Self-aware and warning seamless expansion joint based on flexible mems sensor array

Abstract

The application relates to the technical field of monitoring, and particularly discloses a self-sensing and early-warning seamless expansion joint based on a flexible MEMS sensing array, which comprises a flexible sensing array module and a data concentrator; the flexible sensing array module is composed of a high-performance elastomer base body, a MEMS sensor array embedded in the base body and a flexible printed circuit; the MEMS sensor array is composed of a plurality of sensing nodes distributed in a grid-shaped dot matrix form, each sensing node is integrated with a three-axis MEMS accelerometer, a MEMS temperature sensor, a MEMS strain gauge and a MEMS microphone; the application endows the seamless expansion joint with real-time, multi-dimensional and self-diagnostic intelligent monitoring capability through the cooperation of hardware integration and algorithm innovation; the integrated module of the flexible base body and the multi-parameter MEMS sensing array solves the problem of long-term reliable cooperative work of the sensor and the flexible deformation body, and realizes the synchronous intensive sensing of multiple physical fields such as temperature, strain, vibration and sound.

Description

Technical Field

[0001] This invention belongs to the field of monitoring technology, specifically relating to a self-sensing and early warning seamless expansion joint based on a flexible MEMS sensing array. Background Technology

[0002] In bridge and road engineering, expansion joints are key components for regulating structural deformation and withstanding vehicle impacts. Their long-term performance degradation directly affects driving safety and structural durability. Currently, health monitoring of expansion joints mainly relies on two methods: one is traditional manual periodic inspection, which involves visually inspecting for visible damage, detachment, or leakage using simple tools; the other is advanced sensing technology, such as distributed fiber optic sensing, which embeds sensors within the structure to achieve continuous strain and temperature measurement along the fiber optic path. In addition, some research has explored installing discrete point sensors, such as MEMS accelerometers, on the structural surface to monitor vibration response.

[0003] However, existing technologies have significant limitations. Manual inspection methods are highly subjective and have long intervals, failing to achieve early detection and real-time warnings of defects, representing a typical passive and delayed maintenance approach. While distributed optical fibers can provide continuous spatial information, their high system cost, complex on-site deployment, and long-term reliability heavily influenced by construction and environmental factors, coupled with their relatively limited sensing capabilities, make them insufficiently sensitive to identifying localized damage such as voids and internal micro-cracks, hindering accurate defect type and severity assessment. Furthermore, sparsely arranged conventional point sensors struggle to capture local gradient characteristics of damage, and their rigid encapsulation is incompatible with the flexible deformation of the expansion joint substrate, easily leading to sensor debonding, damage, or measurement signal distortion. Summary of the Invention

[0004] The purpose of this invention is to provide a self-sensing and early warning seamless expansion joint based on a flexible MEMS sensing array, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensing array, comprising:

[0007] Flexible sensor array module and data concentrator;

[0008] The flexible sensing array module consists of a high-performance elastomer matrix, a pre-embedded MEMS sensor array, and a flexible printed circuit.

[0009] The MEMS sensor array consists of multiple sensing nodes distributed in a grid-like array, and each sensing node integrates a triaxial MEMS accelerometer, a MEMS temperature sensor, a MEMS strain gauge, and a MEMS microphone.

[0010] The sensing nodes are connected via flexible printed circuits;

[0011] The data concentrator has a built-in data acquisition chip, microprocessor, and wireless transmission module, which are electrically connected to the flexible sensor array module and encapsulated in a sealed electronic compartment at the end of the expansion joint.

[0012] Preferably, the high-performance elastomer matrix has a loss factor greater than 0.15 within an operating temperature range of -40°C to 80°C, and its adhesion strength to the flexible printed circuit is not less than 5 MPa.

[0013] Preferably, the flexible sensor array module is prefabricated in the factory as an independent, mass-producible structural unit, and then combined with the wear-resistant layer and waterproof layer of the expansion joint during on-site installation.

[0014] A method for monitoring and self-diagnosing seamless expansion joints as described in any of the above claims, comprising the following steps:

[0015] S1. Read the temperature sensor data of each sensing node, generate and analyze the temperature field cloud map of the entire expansion joint section;

[0016] S2. Read the difference in MEMS strain gauge readings between adjacent nodes and calculate the local strain gradient; at the same time, read the static tilt angle value of the MEMS accelerometer to determine the relative settlement or warping of the local area.

