Bridge safety monitoring system, node and method based on differential symmetric self-diagnosis

Abstract

This invention discloses a bridge safety monitoring system, node, and method based on differential symmetry self-diagnosis. The system includes a node comprising a mounting base, a micro-exciter, symmetrically arranged first and second reference resonators, a bridge response sensor, and a processing module. The method includes: applying homogeneous excitation through the micro-exciter; synchronously acquiring the response signals of the dual reference resonators and the bridge; calculating an installation health index characterizing the symmetry of the installation state based on the response signals of the dual reference resonators; and using this index to decouple the bridge response signal and separate structural damage characteristics. This invention, by constructing a symmetrical differential measurement system within the node, transforms the installation coupling state, which is difficult to observe directly, into a calculable electrical signal symmetry index, fundamentally solving the problem of false alarms caused by the confusion between installation drift and structural damage in traditional monitoring, and significantly improving the reliability and operational accuracy of the monitoring system.

Description

Technical Field

[0001] This invention belongs to the field of bridge structural health monitoring technology, specifically relating to a long-term, online safety monitoring system and method for large-scale civil engineering structures such as urban bridges. Background Technology

[0002] During the long-term service of urban bridges, continuous monitoring of parameters such as stress, vibration, and displacement of their key components is a routine technical objective for assessing structural safety and achieving predictive maintenance. To achieve this objective, existing technologies generally employ the fixed installation of sensors such as accelerometers and strain gauges at key parts of the bridge, collecting data over a long period and setting thresholds or using modal analysis to determine the structural health status.

[0003] However, those skilled in the art have discovered in practice that existing technical solutions suffer from the following long-standing defects and bottlenecks that have plagued the industry. For example, false alarms due to damage are a prominent issue. Because sensor nodes are fixed to bridge components for extended periods using bolts or adhesives, long-term environmental temperature cycles, salt spray corrosion, and traffic vibrations inevitably lead to problems such as loosening of bolt preload, aging of adhesives, and localized corrosion of the base. This causes a slow drift in the coupling state between the sensor and the structure. The signal characteristics generated by this installation drift in the monitoring data are highly similar to those caused by early structural damage (such as microcracks or loose connections). Existing diagnostic methods based on thresholds or statistical models cannot effectively distinguish between the two, resulting in a large number of false alarms, severely consuming maintenance resources and reducing system reliability. Furthermore, the lack of in-situ and online diagnostic capabilities for the installation status means that when the monitoring system issues an alarm, maintenance personnel cannot determine from the data level whether the alarm originates from actual damage to the bridge structure or a sensor installation malfunction. Currently, the only way to distinguish between the two is to send personnel to the site for physical inspection of the sensor nodes. This is not only costly and slow in response, but also extremely difficult to conduct inspections for large bridge complexes or sensors installed in inaccessible locations such as high altitudes or underwater. Although some studies have attempted to compensate for installation drift by adding temperature sensors or using complex algorithms, these are essentially post-hoc interpretations of mixed signals and fail to provide independent observations of the installation state from the bottom layer of the measurement system. The reliance on the long-term stability of the sensors themselves is too high. Some schemes that attempt to introduce reference sensors for compensation rely heavily on the long-term absolute accuracy and stability of the reference sensors themselves. In the harsh outdoor environment of bridges, ensuring that the performance of a single sensor does not drift for years or even decades is extremely difficult and costly in engineering practice, which restricts the large-scale application of such technologies. In addition, A. Rama Mohan Rao et al., in their paper "ASensor Fault Detection Algorithm for Structural Health Monitoring Using Adaptive Differential Evolution," proposed that in large sensor networks, the computation time required to isolate faulty sensors using existing algorithms is unacceptable for online sensor fault isolation. This paper combines a principal component analysis (PCA)-based algorithm with an adaptive differential evolution algorithm to improve the performance of sensor fault isolation. This approach still falls under the category of achieving isolation through algorithm analysis based on monitored sensor data and requires a high level of remote computing power.

