A wireless sensor network synchronous acquisition method for structural health monitoring

Abstract

The application discloses a wireless sensor network synchronous acquisition method for structural health monitoring, comprising the following steps: constructing a tree-shaped synchronous network based on signal strength and hop count; performing bidirectional time exchange by using a physical layer frame start delimiter capture function to eliminate interruption delay; monitoring and dynamically compensating crystal oscillator frequency drift in real time by using a temperature sensor; adaptively adjusting a synchronous period according to structural vibration amplitude and temperature change rate; automatically performing fast resynchronization and route repair when a node is out of step or a link is deteriorated; and performing time stamp checking and quality calibration on collected data. The application realizes microsecond-level synchronization accuracy, greatly reduces power consumption, has an abnormal self-healing capability, and is suitable for long-term unattended health monitoring of infrastructures such as bridges and high-rise buildings.

Description

Technical Field

[0001] This invention relates to the field of structural health monitoring technology, and in particular to a method for synchronous data acquisition using a wireless sensor network for structural health monitoring. Background Technology

[0002] Structural health monitoring systems use a network of sensors deployed at critical structural locations to collect real-time data on vibration, strain, temperature, and other responses of the structure for structural condition assessment and damage identification. Traditional monitoring systems often employ wired connections, which suffer from complex wiring, high costs, and difficult maintenance, and are gradually being replaced by wireless sensor networks.

[0003] In data acquisition within wireless sensor networks, the accuracy of time synchronization among multiple nodes directly impacts the accuracy of subsequent modal analysis and damage identification. Existing technologies primarily employ the following synchronization methods: First, GPS time synchronization. This method relies on Global Navigation Satellite System signals, which cannot be used in areas with severe signal obstruction, such as inside steel structures or tunnels. Furthermore, GPS modules consume a lot of power, making them unsuitable for monitoring scenarios requiring long-term battery power.

[0004] Second, software timestamp synchronization based on the IEEE 1588 protocol. This method is implemented at the application layer of the embedded operating system. Affected by the operating system scheduling delay and the uncertainty of interrupt response, the synchronization accuracy is usually in the millisecond range, which is difficult to meet the phase consistency requirements of high-frequency structural vibration signals above 100Hz.

[0005] Third, fixed-period synchronization. This method calibrates the clock at fixed time intervals, which cannot be dynamically adjusted according to environmental changes and operating conditions, resulting in a contradiction between "synchronization accuracy" and "power consumption".

[0006] Furthermore, existing technologies generally neglect the long-term drift effect of ambient temperature on crystal oscillator frequencies, leading to increased cumulative errors between synchronization intervals. They also lack self-healing capabilities under abnormal conditions such as node failures and signal blockages, failing to meet the engineering requirements of long-term unattended operation and high reliability for structural health monitoring. Summary of the Invention The purpose of this invention is to address the shortcomings of existing technologies by proposing a wireless sensor network synchronous acquisition method for structural health monitoring.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A method for synchronous data acquisition using a wireless sensor network for structural health monitoring includes the following steps: Step S1: Network topology construction. The main control gateway sends a broadcast beacon, and multiple wireless acquisition nodes are automatically divided into primary synchronization nodes and secondary synchronization nodes based on the received signal strength and hop count, forming a tree-like synchronization network. Step S2: Physical layer hardware timestamp synchronization. The main control gateway and the wireless acquisition node perform bidirectional time exchange. The wireless acquisition node uses the physical layer frame start delimiter capture function of the wireless transceiver chip to automatically latch the local crystal oscillator count value as the receiving time at the rising edge of the frame start delimiter of the received synchronization frame, eliminating the interrupt delay between the protocol stack and the processor. It also calculates the transmission delay and clock offset based on the four timestamps of the bidirectional time exchange and adjusts the local virtual clock. Step S3: Temperature-sensing dynamic compensation of crystal oscillator frequency. Each wireless acquisition node monitors the temperature in real time through a built-in temperature sensor, calculates the current frequency error according to the preset temperature-frequency offset model, dynamically adjusts the timer reload value, and corrects the crystal oscillator frequency drift within the two synchronization intervals. Step S4: Adaptive synchronization cycle adjustment. The main control gateway dynamically adjusts the network synchronization cycle according to the working conditions of structural health monitoring: when the structural response amplitude exceeds the preset threshold, it switches to high dynamic mode and shortens the synchronization cycle; when the structural response amplitude is lower than the preset threshold and the temperature change rate is lower than the preset threshold, it switches to static monitoring mode and extends the synchronization cycle. The wireless acquisition node enters sleep mode within the synchronization interval. Step S5: Synchronous data acquisition and data transmission. All wireless acquisition nodes, under a unified virtual clock domain, initiate analog-to-digital conversion according to the global trigger command or collaborative acquisition timetable of the main control gateway. After the acquired data is stamped with a global timestamp, it is transmitted back to the main control gateway through the time division multiple access mechanism.

