A bridge support detection system and multi-load joint detection method thereof

Abstract

This application discloses a bridge bearing detection system and a multi-load joint detection method thereof, relating to the technical field of bridge structure health monitoring. The detection system includes a camera, an eddy current thickness sensor, an acceleration sensor, a laser Doppler vibrometer, a strain gauge, a data acquisition module, a data processing module, and a data transmission module. At least one camera is fixed to the side or directly above the bridge bearing via an adjustable bracket, with its optical axis facing the bridge bearing surface to ensure comprehensiveness and clarity of image acquisition. The system uses image processing algorithms to detect cracks on the bearing surface, combines dynamic response analysis with model predictions to predict stiffness degradation and voiding, and assesses the degree of corrosion through thickness monitoring, thereby achieving full life cycle management of the bearing.

Description

Technical Field

[0001] The present application relates to the technical field of bridge structure health monitoring, specifically a bridge bearing detection system and a multi-load joint detection method that integrate multiple sensors, image processing algorithms, dynamic response analysis and optimization models, which can monitor the health status of bridge bearings in real time under multi-dimensional disease characteristics such as cracking, stiffness degradation, voiding and thickness changes. Background Art

[0002] As an essential component of bridge structures, bridge bearings perform the critical functions of transferring loads, adjusting deformation, and reducing shock. Their health directly impacts the overall safety and service life of the bridge. Over their long service lives, bridge bearings are subject to a variety of complex factors, including repeated impacts from traffic loads, environmental erosion (such as rain, salt spray, and chemical corrosion), and material fatigue and aging. These factors can cause bearing defects such as cracking, voids, corrosion, and stiffness degradation. For example, the long-term effects of traffic loads can cause fatigue cracks in bearings, environmental erosion can accelerate corrosion of metal components and deterioration of concrete, and material aging can lead to a gradual degradation of bearing stiffness. These problems not only reduce the bearing's load-bearing capacity but can also cause partial or even complete failure of the bridge structure, seriously threatening the safe operation of the bridge.

[0003] However, traditional bridge bearing inspection methods have obvious limitations. Currently, the inspection of bridge bearings mainly relies on manual inspection or single parameter monitoring. Manual inspection is usually carried out by visual observation, tapping and listening, or simple measuring tools. Although this method is intuitive, it is limited by the subjective experience and skill level of the inspectors, making it difficult to detect hidden defects. In addition, the inspection efficiency is low and the coverage is limited. Single parameter monitoring usually uses displacement sensors, stress sensors, or acceleration sensors, which can only monitor a certain parameter of the bearing (such as displacement, stress, or vibration) and cannot fully reflect the overall condition of the bearing. For example, monitoring only displacement changes may ignore stress concentration or corrosion problems, while monitoring only stress may not capture bearing stiffness degradation or voiding. In addition, traditional methods usually lack real-time and continuity, and cannot promptly detect potential problems with bearings during bridge operation. This causes defects to gradually worsen without being detected, increasing the risk of bridge operation.

[0004] The main reasons for these shortcomings include the limitations of technical means and the complexity of detection systems. Traditional manual inspection methods rely on manual operation and are difficult to automate. Due to technical limitations, single-parameter monitoring systems cannot simultaneously integrate multiple sensors, resulting in a single detection parameter. Furthermore, existing data processing algorithms are often relatively simple and cannot effectively process multi-source heterogeneous data, resulting in insufficient accuracy and reliability of detection results. For example, some systems may only use simple threshold judgment methods, unable to conduct in-depth analysis and prediction of much more complex parameter data. At the same time, existing detection systems also present certain difficulties in installation and maintenance. The installation location and number of sensors are often limited by the bridge structure, making it difficult to cover all key areas. Furthermore, the long-term stability and anti-interference capabilities of sensors are often insufficient, resulting in reduced reliability of detection data.