[0004] The aforementioned existing technologies all point to a more fundamental and hidden technical problem: in long-term, unattended monitoring scenarios, there is a lack of a mechanism built into the monitoring node itself that can directly quantify the sensor-structure installation coupling state online, robustly, and at low cost, and physically separate it from the mixed signal. This leaves structural health monitoring systems facing persistent dilemmas of high false alarm rates and unexplainable maintenance, severely hindering their development towards intelligence and precision.

[0005] Therefore, addressing the long-standing and unresolved core pain point of confusion between installation drift and structural damage in existing bridge monitoring technologies, developing a new type of monitoring node capable of self-diagnosis of installation status and signal decoupling has become a critical technical bottleneck that urgently needs to be overcome in this field, and is clearly necessary and urgent. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a bridge monitoring system and method based on symmetrical differential self-diagnosis. By constructing a physically symmetrical dual-reference resonator measurement system inside the node and directly quantifying the installation status by utilizing the symmetrical changes in its response signal, the system achieves physical identification and separation of installation faults and structural damage, ultimately significantly reducing false alarm rates and achieving precise operation and maintenance.

[0007] The solution to the technical problem of this invention is as follows: a bridge safety monitoring system based on differential symmetry self-diagnosis is adopted, including one or more monitoring nodes, a convergence unit, and a remote management platform; the monitoring nodes are used to install on various test parts of the bridge, and each monitoring node includes: a mounting base for fixed connection with the bridge component; a micro-exciter disposed on the mounting base; a first reference resonator and a second reference resonator, both with identical structural parameters, symmetrically arranged on the mounting base and decoupled from the bridge component; and a bridge response sensor disposed on the mounting base; the processing module is configured to: control the micro-exciter to apply excitation; determine an installation health index characterizing the symmetry of the installation state based on the response signals of the first and second reference resonators; and process the output signal of the bridge response sensor based on the installation health index to separate structural features reflecting the bridge structural state; the convergence unit is communicatively connected to one or more monitoring nodes for aggregating data from each node; the remote management platform is communicatively connected to the convergence unit for receiving data, presenting the installation health index and the structural features, and generating maintenance decisions according to preset rules.

[0008] Preferably, the installation health index is determined based on the coherence function values ​​of the response signals of the first reference resonator and the second reference resonator within a predetermined frequency band.

[0009] Preferably, the aggregation unit is further configured to perform spatiotemporal correlation analysis based on the installation health index and structural characteristics of multiple monitoring nodes to locate abnormal areas of the bridge or identify damage patterns.

[0010] A node for bridge safety monitoring includes: a mounting base for fixed connection to a bridge component; a micro-exciter disposed on the mounting base; a first reference resonator and a second reference resonator with identical structural parameters, symmetrically arranged on the mounting base and decoupled from the bridge component; a first sensing unit and a second sensing unit for acquiring response signals from the first and second reference resonators, respectively; a bridge response sensor for acquiring response signals from the bridge component; and a processing module connected to the micro-exciter, the first sensing unit, the second sensing unit, and the bridge response sensor. The processing module is configured to: control the micro-exciter to apply excitation; determine an installation health index characterizing the symmetry of the mounting base's installation state based on the response signals acquired by the first and second sensing units; and process the response signals acquired by the bridge response sensor based on the installation health index to extract structural features related to damage to the bridge component.

[0011] Preferably, the first reference resonator and the second reference resonator are mounted on the mounting base via vibration damping elements.

[0012] Preferably, the processing module is further configured to: synchronously acquire the signal energy of the bridge response sensor during the calibration period; when the signal energy exceeds the energy threshold determined based on historical data, determine that the installation health index during the period is affected by environmental interference, and mark it.