[0008] Preferably, the wireless acquisition node includes a microcontroller, a wireless transceiver chip with physical layer frame start delimiter capture function, a temperature-compensated crystal oscillator, a temperature sensor, an analog-to-digital conversion acquisition unit, and a power management unit; the capture output terminal of the wireless transceiver chip is connected to the capture input pin of the microcontroller; the analog-to-digital conversion acquisition unit is triggered by a timer of the microcontroller and is linked with the synchronization signal of the wireless transceiver chip.

[0009] Preferably, the bidirectional time exchange in step S2 specifically includes: The main control gateway at all times Send synchronization frames; Wireless acquisition node records reception time ; Wireless data acquisition nodes at any time A response frame is provided, the response frame containing and ; The main control gateway at all times Receive response frame; Calculate transmission delay and clock offset The wireless acquisition node adjusts the timer counter frequency via a software phase-locked loop based on the clock offset.

[0010] Preferably, in step S2, adjusting the local virtual clock uses a software phase-locked loop to adjust the timer counter frequency, rather than directly modifying the system time, to avoid time jumps affecting the continuity of data acquisition.

[0011] Preferably, the temperature-frequency offset model in step S3 is obtained by: placing the wireless acquisition node in a temperature-controlled chamber, measuring the actual frequency of the crystal oscillator at different temperature points, and fitting a quadratic function using the least squares method. ,in The current temperature. For frequency error, , , represents the fitting coefficient.

[0012] Preferably, the adaptive synchronization cycle adjustment in step S4 further includes a temperature transient mode: when the wireless acquisition node detects that the temperature change rate exceeds the second preset threshold, the node actively requests the main control gateway to temporarily shorten the synchronization cycle, and restores it after the temperature stabilizes.

[0013] Preferably, it also includes anomaly handling and self-healing steps: If a wireless acquisition node fails to complete bidirectional time exchange for a preset number of consecutive times, the node is marked as out of sync. A stray node actively initiates a fast resynchronization process with its neighboring synchronized nodes, using the local time of the neighboring nodes as a reference to recalibrate its clock. If fast resynchronization fails, the out-of-sync node enters sleep-retry mode and attempts to rejoin the network at random backoff intervals. When a secondary synchronization node detects that the link quality with its parent node is lower than the link threshold, it automatically searches for other primary or secondary nodes in the vicinity and reconstructs the synchronization tree.

[0014] Preferably, it also includes a health status reporting step: the wireless acquisition node packages the synchronization status, link quality, temperature compensation amount, and crystal oscillator frequency deviation information into a health status frame and periodically transmits it back to the main control gateway.

[0015] Preferably, it also includes a data quality verification step: Each wireless acquisition node contains the theoretical trigger time, the actual trigger time, the clock offset of the last synchronization time, and the temperature compensation correction value in each data packet; After receiving the data, the main control gateway calculates the deviation between the actual trigger time and the theoretical trigger time. If the deviation exceeds the preset accuracy threshold, the data packet is marked as a synchronization quality warning. If multiple data packets consecutively trigger synchronization quality warnings, the master gateway will trigger a forced resynchronization command for that node.

[0016] Preferably, it also includes: the main control gateway has a built-in global navigation satellite system timing module and a temperature-controlled crystal oscillator. When the global navigation satellite system signal is valid, the main control gateway uses a Kalman filter algorithm to tame the temperature-controlled crystal oscillator to form a reference time domain; when the global navigation satellite system signal fails, the main control gateway uses the temperature-controlled crystal oscillator as the reference time source.