[0005] In order to solve the above problems, it is of great engineering significance to develop a system that can monitor the health status of bridge bearings in real time during bridge operation and realize multi-dimensional parameter comprehensive analysis and intelligent early warning. Summary of the Invention

[0006] This application aims to overcome at least one shortcoming of the existing technology by providing a bridge bearing detection system and a multi-load joint detection method. Through multi-sensor data acquisition, real-time analysis, health assessment, and early warning functions, the system enables high-precision monitoring and diagnosis of critical bridge bearing defects. The system uses image processing algorithms to detect surface cracks in bearings, combines dynamic response analysis with model predictions for stiffness degradation and voiding, and assesses corrosion levels through thickness monitoring, thereby enabling full lifecycle management of bearings.

[0007] To achieve the above objectives, in a first aspect, the present application discloses a bridge bearing detection system, which includes a camera, an eddy current thickness sensor, an acceleration sensor, a laser Doppler vibrometer, a strain gauge, a data acquisition module, a data processing module, and a data transmission module. At least one camera is fixed to the side or directly above the bridge bearing via an adjustable bracket, with its optical axis facing the bridge bearing surface to ensure comprehensiveness and clarity of image acquisition; the eddy current thickness sensor is installed on the metal surface of the bridge bearing at a position closest to the slip zone, with the eddy current thickness sensor facing the bridge bearing surface, and its detection head remains parallel to the metal surface being measured. The acceleration sensor is installed in the center of the top of the bridge bearing. The acceleration sensor is connected to the data acquisition module via a cable and outputs a real-time dynamic response signal. The laser Doppler vibrometer is installed on the top of the bridge pier, facing the connection between the bridge support and the beam body. The measuring beam is parallel to the sliding direction of the support. The laser Doppler vibrometer emits a beam perpendicular to the support surface through a laser transmitter. The beam reflection signal is captured by the photoelectric receiver in the laser Doppler vibrometer, and the laser Doppler vibrometer calculates the displacement response of the bridge support under the action of dynamic load. The laser Doppler vibrometer is connected to the data acquisition module via an optical fiber cable. The strain gauges are attached to the bridge support by glue. The upper and lower contact surfaces and edge areas are covered with heat-shrinkable insulating material and are used to measure the compressive stress in the vertical direction and the shear stress in the horizontal direction. Each strain gauge is connected to the data acquisition module through a cable; the data acquisition module is connected to the data processing module through an interface, and the signal of the data acquisition module interacts with the data processing module through a two-way communication protocol; the data processing module is installed in the bridge management facility, and processes the real-time data sent by each sensor for filtering, compensation and feature extraction; the data processing module is connected to the data sending module, and transmits the detection data to the outside through the data sending module to realize the output of the detection data.

[0008] Furthermore, the camera is directly connected to the data acquisition module via a Gigabit Ethernet interface to transmit high-resolution image data in real time.

[0009] Furthermore, the eddy current thickness sensor is fixed by a clamp with a fixing force of 20 N·m, ensuring that the distance between the sensor and the support surface is 1 mm, and the measurement accuracy is ±0.01 mm.

[0010] Furthermore, the bridge support detection system also includes a temperature and humidity sensor installed next to the eddy current thickness sensor. The temperature and humidity sensor is connected to the data acquisition module to provide temperature and humidity information to the data acquisition module.

[0011] Furthermore, the eddy current thickness sensor is connected to the data acquisition module via a shielded twisted pair cable, and the shielding layer of the shielded twisted pair cable is grounded to eliminate electromagnetic interference.

[0012] Furthermore, the acceleration sensor is installed using M8 stainless steel bolts to fix the sensor base on the beam with a torque of 40 Nm, with its sensing axis aligned with the support sliding direction and the vertical direction.

[0013] Furthermore, the strain gauge is attached at a position 50 mm from the center and 20 mm from the bottom of the support edge to ensure that the measurement covers the core area of ​​the support where the force is applied.

[0014] Furthermore, the data acquisition module is installed in a hidden area on top of the pier.

[0015] Furthermore, the data acquisition module is connected to the data processing module via an RS485 interface.

[0016] The components of this application are closely connected and clearly divided into functional groups within the system. All sensors directly collect data from key bearing locations, transmitting it to the data acquisition module via physical connections and a communication network. The acquisition module then connects with the data processing module for processing, ultimately outputting the results. The overall system design enables high-precision monitoring of bridge bearings, reliable data transmission, and timely status assessment, providing technical support for safe bridge operation.