[0013] A bridge safety monitoring method based on differential symmetry self-diagnosis is applied to the system described above. The method includes: applying excitation through micro-exciters deployed at one or more monitoring nodes at various parts of the bridge; synchronously acquiring the response signals of a first reference resonator, a second reference resonator, and a bridge response sensor at each monitoring node; calculating the installation health index of each node based on the response signals of the first and second reference resonators; decoupling the bridge response signals of the corresponding nodes based on the installation health index of each node to extract the structural features of each part; aggregating the installation health index and structural features of each node and uploading them to a management platform; and performing a fusion analysis on the management platform based on the installation health index and structural features to generate maintenance suggestions for the sensor installation status and assessment results for the bridge structural safety.

[0014] Preferably, the calculation of the installation health index of each node specifically includes: converting the response signals of the first reference resonator and the second reference resonator of each node to the frequency domain; calculating the coherence function of the two within a preset analysis frequency band; and using the average value of the coherence function within the preset analysis frequency band as the installation health index of the node.

[0015] Preferably, after the initial installation of the node, an initialization step is performed, which includes: applying excitation and acquiring initial response signals under conditions of no significant environmental interference; calculating and storing an initial installation health index based on the initial response signals of the first and second reference resonators; wherein, the installation health index obtained by subsequent calculation is compared with the initial installation health index to determine the trend of installation status change.

[0016] Preferably, the process of decoupling the bridge response signal of the corresponding node based on the installation health index of each node includes: when the installation health index of a node is higher than a first preset threshold, the node is determined to be in a reliable installation state, and the structural features extracted from its bridge response signal are directly used in the bridge safety assessment; when the installation health index of a node is lower than a second preset threshold, the node is determined to be in an abnormal installation state, a maintenance alarm is triggered for the node, and the extracted structural features are downweighted or required to be reviewed.

[0017] The beneficial effects of this invention are as follows: 1. By introducing a physically symmetrical dual-reference resonator differential measurement structure, this invention achieves, for the first time, direct, in-situ observation of the installation coupling state at the node level. By calculating the real-time installation health index, the system can clearly distinguish whether signal changes originate from installation faults or structural damage. In simulation tests, this solution can reduce false structural alarms caused by loose installation by more than 70%. Through dual-channel alarm splitting for installation deterioration and structural damage, the reliability of the system's structural warning output achieves a qualitative leap.

[0018] 2. It simplifies the complex process of traditional manual on-site inspections. The quantitative installation health index output by the nodes provides maintenance personnel with clear action guidelines, enabling predictive maintenance. This design reduces the stringent requirements for the long-term absolute accuracy of individual sensors, allowing the use of mature, low-cost commercial resonators. Implementation shows that while ensuring high performance, node hardware costs can be optimized, and by reducing ineffective inspections, it is expected to improve related maintenance efficiency by more than 30%, achieving a complete closed-loop optimization of the monitoring system from data acquisition to maintenance actions.

[0019] 3. The symmetrical differential principle endows the system with excellent common-mode rejection capability. Factors affecting both channels simultaneously, such as ambient temperature changes and exciter performance degradation, are largely offset in the differential calculation, thus ensuring that the installation health index remains stable over a wide temperature range of -20℃ to 60℃. This provides a stable and reliable underlying sensing platform for subsequent technology iterations (such as adapting to more complex structural forms, fusing with multiple sensors, and developing more advanced diagnostic algorithms). Attached Figure Description

[0020] Figure 1 This is the overall system architecture diagram of Example 1.

[0021] Figure 2 This is a schematic diagram of the structural cross-section of the monitoring node.

[0022] Figure 3 This is a block diagram of the node circuit connection.

[0023] Figure 4 This is a flowchart of the overall monitoring method.

[0024] Figure 5 This is a flowchart of the node's workflow.

[0025] Figure 6 This is a schematic diagram illustrating the change in the installation health index (Is) as the installation torque decreases.

[0026] Figure 7 This is a schematic diagram of the assembly relationship of the monitoring nodes. Detailed Implementation

[0027] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the principles of the present invention without inventive effort are within the scope of protection of the present invention.