[0017] S3. Determine the vehicle position and speed based on the dynamic acceleration response sequence triggered at each node when the vehicle passes by; estimate the axle load based on the dynamic response amplitude; and roughly classify the vehicle type based on the wheelbase and response mode.

[0018] S4. The noise signal generated by the interaction between the expansion joint and the wheel when the vehicle passes through is collected by a MEMS microphone, and its sound pressure level and spectrum characteristics are analyzed to evaluate driving comfort and identify abnormal noises.

[0019] S5. Internal disease diagnosis, including:

[0020] S5a, By comparing the dynamic acceleration signal spectrum of node i when the vehicle passes over it. With health baseline spectrum Calculate the two in the key frequency band Spectral correlation coefficient within , ,when Below the threshold At that time, it was determined that there was a void in the area where node i was located; S5b, monitor the static strain value of node j. Calculate its average strain compared to surrounding nodes. deviation Simultaneously, the cumulative energy of high-frequency acoustic emission events captured by the MEMS accelerometer at node j per unit time is statistically analyzed. When both conditions are met and At time, an internal crack warning is triggered at node j; S5c, during the vehicle-free silent period, the expansion joint is excited to vibrate freely, and its first N natural frequencies are measured using a MEMS sensor array. Calculate the weighted aging index ,in The weighting coefficients for the nth frequency are... For the initial natural frequency, when Exceeding the threshold At that time, it was determined that the material had undergone significant aging.

[0021] Preferably, in step S3, the rough classification of vehicle types includes distinguishing between passenger cars and multi-axle trucks.

[0022] Preferably, in step S4, the identification of abnormal sounds specifically involves: establishing a noise spectrum fingerprint database of expansion joints under different health conditions; and performing spectrum analysis on the noise signals collected in real time to obtain feature vectors. Calculate its feature vectors with those of each disease template in the fingerprint database. Matching degree If the matching degree If the warning threshold for the corresponding disease is exceeded, an abnormal sound alarm will be triggered.

[0023] Preferably, in step S5a, the key frequency band This was determined by comparing and analyzing the acceleration signal spectra under healthy and known detached states.

[0024] Preferably, in step S5c, the weighting coefficient The sensitivity of each natural frequency to material stiffness degradation is determined by analysis.

[0025] A self-sensing and early warning seamless expansion joint system for performing the method described in any of the above claims, comprising:

[0026] The flexible sensing array module consists of a high-performance elastomer matrix, a pre-embedded MEMS sensor array, and a flexible printed circuit, and is used to sense multi-physics field signals.

[0027] The data acquisition and transmission module is located in the sealed electronic compartment at the end of the expansion joint and is used to power the sensor, acquire data and transmit it wirelessly.

[0028] The data processing and diagnosis module is used to receive data and execute the temperature field analysis, deformation analysis, load and traffic analysis, noise analysis and internal defect diagnosis algorithms as described in any one of claims 4-8.

[0029] The early warning feedback module is used to generate a health status report based on the output of the diagnostic module and send early warning information to the management platform.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] (1) Through the synergy of hardware integration and algorithm innovation, seamless expansion joints are endowed with real-time, multi-dimensional, and self-diagnostic intelligent monitoring capabilities. The integrated module of "flexible substrate-multi-parameter MEMS sensor array" solves the problem of long-term reliable collaborative work between sensors and flexible deformable bodies, realizes synchronous dense sensing of multiple physical fields such as temperature, strain, vibration, and sound, and develops special diagnostic algorithms based on multi-source information fusion (such as vibration spectrum correlation analysis, strain-acoustic emission fusion judgment, noise fingerprint matching, etc.), realizing early and accurate identification and location of hidden defects such as voids, internal cracks and material aging, and completing the fundamental leap from passive "continuous sensing" to active "intelligent diagnosis".

[0032] (2) The traditional periodic and delayed maintenance that relies on manual labor is transformed into data-based preventive and precise intelligent management and maintenance. This not only significantly improves the safety and durability of key structures such as bridges and high-grade highways, but also reduces the total life cycle cost through predictive maintenance. At the same time, it ensures the comfort of travel by monitoring driving noise and abnormal conditions, and provides an efficient and feasible technical path for the digital operation and maintenance of intelligent transportation infrastructure. Attached Figure Description

[0033] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Example:

[0036] Please see Figure 1As shown, a self-sensing and early warning seamless expansion joint based on a flexible MEMS sensing array includes:

[0037] Flexible sensor array module and data concentrator;

[0038] The flexible sensor array module consists of a monolithically formed high-performance elastomer substrate, a pre-embedded MEMS sensor array, and a flexible printed circuit (FPC) integrated therewith.