[0017] In a second aspect, the present application discloses a multi-load joint detection method based on a bridge bearing detection system, the detection method comprising the following steps:

[0018] Step 1: Collect images of the bridge bearing surface and perform crack detection. Use a high-resolution camera to collect images of the bearing surface, ensuring that the image resolution is accurate to within 0.1 mm of the actual size of the bearing, and capture the fine features of the crack; input the collected images into an improved deep learning model, which uses a dynamic weighted loss function to assign higher priority to cracks with larger areas, while adding edge sharpness constraints to improve the recognition accuracy of complex crack shapes; the model outputs crack length, width and area features based on the pixel area of ​​the crack. These pixel features are converted to actual physical dimensions through the camera's calibration ratio, and the crack growth trend is determined by calculating the ratio of the current crack length to the historical crack length.

[0019] Step 2: Evaluate the stiffness changes of the bridge bearings. Under normal bridge operating conditions, use accelerometers and laser Doppler vibrometers to synchronously collect the dynamic response signals of the bridge bearings and beams. The change in bridge bearing stiffness is calculated using the dynamic displacement ratio and the vibration main frequency reduction ratio. The dynamic displacement ratio represents the ratio of the dynamic displacement of the bearing to the dynamic displacement of the beam, and the vibration main frequency reduction ratio quantifies the stiffness degradation by comparing the change in the current vibration main frequency with the initial main frequency. If the dynamic displacement ratio exceeds 2 times or the main frequency reduction ratio exceeds 20%, a stiffness degradation warning signal is generated.

[0020] Step 3: Detect bridge bearing debonding. This phenomenon is analyzed using the flexibility matrix method. The flexibility matrix describes the deformation characteristics of bridge bearings after being subjected to forces in different directions. Diagonal elements reflect the direct deformation of the bridge bearing, while off-diagonal elements reflect the coupled deformation between adjacent components. The normalized diagonal difference ratio of the flexibility matrix—the ratio of the difference between the diagonal and off-diagonal elements to the average value of the initial diagonal elements—is calculated to quantify the degree of debonding. Combined with dynamic response data from vehicle impact loads and temperature fluctuation parameters, a machine learning model is used to predict the expansion trend of the debonding and output a percentage value for the degree of debonding.

[0021] Step 4: Monitor the thickness changes of bridge supports and evaluate the corrosion situation. The eddy current thickness sensor monitors the thickness changes of bridge supports in real time by detecting the relationship between the sensing signal and the distance from the metal surface, and uses a nonlinear compensation model to adjust the thickness. The actual thickness value is calculated based on the quadratic relationship between temperature change and thickness error. Humidity compensation is based on the interference of humidity changes on the sensor signal, and corrections are made using experimental fitting data. When the corrosion rate exceeds 0.1 mm per year or the thickness is reduced to less than 10% of the design value, a corrosion warning is generated.

[0022] Step 5: Fusion of multi-parameter data and comprehensive evaluation. The crack characteristics, stiffness changes, voiding degree, and thickness changes obtained in steps 1 to 4 are integrated to generate a comprehensive health score. The scoring model dynamically assigns weights to each parameter based on its impact on bearing performance.

[0023] Step 6: Generate early warning signals and provide maintenance recommendations. Generate graded early warning signals based on the health scores and test results obtained in step 5.

[0024] The bridge bearing detection method of this application has significant technical advantages, which are mainly reflected in the improvement of detection accuracy, comprehensive evaluation capabilities and intelligent management level. By introducing a dynamic correction model, nonlinear compensation is performed on thickness measurement and stiffness evaluation, effectively eliminating the influence of environmental factors (such as temperature and humidity) on the detection results, and ensuring the accuracy of the data. By adopting an improved target detection algorithm and flexibility matrix analysis method, the recognition accuracy of cracks and voids is greatly improved, especially in the early detection of complex-shaped cracks and slight voids. The comprehensive health scoring model integrates multi-dimensional parameters of crack propagation, stiffness degradation, void degree and thickness thinning, and dynamically adjusts the weights to be closer to the actual service status of the bearing, providing a comprehensive and quantitative health assessment. At the same time, combined with real-time data and historical trends, clear and unambiguous risk warnings and maintenance recommendations are generated, which greatly improves the scientificity and efficiency of bridge operation and maintenance management, and provides reliable technical support for health monitoring of bridges throughout their life cycle.