[0028] In the long-term structural health monitoring of urban bridges, widely deployed distributed sensor networks face a fundamental challenge: the installation status of sensor nodes can unpredictably drift due to environmental factors such as temperature cycling, vibration, and corrosion. This installation drift generates spurious signals in the monitoring data that are highly similar to the signal characteristics of actual early-stage bridge structural damage, leading to an extremely high false alarm rate in existing monitoring systems. Because it is impossible to distinguish between sensor installation failure and bridge structural damage at the data level, maintenance work is forced to rely on inefficient and expensive on-site manual inspections, severely hindering the development of intelligent and accurate monitoring systems. The following embodiments aim to collaboratively address this systemic problem from three levels: system architecture, node hardware, and diagnostic algorithms.

[0029] Example 1: Urban elevated bridge monitoring system based on symmetrical dual resonator nodes. See the overall system architecture section. Figure 1 The system includes multiple monitoring nodes 100 deployed at key locations on the bridge, one or more aggregation units (gateways) 200, and a remote management platform 300. The monitoring nodes 100 are connected to the nearest aggregation unit 200 via wired (e.g., RS-485, Ethernet) or wireless (e.g., LoRa, Zigbee) methods. Each aggregation unit 200 communicates with the remote management platform 300 deployed on a cloud server or local data center via a 4G / 5G or fiber optic network. The remote management platform 300 has data storage, analysis, visualization, and alarm management functions.

[0030] The perception layer (monitoring nodes) consists of multiple intelligent monitoring nodes distributed across key parts of the bridge (such as the main beam, supports, and cable anchorage zones). The core feature of each node is its integrated symmetrically arranged dual-reference resonators, capable of active excitation, synchronous data acquisition, and local computation, outputting decoupled "Installation Health Index (Is)" and "Structural Characteristics." These multiple nodes collectively form the data perception nerve endings of the system.

[0031] The network layer (data aggregation unit) is typically handled by gateway devices deployed at the bridge site. It is responsible for receiving data from all monitoring nodes within its communication range, performing local aggregation, caching, and protocol conversion (such as converting LoRa data to TCP / IP data). It can handle simple edge computing tasks (such as data filtering and compression) and stably transmit data to the remote management platform via WAN links such as 4G, 5G, or fiber optics. This layer acts as a "bridge" connecting the field and the backend.

[0032] The platform layer (remote management platform) is the "brain" of the system, typically deployed in the cloud or on-premises data center. Its core is a software system containing multiple functional modules: the data access and storage module, responsible for receiving, parsing, and storing massive amounts of time-series data from various aggregation units; and the health diagnosis engine, which is the core of the algorithm, outlining the main steps of the method (such as...). Figure 4 (As shown). Based on the received Is and structural features, it runs fusion diagnostic logic: when Is is abnormal, it is determined to be a node failure; when the structural features are abnormal and Is is normal, it is determined to be bridge structural damage. Visualization and Alarm Center: Displays the health status of all nodes on the entire bridge using a graphical interface (such as GIS maps and trend curves). Based on the results of the diagnostic engine, it generates two types of structured outputs: ① "Installation and Maintenance Work Orders" for sensor nodes; ② "Safety Warnings" for the bridge structure. This achieves a closed loop of monitoring and maintenance.

[0033] To clearly demonstrate the internal structure of monitoring node 100, please refer to the detailed structure and working principle of the monitoring node. Figure 2 and Figure 7 The diagram shows a cross-sectional view of a single monitoring node. This monitoring node is the core sensing unit of the system and mainly includes a mounting base 101, a micro-exciter 102, a first reference resonator 103a, a second reference resonator 103b, and a bridge response sensor 104. The mounting base 101 is made of weathering steel or anodized aluminum alloy and is firmly connected to bridge components (such as the lower flange of the main beam) via high-strength bolts 108 and washers 109. Grooves are provided at the bolt positions on the bottom of the base 101, and conductive washers 111 are fitted into these grooves, directly supporting the contact surface of the bridge components. The micro-exciter 102 is a compact inertial electromagnetic vibrator (voice coil motor), fixed at the center of the base 101, with a peak excitation force of 0.5N and an operating frequency band of 5-800Hz. A waterproof outer shell 110 is sealed and fixed to the base 101, providing protection.

