assessment of its dynamic behavior and seismic resistance under seismic loading.

[0062] Secondly, please refer to Figure 2 This application proposes a sensor-based seismic performance testing system for mountainous highway bridges, aiming to provide a concrete implementation platform capable of efficiently and accurately executing the aforementioned sensor-based seismic performance testing method for mountainous highway bridges. Through modular design, the system transforms each key step in the testing method into operable system components, thereby achieving automation and standardization of the testing process.

[0063] Specifically, the sensor-based seismic performance testing system for mountain highway bridges in this application includes: The local response signal acquisition module 210 is used to apply a local excitation signal to the connection interface between the sensor and the main structure of the bridge during non-earthquake periods, and to acquire the local response signal generated by the local excitation signal. The connection status parameter generation module 220 is used to extract feature parameters from the local response signal and compare the feature parameters with a preset benchmark to generate connection status parameters that quantify the tightness of the mechanical connection of the connection interface. The seismic response signal calibration module 230 is used to acquire the seismic response signal collected by the sensor during the earthquake, and to calibrate the seismic response signal according to the connection status parameters to obtain the calibrated seismic response signal. The seismic performance parameter extraction module 240 is used to extract the seismic performance parameters of the bridge from the calibrated seismic response signal.

[0064] The local response signal acquisition module 210 can be configured to include an excitation source and one or more sensors. The excitation source is responsible for generating and applying a local excitation signal to the connection interface; for example, it can be a piezoelectric exciter, an electromagnetic exciter, or a mechanical vibrator. The sensors are used to acquire the local response signal generated at the connection interface by the local excitation signal in real time; these sensors can be accelerometers, strain gauges, or displacement sensors, etc. This module provides raw data for subsequent connection status assessment by precisely controlling the application of the excitation signal and the acquisition of the response signal.

[0065] The connection state parameter generation module 220 can be implemented as a data processing unit that receives local response signals from the local response signal acquisition module. This module integrates signal processing algorithms and comparison logic, enabling it to extract characteristic parameters reflecting the tightness of the mechanical connection at the connection interface from the local response signal, such as nonlinear damage characteristics, attenuation rate, and energy dissipation characteristics. The extracted characteristic parameters are then compared with preset benchmark values ​​to quantify the current state of the connection interface and generate connection state parameters. This module can also further diagnose potential damage types or ice crystal states, thus providing a refined basis for subsequent seismic response signal calibration.

[0066] The seismic response signal calibration module 230 is designed to acquire the raw seismic response signal collected by sensors during an earthquake. This module calibrates the raw seismic response signal based on the connection status parameters provided by the connection status parameter generation module. The calibration process may include amplitude compensation, phase compensation, or nonlinear noise suppression, with the aim of eliminating or reducing measurement errors introduced by changes in the interface between the sensor and the bridge structure (e.g., loosening, damage, or the influence of ice crystals), ensuring the accuracy of the seismic response signal.

[0067] The seismic performance parameter extraction module 240 is responsible for receiving the calibrated seismic response signal. This module uses advanced signal processing techniques such as modal decomposition and time-frequency analysis to extract the bridge's seismic performance parameters from the calibrated seismic response signal, including natural frequencies, damping ratios, and mode shapes. These parameters are key indicators for assessing the structural health and seismic resistance of the bridge.

[0068] As a specific implementation, in a practical application scenario of a mountainous highway bridge, this sensor-based seismic performance detection system for mountainous highway bridges can be deployed at key structural parts of the bridge. During non-earthquake periods, the local response signal acquisition module periodically applies weak local excitation signals to the connection interface between the sensors installed on the bridge and the main bridge structure, and collects the resulting local response signals. For example, a local excitation test can be performed once a month. Subsequently, the connection status parameter generation module analyzes these local response signals, extracts nonlinear damage characteristics such as harmonic distortion ratios, and compares them with a preset health benchmark to generate connection status parameters that quantify the tightness of the connection. If loosening or micro-damage is detected at the connection interface, this parameter will reflect the corresponding anomaly. When an earthquake occurs, the seismic response signal calibration module immediately acquires the seismic response signals collected by the sensors. At this time, the module performs adaptive calibration processing on the seismic response signals based on the previously generated connection status parameters. For example, if the connection status parameters indicate slight loosening of the sensor connection, corresponding amplitude compensation and phase compensation will be applied to eliminate signal distortion caused by loosening. Ultimately, the seismic performance parameter extraction module extracts seismic performance parameters such as the bridge's natural frequency and damping ratio from the calibrated seismic response signal, providing bridge management departments with an accurate seismic performance assessment report. In this way, even with minor changes in sensor connection status, the system can ensure the accuracy of seismic response data, thereby avoiding misjudgments caused by sensor coupling issues.