[0025] The above-listed beneficial effects are not exhaustive and other potential beneficial effects and detailed technical implementations will be further disclosed in the examples or other description sections of this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] After reading the following detailed description in conjunction with the accompanying drawings, you will better understand the various aspects of the present disclosure. The positions, sizes, and ranges of various structures shown in the drawings and the like sometimes do not represent the actual positions, sizes, and ranges. In the drawings:

[0027] Figure 1 It is a system block diagram of an embodiment disclosed in this application. DETAILED DESCRIPTION

[0028] The present disclosure will be described below with reference to the accompanying drawings, which illustrate several embodiments of the present disclosure. However, it should be understood that the present disclosure can be presented in many different ways and is not limited to the embodiments described below; in fact, the embodiments described below are intended to make the present disclosure more complete and fully illustrate the scope of protection of the present disclosure to those skilled in the art. It should also be understood that the embodiments disclosed herein can be combined in various ways to provide many additional embodiments.

[0029] It should be understood that like reference numerals refer to like elements throughout the drawings. In the drawings, the dimensions of some features may be distorted for clarity.

[0030] It should be understood that the terms used in this specification are intended only to describe specific embodiments and are not intended to limit the present disclosure. All terms (including technical and scientific terms) used in this specification have the meanings commonly understood by those skilled in the art, unless otherwise defined. For the sake of brevity and / or clarity, technologies, methods, and devices known to those skilled in the relevant art may not be discussed in detail; however, where appropriate, such technologies, methods, and devices should be considered part of this specification.

[0031] As used in this specification, the singular forms "a," "an," "said," and "the" include the plural forms unless otherwise expressly stated. The terms "include," "comprise," and "contain" as used in this specification indicate the presence of the claimed features, but do not exclude the presence of one or more additional features. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0032] In this embodiment, the bridge bearing detection system is a high-precision device used to monitor the health of bridge bearings in real time. Its overall structure consists of multiple high-precision sensors, a data acquisition module, a data processing module, and a data transmission module. These components work closely together through physical connections and a communication network to form a complete monitoring system.

[0033] The camera is fixed to the side or directly above the bridge support via an adjustable bracket, with its optical axis facing the bridge support surface, ensuring comprehensive and clear image acquisition. The camera's optical system uses multi-layer coating technology, capable of capturing clear images of the support surface under complex lighting conditions, with a resolution accurate to within 0.1 mm. The camera is directly connected to the data acquisition module via a Gigabit Ethernet interface, transmitting high-resolution image data in real time. The adjustable bracket is made of high-strength aluminum alloy material with corrosion and vibration resistance, ensuring that the camera remains stable during long-term use. The bracket's adjustment mechanism uses a precision thread design, which can achieve fine-tuning of the camera's position and angle, ensuring that the optical axis is perpendicular to the support surface to avoid angular deviation.

[0034] The eddy current thickness sensor is installed on the metal surface of the bridge support at the position closest to the slip area. It is fixed with a special clamp with a fixing force of 20 Nm to ensure that the distance between the sensor and the support surface is 1 mm and the measurement accuracy is ±0.01 mm. The eddy current thickness sensor faces the surface of the bridge support, and its detection head remains parallel to the metal surface being measured. The sensor is connected to the data acquisition module via a shielded twisted pair cable, and the shielding layer is grounded to eliminate electromagnetic interference. The sensitive element of the sensor is made of highly magnetic permeable material and can accurately measure small changes in the thickness of the support. The signal processing unit has a built-in temperature compensation algorithm, which corrects the measurement results based on the thermal expansion effect of the metal and humidity changes to ensure that the measurement accuracy reaches ±0.01 mm. Thickness measurement signal and support thickness The relationship is:

[0035]

[0036] in, and is the sensor calibration factor.

[0037] Temperature compensation based on temperature changes Thickness measurement error Make corrections:

[0038]

[0039] in, and is the experimental fitting coefficient.

[0040] Humidity compensation based on humidity changes Thickness measurement error Make corrections:

[0041]

[0042] in, is the experimental fitting coefficient.