[0039] In this embodiment, the elastomer matrix is ​​made of polyurethane elastomer material with excellent weather resistance. It maintains elasticity in a wide temperature range of -40℃ to 80℃, and has a loss factor (tanδ) greater than 0.15, which can effectively dissipate some vibration energy and protect the internal sensor. The matrix is ​​bonded to the flexible printed circuit with a special adhesive. The bonding strength is tested to be no less than 8MPa, ensuring reliable bonding under long-term cyclic deformation.

[0040] The MEMS sensor array consists of 64 (8×8 grid) sensing nodes embedded in an elastomer matrix in a grid pattern with equal spacing (e.g., 100 mm). Each sensing node is a multifunctional microsystem chip with a size of approximately 5 mm × 5 mm × 1 mm. It integrates a triaxial MEMS accelerometer (range ±50 g, bandwidth 0-5 kHz), a MEMS digital temperature sensor (accuracy ±0.5 ℃), a MEMS piezoresistive strain gauge (gauge length 10 mm, sensitivity coefficient 2.0), and a MEMS silicon microphone (frequency response 20 Hz-20 kHz). All sensing nodes are interconnected for power and signals through flexible printed circuits etched on a polyimide substrate. The circuit traces adopt a serpentine layout to adapt to the stretching of the substrate.

[0041] The data concentrator is a sealed metal housing with an IP67 protection rating, installed in a prefabricated electronic compartment at one end of the expansion joint. It integrates a multi-channel data acquisition chip (such as a 24-bit ADC), an ARM Cortex-M7 series microprocessor, a 4G / 5G wireless communication module, and a rechargeable lithium battery pack. The data concentrator is connected to the FPC outlet of the flexible sensor array module through a waterproof connector. It is responsible for powering all sensors, synchronously acquiring multiple signals, performing preliminary data preprocessing (such as filtering and downsampling), and sending data packets to a remote cloud server or edge computing gateway via a wireless network.

[0042] The flexible sensor array module is molded in the factory in one go using a molding process to form an independent, flexible "smart pad" unit that is about 2 meters long, 0.3 meters wide, and 15 mm thick. This unit can be mass-produced and laid directly at the bottom of the expansion joint groove on the construction site. Then, an epoxy resin wear-resistant layer and a polymer-modified asphalt waterproof layer are laid on top of it, thereby realizing the integrated integration of the sensing system and the expansion joint structure without changing the appearance and basic function of the original expansion joint.

[0043] A monitoring and self-diagnosis method for any of the above-mentioned seamless expansion joints includes the following steps:

[0044] S1. Read the temperature sensor data of each sensing node, generate and analyze the temperature field cloud map of the entire expansion joint section;

[0045] Specifically, the data concentrator synchronously reads temperature sensor data from all 64 nodes every 5 minutes. After receiving the data, the server uses the Kriging spatial interpolation algorithm to generate a high-resolution (e.g., 0.01m grid) two-dimensional temperature field cloud map covering the entire expansion joint cross-section. By analyzing the cloud map, localized temperature anomaly areas caused by uneven sunlight or internal water accumulation can be identified, providing environmental load input for deformation analysis.

[0046] S2. Read the difference in MEMS strain gauge readings between adjacent nodes and calculate the local strain gradient; at the same time, read the static tilt angle value of the MEMS accelerometer to determine the relative settlement or warping of the local area.

[0047] Specifically, strain gradient analysis: Static strain values ​​for all nodes are collected every 30 seconds. For any given node, the difference between its strain readings and those of its four adjacent nodes (east, south, west, and north) is calculated. Local strain gradient vector. Approximate calculation using the central difference method; if the gradient magnitude of a certain region... If the stress consistently exceeds the threshold (e.g., 50 microstrain / meter), it indicates that there is abnormal stress concentration or local debonding at that location.