[0034] The first reference resonator 103a and the second reference resonator 103b are two commercially available tuning fork quartz resonators (e.g., 32.768kHz) with matched parameters. They are symmetrically mounted on the base 101 in pre-set left and right slots via low-stiffness silicone damping pads 105, their positions being strictly mirror-symmetrical with respect to the center of the micro-exciter 102. This symmetrical decoupled mounting is the physical basis for differential diagnostics. The bridge response sensor 104 is an IEPE type accelerometer, mounted on the base 101 near the micro-exciter 102. A signal line 107 (e.g., RS-485, Ethernet, or wireless such as LoRa, Zigbee) is fixed to one side of the base 101 via a terminal block 106, and the signal line 107 connects to the nearest data aggregation unit 200. One set of data aggregation units 200 can simultaneously connect to the signal lines of multiple monitoring nodes. Each aggregation unit 200 communicates with a remote management platform 300 deployed on a cloud server or local data center via a 4G / 5G or fiber optic network. The remote management platform 300 has data storage, analysis, visualization, and alarm management functions.

[0035] The data aggregation unit 200 includes a sealed electronic compartment, where signal conditioning and acquisition circuitry, as well as processing and communication modules, are integrated. The circuit block diagram is shown below. Figure 3 As shown, it mainly includes: the core processor (U1) adopts a low-power ARM Cortex-M4 MCU (such as the STM32L4 series), which is responsible for overall control and algorithm execution. The excitation drive circuit is amplified by the PWM output of the MCU through the Class D audio power amplifier chip (U2) to drive the micro exciter 102.

[0036] The signal acquisition module employs three synchronous acquisition channels. Channel 1 connects to the first reference resonator, and the signal, after passing through a charge amplifier (U3a) and a second-order anti-aliasing filter with a cutoff frequency of 500Hz, is sent to the MCU's synchronous ADC. Channel 2 connects to the second reference resonator, and the signal, after passing through a charge amplifier and a second-order anti-aliasing filter with a cutoff frequency of 500Hz, is sent to the synchronous ADC. Channel 3 connects to the bridge response sensor, and the signal, after passing through an instrumentation amplifier (U4) and an anti-aliasing filter with a cutoff frequency of 200Hz, is sent to the synchronous ADC. The communication unit (U5) uses a low-power NB-IoT module, responsible for uploading the processed results to the aggregation unit.

[0037] The node's workflow is controlled by the MCU firmware, such as... Figure 4 and Figure 5 As shown in the flowchart: Step S401, Initialization Self-Learning: After the node is installed and secured, upon first power-on during a preset no-traffic period (e.g., 2:00 AM), the MCU controls the micro-exciter 102 to apply a 5-second pseudo-random binary sequence (PRBS) excitation. Three initial response signals are simultaneously acquired, and the initial installation health index is calculated and stored. (Based on the average coherence function of the 103a and 103b responses in the 50-400Hz frequency band) and the static load environmental energy baseline .

[0038] Step S402, Periodic Active Calibration: The node automatically performs calibration daily during low-interference periods. The MCU controls the application of the same PRBS excitation and synchronously acquires the real-time response signal. .

[0039] Step S403, Environmental Interference Judgment and Installation Health Index Calculation: The MCU calculates the current bridge response signal energy. .like If the current calibration is determined to be affected by severe traffic interference, the data is marked and the process awaits the next cycle. Otherwise, based on... and Calculate the current installation health index .

[0040] Step S404, Installation Status Determination: and Comparison, setting threshold tolerance .like The node installation status is determined to be "healthy"; if it remains "healthy" for three consecutive cycles... If so, it is determined that "installation is degraded", and a generator containing the node ID and... Local alarms are triggered.