[0017] The present invention has the following beneficial effects: 1. This invention utilizes physical layer hardware timestamp capture technology to automatically latch the count value the instant the wireless transceiver chip receives the frame start delimiter. This completely eliminates the uncertain delays caused by protocol stack processing and operating system interrupt responses. Combined with dynamic frequency compensation from a temperature-compensated crystal oscillator, it can still stably control the synchronization accuracy of the entire network within a certain range even in structures without global navigation satellite system signal coverage. This meets the phase consistency requirements for structural vibration modal analysis above 100Hz, providing a reliable data foundation for high-precision structural damage identification.

[0018] 2. This invention adopts a working condition adaptive synchronization cycle adjustment strategy. The main control gateway dynamically adjusts the synchronization frequency according to the structural vibration amplitude and temperature change rate. In the static monitoring mode where the structure is at rest and the environment is stable, the synchronization cycle is extended. The node enters a deep sleep state within the synchronization interval. Compared with the traditional fixed high-frequency synchronization mode, the average power consumption of the node is reduced, which significantly extends the service life of the battery-powered wireless monitoring system and reduces long-term operation and maintenance costs.

[0019] 3. The invention’s improved anomaly handling and self-healing mechanism enables the system to automatically and quickly resynchronize and dynamically repair routes in abnormal situations such as node failure, signal obstruction, and link quality degradation, achieving robust operation under unattended conditions. At the same time, through the timestamp verification field built into the data packet, the main control gateway can perform synchronization quality assessment and marking on each data packet, ensuring that subsequent structural health analysis can identify and remove low-quality data, thereby improving the accuracy and reliability of monitoring results. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] Example 1: Health Monitoring System for Large Suspension Bridge Structures In a large-scale suspension bridge health monitoring project, 50 wireless acquisition nodes and one main control gateway were deployed on the bridge deck and key sections of the main cable. The main control gateway was installed on the top of the bridge tower and was equipped with 4G backhaul and a GNSS antenna. The monitored parameters included triaxial acceleration, strain, and temperature.

[0023] Step S1: Network Topology Construction The master gateway periodically broadcasts beacon frames, which contain the gateway ID, network identifier, and synchronization level information. Upon receiving a beacon frame, each wireless acquisition node measures the Received Signal Strength Indicator (RSSI). Nodes in open areas of the bridge deck with RSSI values ​​higher than -75 dBm are automatically classified as Level 1 synchronization nodes (15 in total). Nodes inside the main cable and box girder have RSSI values ​​lower than -75 dBm and are classified as Level 2 synchronization nodes (35 in total), forwarding synchronization signals through the Level 1 nodes. After determining their level, each node sends a registration frame to the master gateway, completing the network topology construction.

[0024] Step S2: Physical layer hardware timestamp synchronization The gateway acquires UTC time via GNSS, tames the cryogenic crystal oscillator, and establishes a reference time domain. All nodes complete their first bidirectional time exchange within one minute of startup: the gateway at time... When sending a synchronization frame, the node utilizes the SFD capture function of the CC1352R wireless transceiver chip to automatically latch the local crystal oscillator count value as the reception time when the rising edge of the start-of-frame delimiter of the synchronization frame is detected. Interruption delays are completely eliminated; nodes at time... The response frame contains and The gateway at any time Receive the response frame. The node calculates the transmission delay. and clock offset By adjusting the timer counter frequency through a software phase-locked loop, the initial synchronization accuracy reaches ±500 nanoseconds.

[0025] Step S3: Temperature-Sensitive Dynamic Compensation of Crystal Oscillator Frequency Each wireless data acquisition node has a built-in TMP117 temperature sensor with a resolution of ±0.1℃ and a sampling period of 1 second. During the hottest part of the daytime, at an ambient temperature of 40℃, the node's crystal oscillator frequency drift reaches +15ppm. The node reads the current temperature. Substitute into the factory-calibrated temperature-frequency model The calculated frequency error was +15.2ppm. By dynamically adjusting the timer reload value, the ADC sampling frequency was precisely maintained at 1000.000Hz from 1000Hz, effectively offsetting hardware drift.