[0048] Static tilt settlement judgment: Using the static output of the triaxial accelerometer when there are no vehicles, the tilt angle (Roll and Pitch) of each node relative to the horizontal reference plane is calculated. By comparing the tilt angle difference between adjacent nodes, if the tilt angle of a certain node area continues to deviate from the average value of the surrounding area by more than 0.5 degrees, and combined with the strain anomaly at that location, it can be judged that the local area has experienced relative settlement or warping.

[0049] S3. Determine the vehicle position and speed based on the dynamic acceleration response sequence triggered at each node when the vehicle passes by; estimate the axle load based on the dynamic response amplitude; and roughly classify the vehicle type based on the wheelbase and response mode.

[0050] Specifically, when a vehicle passes by, the system triggers a high-speed acquisition mode (such as a 1kHz sampling rate) to record the dynamic response of the accelerometers at each node.

[0051] Vehicle positioning and speed estimation: Based on the time sequence of acceleration signals being triggered at each node, combined with the known coordinates of the nodes, the position of the vehicle wheel track is calculated using the time difference positioning method. The instantaneous speed of the vehicle is estimated based on the time difference between the triggering of different nodes by two wheel signals on the same axle and the distance between the nodes.

[0052] Axle load estimation and vehicle classification: Vehicle axle load The estimation is based on the relationship between the dynamic response amplitude and the calibration coefficient, using a simplified formula. ,in and The parameters are pre-calibrated for a calibration vehicle of known weight. The peak value of the vertical acceleration at the node when the vehicle axle passes through is used to roughly classify vehicles into "passenger cars" (two axles, short wheelbase, light axle load) and "multi-axle trucks" (three or more axles, long wheelbase, heavy axle load) by identifying the time interval (wheelbase) and axle load pattern of continuous axle signals.

[0053] S4. The noise signal generated by the interaction between the expansion joint and the wheel when the vehicle passes through is collected by a MEMS microphone, and its sound pressure level and spectrum characteristics are analyzed to evaluate driving comfort and identify abnormal noises.

[0054] Specifically, as the vehicle passes by, the audio signal from the MEMS microphone is collected simultaneously (sampling rate 44.1kHz).

[0055] Comfort assessment: Calculate the A-weighted sound pressure level of the noise signal ( As a quantitative indicator of driving comfort.

[0056] Abnormal sound identification and disease cross-validation: Establishing a noise spectrum "fingerprint database" including typical scenarios such as "healthy state," "void state," and "cracked state." Fingerprint feature vector This may include the amplitude, centroid, and roll-off rate of the 1 / 3 octave band spectrum;

[0057] After preprocessing and FFT transformation, the same feature vectors are extracted from the noise signals acquired in real time. Cosine similarity is used to calculate its similarity with each template vector in the fingerprint database. Matching degree .like If the similarity to a certain disease template exceeds a preset threshold (e.g., 0.85), an abnormal sound alarm for that disease will be triggered. This result can be cross-validated with the diagnostic results in S5 to improve the reliability of the early warning.

[0058] S5. Internal disease diagnosis, including:

[0059] S5a, By comparing the dynamic acceleration signal spectrum of node i when the vehicle passes over it. With health baseline spectrum Calculate the two in the key frequency band Spectral correlation coefficient within , ,when Below the threshold When node i is located, it is determined that there is a void in the region.

[0060] In the initial healthy state after the expansion joint installation, at least 100 vertical acceleration time-history signals of each node i were collected when a standard vehicle (such as a passenger car) passed by at a constant speed. Each signal was subjected to an FFT transform, and its average power spectral density was calculated as the healthy baseline spectrum of that node. ;

[0061] By comparing and analyzing the acceleration spectra of healthy and laboratory-simulated vacuolation specimens, it was found that the 200Hz to 800Hz frequency band is most sensitive to the structural impedance changes caused by vacuolation. Therefore, a critical frequency band was set. [200Hz, 800Hz];

[0062] As a vehicle passes, the current spectrum of node i is calculated in real time. ,calculate and Spectral correlation coefficient within the key frequency band In a healthy state Typically greater than 0.92, a threshold is set. =0.85. If in three consecutive vehicle events, the value of a certain node... If the mean value is below 0.85, the system determines that there is voiding in the area covered by the node and marks its location on the expansion joint plan; S5b, the static strain value of monitoring node j. Calculate its average strain compared to surrounding nodes. deviation Simultaneously, the cumulative energy of high-frequency acoustic emission events captured by the MEMS accelerometer at node j per unit time is statistically analyzed. When both conditions are met and At that time, an early warning of internal cracking is triggered at node j;

[0063] Specifically, strain deviation monitoring: real-time monitoring of the static strain of all nodes. For node j, calculate the average strain between it and its 8 neighboring nodes (3×3 window). deviation threshold Set to 3 times the standard deviation of the long-term average deviation (e.g., about 30 microstrain).