[0041] Step S405, Structural feature extraction and confidence fusion: If the installation status is "healthy", then based on Extracting high-confidence structural features (such as first-order frequencies) Damping ratio (etc.). If the installation status is "degraded", then any structural features extracted at the same time will be marked with "confidence downweight".

[0042] Step S406, Data Reporting: The MCU will report the data from this calibration. The structural characteristic values, node status (health / deterioration) and timestamp are packaged and sent to the aggregation unit 200 through the NB-IoT module.

[0043] The aggregation unit 200 is responsible for collecting data from all monitoring nodes 101 within its wireless coverage area, performing local caching and protocol conversion, and then forwarding it to the remote management platform 300.

[0044] The remote management platform 300's software system performs the following core functions: data storage and visualization: displaying the real-time data of each node in the form of maps and charts. Historical curves of trends and structural characteristics. Health diagnosis and alarms: When a node installation degradation alarm is received, the platform automatically generates a sensor maintenance work order to be dispatched to the designated location. When the structural characteristics of a node (such as...) are detected... When a significant drift exceeding 5% occurs, and the node is in a healthy state, the platform triggers a high-confidence structural anomaly warning and locates the specific bridge component. When structural characteristics are abnormal but the node state is deteriorated, the platform marks the warning as pending review; please prioritize checking the node installation. Statistical analysis: The platform can perform analysis on all nodes... Statistical analysis was conducted to identify areas with generally poor installation conditions across the entire bridge, providing decision support for batch maintenance.

[0045] A verification experiment on the correlation between installation health index and torque loss (based on ANSYS Workbench) was conducted to quantitatively verify the relationship between torque loss of installation bolts at monitoring nodes and the installation health index. To determine the correlation and ensure the rationality of the decision threshold in step S404, a static simulation experiment was conducted using ANSYS Workbench software. The specific process is as follows.

[0046] 1. Model Construction: In the ANSYS Workbench statics module, a simplified 3D model of the monitoring node mounting base, bolts, and the lower flange of the bridge main beam was established. The bolt material was set to carbon steel, the mounting base material to weathering steel, and the bridge main beam material to aluminum alloy. Physical parameters such as elastic modulus and Poisson's ratio were assigned to the corresponding materials. The installation method of "high-strength bolt clamping" in the embodiment was strictly matched, and the standard bolt preload torque of 30 N·m was converted into an axial preload force applied to the bolt head. Simultaneously, the bottom surface of the lower flange of the bridge main beam was fixed as a boundary constraint.

[0047] 2. Variable Simulation: By progressively reducing the bolt preload percentage, equivalent simulations of torque loss gradient conditions (0%, 10%, 20%...100%) are performed. For each condition, a static solver is used to focus on solving for the connection stiffness values ​​between the mounting base and bridge components under different torque loss states (directly related to the coherence characteristics of the resonator response, which is crucial for calculation). (Core physical foundation).

[0048] 3. Health Index Conversion: Based on the example... Definition formula ( =Current stiffness / Initial stiffness), converting the connection stiffness value under each torque loss condition into the corresponding installation health index. To obtain the "percentage of torque loss" The associated dataset; where the initial stiffness corresponds to the solution result of the 0% torque loss condition, and corresponds to the initial installation health index. =0.98.

[0049] 4. Data Visualization and Threshold Validation: Export the correlation dataset obtained from the simulation as a CSV file, and plot the "Installation Health Index" using Matplotlib software. The curve showing the change in percentage of torque loss (i.e.) Figure 6 The corresponding experimental curve), and add the decision threshold set in the embodiment ( -Δ=0.88). The curve clearly shows... The nonlinear decreasing trend of torque loss verifies the scientific validity of the threshold setting in step S404—when the torque loss reaches 30%, A value reduced to 0.91 (≥0.88) is still considered healthy; when the torque loss reaches 40%, When the value drops to 0.85 (<0.88), a deterioration warning is triggered, which matches the risk level of bolt loosening in actual engineering projects.