[0043] Actual thickness after correction for:

[0044]

[0045] like ( is the designed thickness), the system triggers a corrosion warning.

[0046] The acceleration sensor is installed in the center of the top of the bridge support and fixed in the prefabricated mounting hole at the bottom of the beam. During installation, M8 stainless steel bolts are used to fix the sensor base to the beam with a torque of 40 Nm, with its sensing axis aligned with the sliding direction and vertical direction of the support. The acceleration sensor is connected to the data acquisition module via a cable and outputs a real-time dynamic response signal. The sensitive unit of the sensor is made of piezoelectric ceramic material, which can capture the vibration signal of the support within a wide frequency range. The sealed housing has an IP67 protection grade and can withstand the influence of harsh environments. Stiffness evaluation is performed by dynamic displacement ratio and vibration main frequency reduction ratio Comprehensive calculation:

[0047]

[0048] in, and are the dynamic displacements of the supports and beam, and They are the initial main frequency and the current main frequency respectively.

[0049] Comprehensive stiffness index for:

[0050]

[0051] in, and is the weight coefficient.

[0052] like or , the system generates a stiffness degradation warning.

[0053] A laser Doppler vibrometer is installed atop the pier, directly opposite the connection between the bridge support and the beam, with the measuring beam parallel to the direction of support slip. The laser Doppler vibrometer uses a laser transmitter to emit a beam perpendicular to the support surface. The reflected beam signal is captured by a photoelectric receiver, which calculates the displacement response of the support under dynamic loads. The laser Doppler vibrometer is connected to a data acquisition module via a fiber optic cable. Its control unit is fixed to the sidewall of the pier, and all components are protected by a steel casing to prevent mechanical damage and environmental corrosion. The laser Doppler vibrometer's optical system uses a high-precision lens system, enabling micron-level displacement measurement over long distances.

[0054] Strain gauges are glued to the upper and lower contact surfaces and edge areas of the bridge support and covered with heat-shrink insulation. They measure vertical compressive stress and horizontal shear stress. Attachment is made 50 mm from the center of the support edge and 20 mm from the bottom, ensuring coverage of the core stress-bearing area of ​​the support. The strain gauges are connected to the data acquisition module via cables, with all connection points welded to ensure signal integrity. The strain gauge's sensitive grid is made of a highly elastic alloy material, ensuring stable output characteristics under high stress conditions.

[0055] The data acquisition module is installed in a concealed area atop the bridge pier, no more than 5 meters from the sensor to ensure signal strength and stability. It connects to the data processing module via an RS485 interface, supporting 1000 sensor data reads and analysis per second. The module incorporates a multi-channel signal conditioning circuit to amplify, filter, and digitize sensor signals. The data acquisition module's enclosure features a waterproof and dustproof design with an IP65 rating, ensuring stable operation in harsh environments.

[0056] The data processing module, installed within the bridge management facility, provides processing capabilities for filtering, compensating, and extracting features from real-time data. It interacts with the data acquisition module via a bidirectional communication protocol, ensuring real-time and reliable data transmission. The data processing module's hardware platform utilizes a high-performance embedded processor that supports multi-threaded parallel computing, enabling rapid processing of large amounts of sensor data.

[0057] The data transmission module is connected to the data processing module, transmitting test data externally through the data transmission module, achieving test data output. The data transmission module transmits test data to the cloud-based analysis platform via the 5G high-speed network. The platform stores the test data in a distributed storage system, analyzes trends using pre-set algorithms, and generates health assessment reports. The data transmission module's communication interface supports multiple protocols, including TCP / IP, MQTT, and HTTP, ensuring compatibility with external systems.

[0058] During implementation, the system first uses a camera to capture images of the bridge bearing surface. It then uses a deep learning model to identify crack length, width, and area characteristics and calculate the crack growth rate. If the crack growth rate exceeds 2 mm per month or the crack area exceeds 10% of the bearing's total area, the system generates a crack growth warning signal. Next, the system uses accelerometers and laser Doppler vibrometers to synchronously collect dynamic response signals from the bearing and beam, comprehensively evaluating bearing stiffness changes using the dynamic displacement ratio and the vibration main frequency reduction ratio. If the dynamic displacement ratio exceeds 2 times or the main frequency reduction ratio exceeds 20%, the system generates a stiffness degradation warning signal.