[0064] Acoustic emission event monitoring: Bandpass filtering (50kHz-200kHz) is applied to the accelerometer signal, and an amplitude threshold is set to capture high-frequency acoustic emission (AE) transients that may be generated by microcracks. The cumulative energy of AE events within a unit time (e.g., 1 hour) at node j is statistically analyzed. Threshold Set according to the background noise level.

[0065] Fusion early warning: When the system detects that for a certain node j, within a time window Δt (e.g., 1 hour), simultaneously satisfying... and If this occurs, an early warning of "suspected internal crack initiation or propagation at node j" will be immediately triggered, with the warning level being "high"; S5c, during the vehicle-free silent period, the expansion joint is excited to vibrate freely, and its first N natural frequencies are measured using a MEMS sensor array. Calculate the weighted aging index ,in The weighting coefficients for the nth frequency are... For the initial natural frequency, when Exceeding the threshold At that time, it was determined that the material had undergone significant aging;

[0066] Specifically, initial frequency measurement: Within one week after the expansion joint is installed, during a quiet period at night when there are no vehicles, a miniature electromagnetic exciter built into the data concentrator is triggered remotely to provide a pulse excitation to the expansion joint. The free-dampening vibration response of the structure is recorded using all accelerometers, and the first three (N=3) vertical bending natural frequencies are identified using either the peak power method (PP method) or the random subspace identification method (SSI). , as the initial reference;

[0067] Regular monitoring and aging index calculation: Repeat the above incentive and measurement process every quarter to obtain the current inherent frequency. Based on finite element analysis and mechanics of materials, the first-order frequency is typically most sensitive to overall stiffness, while the second and third-order frequencies are more sensitive to local material degradation. Therefore, weighting coefficients are set... Calculate the weighted aging index using the formula above. ;

[0068] Aging assessment: Studies show that when the stiffness of concrete or elastomer materials decreases by 5%, their low-order natural frequencies change by approximately 2-3%. (Setting a threshold) =0.03 (corresponding to approximately 3% average frequency shift); if If the system generates a diagnostic report stating "significant aging / stiffness degradation of the material," it recommends a material performance review.

[0069] In an embodiment of the present invention, step S3, the rough classification of vehicle types includes distinguishing between passenger cars and multi-axle trucks.

[0070] In an embodiment of the present invention, step S4, identifying abnormal sounds specifically involves: establishing a noise spectrum fingerprint database of expansion joints under different health conditions; and performing spectrum analysis on the noise signals collected in real time to obtain feature vectors. Calculate its feature vectors with those of each disease template in the fingerprint database. Matching degree If the matching degree If the warning threshold for the corresponding disease is exceeded, an abnormal sound alarm will be triggered.

[0071] In an embodiment of the present invention, in step S5a, the key frequency band This was determined by comparing and analyzing the acceleration signal spectra under healthy and known detached states.

[0072] In an embodiment of the present invention, in step S5c, the weighting coefficient The sensitivity of each natural frequency to material stiffness degradation is determined by analysis.

[0073] A self-sensing and early warning seamless expansion joint system for performing any of the above methods, comprising:

[0074] The flexible sensing array module consists of a high-performance elastomer matrix, a pre-embedded MEMS sensor array, and a flexible printed circuit, and is used to sense multi-physics field signals.

[0075] Specifically, this module is responsible for the original sensing of multi-physics field signals. Its specific structure consists of a high-strength, high-elasticity, and aging-resistant polyurethane matrix, which is formed in one piece by a precision mold. The internal structure is pre-embedded with multi-functional MEMS sensor chips in an 8×8 grid matrix (a total of 64 nodes). Each chip encapsulates three-dimensional acceleration, temperature, strain, and sound pressure sensing units. The pins of the sensor chips are firmly connected to the pads of the polyimide flexible printed circuit (FPC) below through gold wire bonding. The FPC adopts a hollow serpentine trace design to ensure that the circuit connection is reliable and does not affect the signal integrity when the matrix is ​​subjected to tensile, compressive, and torsional deformation. Before leaving the factory, the module undergoes full-function testing and basic calibration to form a standardized "intelligent sensing pad".