[0050] Example 2: An integrated system for monitoring cable force in long-span cable-stayed bridges. This example demonstrates the application of the invention to complex bridge structures. Cable force variation in cable-stayed bridges is a key monitoring indicator, but sensors mounted on cable clamps are susceptible to changes in installation status due to wind vibration and temperature variations. In this example, the monitoring nodes adopt the same core design as in Example 1, but their shape is adapted to the curved surface of the cable clamps. The nodes are densely deployed in the anchorage zones of each cable and at the cable clamp locations.

[0051] Systematic diagnostics: The remote management platform 300 can simultaneously monitor all cable monitoring nodes. If only a single root node A decrease in load indicates a problem with the sensor installation at that location; if multiple cable nodes simultaneously exhibit abnormal structural characteristics while maintaining a healthy installation status, it may indicate tower displacement or changes in overall load distribution, providing a system-level insight that single-point monitoring cannot achieve.

[0052] Closed-loop maintenance: The "Installation Deterioration" work orders generated by the platform can directly guide maintenance personnel to tighten specific cable clamps, avoiding blind inspection of the entire line.

[0053] Through the systematic implementation described above, the systemic false alarm rate is reduced: Because each node possesses self-diagnostic capabilities, the system can filter installation fault signals at the source. Simulation data shows that in a system with 100 nodes, this solution can reduce the number of systemic false alarms caused by installation problems by more than 70%. Maintenance costs are significantly reduced and efficiency is improved: Installation degradation alarms directly generate accurate maintenance work orders, changing the inefficient traditional alarm → full-line inspection → fault location model. It is expected to reduce sensor-related maintenance and inspection costs by more than 30%. Data credibility and decision support capabilities are enhanced: The management platform provides a view that separates the installation health index from structural characteristics, allowing managers to clearly distinguish between the performance degradation of the bridge itself and problems with the monitoring system, greatly enhancing the credibility and practical value of monitoring data in structural safety assessments.

[0054] It should be noted that the above embodiments and accompanying drawings are merely illustrative examples of the core principles and key structures of the system, nodes, and methods of the present invention. The accompanying drawings are simplified schematic diagrams, intended to clearly illustrate the structural, process, or data flow relationships related to the innovative points of the technical solution, and are not intended to limit the complete form of the actual product. This specification focuses on the innovative technical means necessary to achieve the invention's objectives and solve the technical problems. While auxiliary or common-sense details such as specific configurations of network communication protocols, database selection, user interface design, and three-proof casing treatments, which can be implemented without creative effort by those skilled in the art, are not elaborated upon, they should be understood as naturally included in the specific implementation of the present invention and fall within the protection and implementation scope of this technical solution.

Claims

1. A bridge safety monitoring system based on differential symmetric self-diagnosis, characterized by, include: One or more monitoring nodes, aggregation units, and a remote management platform; The monitoring nodes are used to be installed at various test locations on the bridge. Each monitoring node includes: a mounting base for fixed connection with the bridge component; a micro exciter disposed on the mounting base; a first reference resonator and a second reference resonator, which have the same structural parameters and are symmetrically arranged on the mounting base and decoupled from the bridge component. A bridge response sensor is mounted on the mounting base; A processing module is configured to: control the micro-exciter to apply excitation; determine an installation health index characterizing the symmetry of the installation state based on the response signals of the first and second reference resonators; and process the output signal of the bridge response sensor based on the installation health index to separate structural features reflecting the bridge structural state. The aggregation unit is communicatively connected to one or more of the monitoring nodes and is used to aggregate data from each node; The remote management platform is communicatively connected to the aggregation unit and is used to receive data, present the installation health index and the structural features, and generate maintenance decisions according to preset rules. The installation health index of each node is calculated by: converting the response signals of the first and second reference resonators of each node to the frequency domain; calculating the coherence function of the two resonators within a preset analysis frequency band; and taking the average value of the coherence function within the preset analysis frequency band as the installation health index of the node.