[0069] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A sensor-based method for detecting the seismic performance of mountain highway bridges, characterized in that, include: During non-earthquake periods, a local excitation signal is applied to the connection interface between the sensor and the main structure of the bridge, and the local response signal generated by the local excitation signal is collected. Feature parameters are extracted from the local response signal and compared with a preset benchmark to generate connection status parameters that quantify the tightness of the mechanical connection of the connection interface. The seismic response signal collected by the sensor during the earthquake is acquired, and the seismic response signal is calibrated according to the connection status parameters to obtain the calibrated seismic response signal; The seismic performance parameters of the bridge are extracted from the calibrated seismic response signal.

2. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 1, characterized in that, The procedure further includes, prior to the step of applying a local excitation signal to the connection interface between the sensor and the main bridge structure: An initial excitation signal is applied to the connection interface, and an initial response signal is acquired; Spectral analysis is performed on the preliminary response signal to identify preliminary abnormal frequency bands; The local excitation signal is generated based on the initial abnormal frequency band.

3. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 1, characterized in that, The step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface includes: Extract the nonlinear damage features of the local response signal, wherein the nonlinear damage features include at least one of the following: harmonic distortion ratio, amplitude-frequency response nonlinearity index, and instantaneous frequency modulation features; The nonlinear damage characteristics are compared with the preset benchmark to update the connection state parameters and diagnose the damage type.

4. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 3, characterized in that, The steps for diagnosing the type of injury include: A damage mode feature library is pre-established, which contains nonlinear damage features corresponding to different types of microscopic damage. The nonlinear damage features extracted in real time are matched with the damage pattern feature library to diagnose the specific damage type.

5. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 3, characterized in that, The step following the comparison of the nonlinear damage characteristics with the preset benchmark to update the connection state parameters and diagnose the damage type includes: Based on the connection status parameters, determine whether a preset calibration threshold has been reached to initiate calibration and determine the basic calibration amount; The seismic response signal is adaptively calibrated based on the basic calibration values ​​and the diagnosed damage type. The adaptive calibration process includes at least one of amplitude compensation, phase compensation, and nonlinear noise suppression.

6. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 3, characterized in that, The step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface further includes: When ice crystals are present, identify specific frequency band drops or abnormal high-frequency peaks caused by ice crystals. An adaptive frequency focusing excitation and dynamic amplitude scanning are applied to extract nonlinear ice crystal features, which include at least one of the following: subharmonic or superharmonic energy ratio, amplitude-frequency response hysteresis loop area, and response amplitude jump threshold. Based on the nonlinear ice crystal characteristics, the ice crystal state is determined, and the connection state parameters are updated according to a predetermined strategy based on the ice crystal state.

7. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 6, characterized in that, The step of determining the ice crystal state based on the nonlinear ice crystal characteristics and updating the connection state parameters according to the ice crystal state using a predetermined strategy includes: The seismic response signal is calibrated based on the updated connection status parameters and the ice crystal state; When the state is determined to be ice crystal, frequency-selective gain compensation or narrowband filtering is applied to the seismic response signal for specific frequency band collapses or abnormal high-frequency peaks caused by the ice crystals. When the connection status parameter indicates an abnormal mechanical connection, the overall amplitude and phase compensation are performed on the seismic response signal.

8. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 1, characterized in that, The step of extracting feature parameters from the local response signal and comparing the feature parameters with a preset benchmark to generate connection state parameters that quantify the tightness of the mechanical connection at the connection interface includes: Extract at least one of the following characteristic parameters from the local response signal: attenuation rate, energy dissipation characteristics, high-frequency harmonic components, and instantaneous frequency change characteristics of the local response signal; Based on the deviation between the feature parameters and the preset benchmark, multiple sub-health scores are calculated; The connection status parameter is obtained by weighted averaging of the multiple sub-health scores.

9. The sensor-based method for detecting the seismic performance of mountain highway bridges according to claim 1, characterized in that, The step of extracting the seismic performance parameters of the bridge from the calibrated seismic response signal includes: Modal decomposition is performed on the calibrated seismic response signal to obtain the modal response signal; Time-frequency analysis is performed on the modal response signal to obtain the natural frequency and damping ratio of the modal response signal as the seismic performance parameters of the bridge.

10. A sensor-based seismic performance testing system for mountain highway bridges, used to execute the sensor-based seismic performance testing method for mountain highway bridges as described in any one of claims 1 to 9, characterized in that, The system includes: The local response signal acquisition module is used to apply a local excitation signal to the connection interface between the sensor and the main structure of the bridge during non-earthquake periods, and to acquire the local response signal generated by the local excitation signal. The connection status parameter generation module is used to extract feature parameters from the local response signal and compare the feature parameters with a preset benchmark to generate connection status parameters that quantify the tightness of the mechanical connection of the connection interface. The seismic response signal calibration module is used to acquire the seismic response signal collected by the sensor during an earthquake, and to calibrate the seismic response signal according to the connection status parameters to obtain the calibrated seismic response signal. The seismic performance parameter extraction module is used to extract the seismic performance parameters of the bridge from the calibrated seismic response signal.

Images