[0076] The data acquisition and transmission module is located in the sealed electronic compartment at the end of the expansion joint and is used to power the sensor, acquire data and transmit it wirelessly.

[0077] Specifically, this module is responsible for power supply and signal conditioning and transmission. Its main hardware component is a circuit board integrated within a cast aluminum sealed shell, installed in a prefabricated dedicated electronic compartment at one end of the expansion joint, which features drainage and impact protection. Specifically, it includes:

[0078] The power management unit incorporates a high-capacity lithium thionyl chloride battery pack, working in conjunction with a solar panel charging controller to achieve long-term maintenance-free power supply. It provides multiple regulated outputs to power the sensors and its own circuitry.

[0079] The multi-channel data acquisition unit employs a high-precision, synchronously sampled analog front-end chip, supporting synchronous or on-demand acquisition of acceleration (dynamic and static), strain, temperature, and audio signals from all 64 nodes. The sampling rate can be dynamically adjusted according to the monitoring mode (e.g., 1Hz for static monitoring, 1kHz for dynamic event capture, and 50kHz for acoustic emission analysis).

[0080] The microcontroller unit (MCU) uses a high-performance ARM Cortex-M series processor to control the acquisition timing, perform preliminary data preprocessing (such as digital filtering, noise reduction, and compression), encapsulate data packets, and manage communication protocols.

[0081] The wireless transmission unit integrates a 4G / 5G cellular communication module and a Bluetooth Low Energy (BLE) module. The cellular network is used to periodically send data to and receive commands from a remote server; BLE is used for near-field device configuration and data retrieval during on-site maintenance. The transport layer protocol ensures data reliability and real-time performance.

[0082] The data processing and diagnosis module is used to receive data and execute temperature field analysis, deformation analysis, load and traffic analysis, noise analysis and internal defect diagnosis algorithms as described above.

[0083] Specifically, this module runs on a cloud server or edge computing gateway, responsible for in-depth information processing and intelligent diagnostics. It adopts a modular software architecture, which includes:

[0084] The data access and storage service receives data streams from multiple field devices, parses and verifies them, and stores them in a time-series database.

[0085] Algorithm engine cluster:

[0086] The physics analysis engine performs tasks such as temperature field interpolation and plotting, strain gradient calculation, and static tilt angle calculation.

[0087] Traffic parameter extraction engine operates vehicle detection, positioning, speed measurement, axle load estimation, and vehicle type classification algorithms;

[0088] The acoustic analysis engine performs feature extraction, spectral analysis, and fingerprint database matching of noise signals;

[0089] The core disease diagnosis engine encapsulates a series of dedicated diagnostic algorithm models, such as the spectrum correlation analysis, strain-acoustic emission fusion judgment, and natural frequency aging index calculation described above. These models use historical health data and preset thresholds to automatically analyze the received multi-source fusion data.

[0090] The diagnostic decision service integrates the output results of various algorithm engines and forms the final structural health status assessment conclusion and warning level based on the preset warning logic (such as multi-indicator cross-validation).

[0091] The early warning feedback module is used to generate a health status report based on the output of the diagnostic module and send early warning information to the management platform.

[0092] Specifically, this module is responsible for transforming diagnostic conclusions into actionable decision-making information. The specific implementation is as follows:

[0093] The early warning information generator automatically generates a structured health status report based on the status code and severity level output by the diagnostic module. The report includes: disease type (such as "local void"), location (located to specific grid coordinates), severity, development trend chart, and maintenance recommendations.

[0094] The multi-channel information publishing interface pushes early warning reports and real-time data to the road maintenance management platform (CMMS) via standard APIs; at the same time, it can send instant alarm information to designated maintenance management personnel via configured SMS, email or application push methods.

[0095] The optional human-computer interaction interface provides a visual dashboard on the web or mobile device, displaying the real-time status cloud map of the expansion joint, historical data curves, a list of early warning events, and the system's own operating status, and supports interactive querying and report downloading.