[0059] The system also analyzes bearing voids using the flexibility matrix method, calculating the normalized diagonal difference ratio of the flexibility matrix to quantify the extent of voids. When the normalized diagonal difference ratio exceeds 30%, the system determines that the bearing may be voided and predicts the potential for void expansion based on vehicle impact loads and temperature fluctuation parameters. Eddy current thickness sensors monitor bearing thickness changes in real time, using a nonlinear compensation model to correct for measurement errors caused by environmental changes. The system triggers a corrosion warning when the corrosion rate exceeds 0.1 mm per year or the thickness decreases to less than 10% of the design value.

[0060] Finally, the system integrates test results, including crack characteristics, stiffness changes, voiding levels, and thickness variations, to generate a comprehensive health score. The scoring model dynamically assigns weights to various parameters based on their impact on bearing performance, with the health score ranging from 0 to 100. If the score falls below 30, the system indicates a high-risk condition and recommends immediate repair or replacement of the bearing. Early warning signals and maintenance recommendations are transmitted to the operation and maintenance management center via display screens and remote terminals, along with detailed inspection areas and priority levels. The system automatically generates inspection reports and maintenance plans for management personnel to follow.

[0061] Details not disclosed in this embodiment, such as the specific circuit design of the sensors, the software algorithm implementation of the data processing module, and the specific configuration of the communication protocol, are known to those skilled in the art or are currently available and require no further detailed description. Through the above-described implementation, the bridge bearing detection system achieves high-precision monitoring, reliable data transmission, and timely status assessment, providing technical support for the safe operation of bridges. The system's dynamic correction, improved detection, comprehensive assessment, and intelligent early warning capabilities significantly improve the detection accuracy and operational efficiency of bridge bearings.

[0062] Although exemplary embodiments of the present disclosure have been described, it will be understood by those skilled in the art that various changes and modifications may be made to the exemplary embodiments of the present disclosure without departing substantially from the spirit and scope of the present disclosure. Therefore, all such changes and modifications are intended to be within the scope of protection of the present disclosure as defined by the appended claims. The present disclosure is defined by the appended claims, and equivalents of these claims are intended to be included therein.