2. The bridge safety monitoring system of claim 1, wherein The installation health index is determined based on the coherence function values ​​of the response signals of the first reference resonator and the second reference resonator within a predetermined frequency band.

3. The bridge safety monitoring system of claim 1, wherein, The aggregation unit is also configured to perform spatiotemporal correlation analysis based on the installation health index and structural characteristics of multiple monitoring nodes to locate abnormal areas of the bridge or identify damage patterns.

4. A node for bridge safety monitoring, characterized in that include: Mounting base, used for fixed connection with bridge components; A micro-actuator is mounted on the mounting base; The first reference resonator and the second reference resonator have the same structural parameters and are symmetrically arranged on the mounting base and decoupled from the bridge component; The first sensing unit and the second sensing unit are respectively used to acquire the response signals of the first reference resonator and the second reference resonator; A bridge response sensor is used to acquire the response signals of the bridge components; The processing module is connected to the micro-exciter, the first sensing unit, the second sensing unit, and the bridge response sensor, respectively. The processing module is configured to: control the micro-exciter to apply excitation; determine an installation health index characterizing the symmetry of the installation state of the mounting base based on the response signals acquired by the first sensing unit and the second sensing unit; and process the response signals acquired by the bridge response sensor based on the installation health index to extract structural features related to the damage of the bridge components. The bridge response sensor is an IEPE type accelerometer, which is mounted on the base near the micro exciter. The installation health index of each node is calculated by: converting the response signals of the first and second reference resonators of each node to the frequency domain; calculating the coherence function of the two resonators within a preset analysis frequency band; and taking the average value of the coherence function within the preset analysis frequency band as the installation health index of the node.

5. The node of claim 4, wherein, The first reference resonator and the second reference resonator are mounted on the mounting base via vibration damping elements.

6. The node of claim 4 or 5, characterized in that, The processing module is also configured to: synchronously acquire the signal energy of the bridge response sensor during the calibration period; when the signal energy exceeds the energy threshold determined based on historical data, determine that the installation health index during the period is affected by environmental interference, and mark it.

7. A bridge safety monitoring method based on differential symmetric self-diagnosis, characterized by, The method, applied to the system as described in any one of claims 1-3, comprises: Excitation is applied through micro-exciters deployed at one or more monitoring nodes at various parts of the bridge; At each monitoring node, the response signals of its first reference resonator, second reference resonator, and bridge response sensor are collected synchronously. Based on the response signals of the first and second reference resonators, the installation health index of each node is calculated. Based on the installation health index of each node, the bridge response signal of the corresponding node is decoupled and the structural features of each part are extracted. The installation health index and structural characteristics of each node are collected and uploaded to the remote management platform; On the remote management platform, based on the installation health index and structural characteristics, a fusion analysis is performed to generate maintenance recommendations for the sensor installation status and assessment results for the safety of the bridge structure.

8. The method of claim 7, wherein, The calculation of the installation health index of each node includes: Convert the response signals of the first and second reference resonators of each node to the frequency domain; Calculate the coherence function of both within the preset analysis frequency band; The average value of the coherence function within the preset analysis frequency band is used as the installation health index of the node.

9. The method of claim 7, wherein, After the initial installation of the node, an initialization step is performed, which includes: applying excitation and acquiring initial response signals under conditions of no significant environmental interference; calculating and storing an initial installation health index based on the initial response signals of the first and second reference resonators; wherein, the installation health index obtained subsequently is compared with the initial installation health index to determine the trend of installation status changes.

10. The method of claim 7, wherein, The process of decoupling the bridge response signal of the corresponding node based on the installation health index of each node includes: when the installation health index of a node is higher than a first preset threshold, the node is determined to be in a reliable installation state, and the structural features extracted from its bridge response signal are directly used in the bridge safety assessment; when the installation health index of a node is lower than a second preset threshold, the node is determined to be in an abnormal installation state, a maintenance alarm is triggered for the node, and the extracted structural features are downweighted or required to be reviewed.

Images