[0096] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array, characterized in that, include: Flexible sensor array module and data concentrator; The flexible sensing array module consists of a high-performance elastomer matrix, a pre-embedded MEMS sensor array, and a flexible printed circuit. The MEMS sensor array consists of multiple sensing nodes distributed in a grid-like array, and each sensing node integrates a triaxial MEMS accelerometer, a MEMS temperature sensor, a MEMS strain gauge, and a MEMS microphone. The sensing nodes are connected via flexible printed circuits; The data concentrator has a built-in data acquisition chip, microprocessor, and wireless transmission module, which are electrically connected to the flexible sensor array module and encapsulated in a sealed electronic compartment at the end of the expansion joint.

2. The self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 1, characterized in that: The high-performance elastomer matrix has a loss factor greater than 0.15 within a working temperature range of -40℃ to 80℃, and its adhesion strength to the flexible printed circuit is not less than 5MPa.

3. The self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 1, characterized in that: The flexible sensor array module is prefabricated in the factory as an independent, mass-producible structural unit, and then combined with the wear-resistant and waterproof layers of the expansion joint during on-site installation.

4. A method for monitoring and self-diagnosing the seamless expansion joint according to any one of claims 1-3, characterized in that, Includes the following steps: S1. Read the temperature sensor data of each sensing node, generate and analyze the temperature field cloud map of the entire expansion joint section; S2. Read the difference in MEMS strain gauge readings between adjacent nodes and calculate the local strain gradient; at the same time, read the static tilt angle value of the MEMS accelerometer to determine the relative settlement or warping of the local area. S3. Determine the vehicle position and speed based on the dynamic acceleration response sequence triggered at each node when the vehicle passes by; estimate the axle load based on the dynamic response amplitude; and roughly classify the vehicle type based on the wheelbase and response mode. S4. The noise signal generated by the interaction between the expansion joint and the wheel when the vehicle passes through is collected by a MEMS microphone, and its sound pressure level and spectrum characteristics are analyzed to evaluate driving comfort and identify abnormal noises. S5. Internal disease diagnosis, including: S5a, By comparing the dynamic acceleration signal spectrum of node i when the vehicle passes over it. With health baseline spectrum Calculate the two in the key frequency band Spectral correlation coefficient within , ,when Below the threshold At that time, it was determined that there was a void in the area where node i was located; S5b, monitor the static strain value of node j. Calculate its average strain compared to surrounding nodes. deviation Simultaneously, the cumulative energy of high-frequency acoustic emission events captured by the MEMS accelerometer at node j per unit time is statistically analyzed. When both conditions are met and At time, an internal crack warning is triggered at node j; S5c, during the vehicle-free silent period, the expansion joint is excited to vibrate freely, and its first N natural frequencies are measured using a MEMS sensor array. Calculate the weighted aging index ,in The weighting coefficients for the nth frequency are... For the initial natural frequency, when Exceeding the threshold At that time, it was determined that the material had undergone significant aging.

5. A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 4, characterized in that: In step S3, the rough classification of vehicle types includes distinguishing between passenger cars and multi-axle trucks.

6. A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 4, characterized in that: In step S4, the identification of abnormal sounds specifically involves: establishing a noise spectrum fingerprint database for expansion joints under different health conditions; and performing spectrum analysis on the noise signals collected in real time to obtain feature vectors. Calculate its feature vectors with those of each disease template in the fingerprint database. Matching degree If the matching degree If the warning threshold for the corresponding disease is exceeded, an abnormal sound alarm will be triggered.

7. A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 4, characterized in that: In step S5a, the key frequency band This was determined by comparing and analyzing the acceleration signal spectra under healthy and known detached states.

8. A self-sensing and early warning seamless expansion joint based on a flexible MEMS sensor array according to claim 4, characterized in that: In step S5c, the weighting coefficient The sensitivity of each natural frequency to material stiffness degradation is determined by analysis.

9. A self-sensing and early warning seamless expansion joint system for performing the method according to any one of claims 1-8, characterized in that, include: The flexible sensing array module consists of a high-performance elastomer matrix, a pre-embedded MEMS sensor array, and a flexible printed circuit, and is used to sense multi-physics field signals. The data acquisition and transmission module is located in the sealed electronic compartment at the end of the expansion joint and is used to power the sensor, acquire data and transmit it wirelessly. The data processing and diagnosis module is used to receive data and execute the temperature field analysis, deformation analysis, load and traffic analysis, noise analysis and internal defect diagnosis algorithms as described in any one of claims 4-8. The early warning feedback module is used to generate a health status report based on the output of the diagnostic module and send early warning information to the management platform.

Images