Claims

1. A multi-load joint detection method is applied to a bridge bearing detection system, which includes a camera, an eddy current thickness sensor, an acceleration sensor, a laser Doppler vibrometer, a strain gauge, a data acquisition module, a data processing module, and a data transmission module, wherein: At least one camera is fixed on the side or directly above the bridge support through an adjustable bracket, with its optical axis facing the surface of the bridge support to ensure the comprehensiveness and clarity of image acquisition; the eddy current thickness sensor is installed on the metal surface of the bridge support at a position closest to the slip area, the eddy current thickness sensor faces the surface of the bridge support, and its detection head remains parallel to the metal surface being measured; the acceleration sensor is installed in the center of the top of the bridge support; the acceleration sensor is connected to the data acquisition module through a cable to output a real-time dynamic response signal; the laser Doppler vibrometer is installed on the top of the pier, facing the connection between the bridge support and the beam body, and the measuring beam is parallel to the sliding direction of the support. The laser Doppler vibrometer emits a beam perpendicular to the support surface through a laser transmitter, and the beam reflection signal is captured by the photoelectric receiver in the laser Doppler vibrometer, and the laser Doppler vibrometer calculates the displacement response of the bridge support under dynamic load; the laser Doppler vibrometer The vibrometer is connected to the data acquisition module through an optical fiber cable; the strain gauges are respectively attached to the upper and lower contact surfaces and edge areas of the bridge support by gluing and covered with heat-shrinkable insulating material, which are used to measure the compressive stress in the vertical direction and the shear stress in the horizontal direction. Each strain gauge is connected to the data acquisition module through a cable; the data acquisition module is connected to the data processing module through an interface, and the signal of the data acquisition module interacts with the data processing module through a two-way communication protocol; the data processing module is installed in the bridge management facility, and processes the real-time data sent by each sensor for filtering, compensation and feature extraction; the data processing module is connected to the data sending module, and transmits the detection data to the outside through the data sending module to realize the output of the detection data; the bridge support detection system also includes a temperature and humidity sensor installed next to the eddy current thickness sensor, which is connected to the data acquisition module to provide temperature and humidity information to the data acquisition module, and is characterized in that: The detection method comprises the following steps: Step 1: Capture images of the bridge bearing surface and perform crack detection. A high-resolution camera captures images of the bearing surface, ensuring image resolution within 0.1 mm of the bearing's actual dimensions to capture the finest crack features. The captured images are fed into an improved deep learning model, which uses a dynamically weighted loss function to prioritize larger cracks while also increasing edge sharpness constraints. The model outputs crack length, width, and area features based on the crack's pixel area. These pixel features are converted to actual physical dimensions using the camera's calibration scale. The crack growth trend is determined by calculating the ratio of the current crack length to the historical crack length. Step 2: Evaluate changes in bridge bearing stiffness. Under normal bridge operating conditions, use accelerometers and laser Doppler vibrometers to synchronously collect dynamic response signals from the bridge bearings and beams. Changes in bridge bearing stiffness are calculated using a dynamic displacement ratio and a vibration main frequency reduction ratio. The dynamic displacement ratio represents the ratio of the bearing's dynamic displacement to the beam's dynamic displacement, while the vibration main frequency reduction ratio quantifies stiffness degradation by comparing the change in the current vibration main frequency to the initial vibration main frequency. If the dynamic displacement ratio exceeds 2 times or the main frequency reduction ratio exceeds 20%, a stiffness degradation warning signal is generated. Step 3: Detect the debonding of bridge bearings and analyze the debonding phenomenon using the flexibility matrix method. The flexibility matrix describes the deformation characteristics of bridge bearings after being subjected to forces in different directions. The diagonal elements reflect the direct deformation of the bridge bearing, and the off-diagonal elements reflect the coupled deformation between adjacent components. The standardized diagonal difference ratio of the flexibility matrix is ​​calculated, that is, the ratio of the difference between the diagonal elements and the off-diagonal elements to the average value of the initial diagonal elements. The degree of debonding is quantified. Combined with the dynamic response data of the vehicle impact load and the temperature fluctuation parameters, the debonding expansion trend is predicted through a machine learning model, and the percentage value of the degree of debonding is output. Step 4: Monitor changes in bridge bearing thickness and assess corrosion. Eddy current thickness sensors monitor changes in bridge bearing thickness in real time by detecting the relationship between the sensing signal and the distance from the metal surface, and use a nonlinear compensation model to adjust the thickness. The actual thickness value is calculated based on the quadratic relationship between temperature change and thickness error. Humidity compensation is based on the interference of humidity changes on the sensor signal, and corrections are made using experimental fitting data. When the corrosion rate exceeds 0.1 mm per year or the thickness is reduced to less than 10% of the design value, a corrosion warning is generated. Step 5: Fusion of multi-parameter data and comprehensive evaluation. The crack characteristics, stiffness changes, voiding degree, and thickness changes obtained in steps 1 to 4 are integrated to generate a comprehensive health score. The scoring model dynamically assigns weights to each parameter based on its impact on bearing performance. Step 6: Generate early warning signals and provide maintenance recommendations. Generate graded early warning signals based on the health scores and test results obtained in step 5.

2. A multi-load joint detection method as claimed in claim 1, characterized in that: The camera is directly connected to the data acquisition module via a Gigabit Ethernet interface.

3. The multi-load joint detection method as claimed in claim 1, characterized in that: The eddy current thickness measuring sensor is connected to the data acquisition module via a shielded twisted pair cable, and the shielding layer of the shielded twisted pair cable is grounded to eliminate electromagnetic interference.

4. The multi-load joint detection method according to claim 1, wherein: The acceleration sensor is installed by using M8 stainless steel bolts to fix the acceleration sensor base on the beam body with a torque of 40 Nm, with its sensing axis aligned with the support sliding direction and the vertical direction.

5. The multi-load joint detection method as claimed in claim 1, characterized in that: The data acquisition module is connected to the data processing module via an RS485 interface.

Images