[0026] Step S4: Adaptive Synchronization Period Adjustment During the nighttime rest period, from 23:00 to 05:00, the gateway detected a structural response amplitude below 0.01g and a temperature change rate of all nodes. The network-wide synchronization cycle is switched to 60 seconds via broadcast control frames. Nodes enter deep sleep mode between synchronizations, retaining only temperature compensation and timer operation, reducing power consumption to 1.2mW. When a sudden heavy vehicle crosses the bridge, the peak vibration acceleration exceeds 0.05g. The gateway detects this within 0.5 seconds and immediately broadcasts a switching command, restoring the network-wide synchronization cycle to 0.5 seconds. Nodes complete clock calibration in the next synchronization cycle.

[0027] Step S5: Synchronous Acquisition and Data Feedback All nodes, operating within a unified virtual clock domain, synchronously initiate ADC acquisition at a specified time based on the gateway's global trigger command. After acquisition is complete, each data packet is encapsulated with its theoretical trigger time. Actual triggering time Clock offset at the last synchronization time and temperature compensation correction value Data backhaul employs a time-division multiple access (TDMA) mechanism, with the gateway allocating an independent time slot to each node to avoid data collisions. After six months of operation, 98.7% of data packets achieved satisfactory synchronization quality. Modal analysis showed that the phase difference of the vibration signals at each node was less than 0.1 degrees, which met the requirements for structural damage identification.

[0028] Example 2: Wind Vibration Monitoring System for Super High-Rise Buildings In a health monitoring project for a 528-meter-high super high-rise building, 30 wireless acquisition nodes were deployed on different floors of the building, namely the 10th, 30th, 50th, 70th, and 90th floors and the crown. The main control gateway was installed in the equipment room on the top floor of the building and connected to the building's monitoring center via fiber optic cable.

[0029] Step S1: Network Topology Construction The main control gateway sends a broadcast beacon, and each node measures its RSSI value. Due to severe signal attenuation caused by the reinforced concrete structure inside the building, the RSSI values ​​of nodes located on the exterior walls are higher than -80dBm, and these nodes are classified as Level 1 synchronization nodes. There are 5 Level 1 synchronization nodes, located on the exterior walls of the 10th, 30th, 50th, 70th, and 90th floors respectively. The RSSI values ​​of nodes located indoors and in the core tube are lower than -80dBm, and these nodes are classified as Level 2 synchronization nodes. There are 25 Level 2 synchronization nodes, which forward synchronization signals through the Level 1 nodes on the corresponding floors.

[0030] Step S2: Physical layer hardware timestamp synchronization After the gateway starts up, it acquires UTC time through the GNSS timing module and tames the temperature-controlled crystal oscillator. Each node completes its first bidirectional time exchange within 2 minutes, using physical layer SFD capture technology to eliminate interruption delays, achieving an initial synchronization accuracy of ±800 nanoseconds. For secondary synchronization nodes, synchronization frames are forwarded through primary nodes, and a time correction term for relay forwarding is added when calculating transmission delay.

[0031] Step S3: Temperature-Sensitive Dynamic Compensation of Crystal Oscillator Frequency Due to significant temperature differences between different floors, each node underwent independent temperature and frequency model calibration. During factory calibration, a calibration point was established every 10°C within the range of -10°C to 50°C to fit a personalized quadratic model. During operation, the node collected temperature data every 2 seconds, and compensation values ​​were calculated based on its respective model. Actual measurements showed that after compensation, the frequency drift of each floor node under diurnal temperature variations decreased from ±25ppm to within ±2ppm.

[0032] Step S4: Adaptive Synchronization Period Adjustment Under windless conditions, the gateway detects that the structural response amplitude is below 0.005g and the temperature change rate of each node is below 0.1℃ / min, switches the synchronization cycle to 90 seconds, and reduces node power consumption to 0.8mW. When a typhoon passes, the building acceleration response amplitude exceeds 0.03g, and the system switches to high dynamic mode within 1 second, shortening the synchronization cycle to 1 second, and continuously records 72 hours of typhoon response data.

[0033] Step S5: Synchronous Acquisition and Data Feedback All nodes synchronously start data collection according to the gateway's collaborative collection schedule, with the sampling frequency set to 50Hz. Data backhaul employs a hybrid mechanism of TDMA and frequency division multiple access, with different communication channels used for different floors to reduce interference. Health status frames are reported every 30 minutes, and maintenance personnel can view the synchronization status of each node in real time through the monitoring center. Post-event modal analysis results show that the measurement error of the vibration phase difference between floors is less than 0.2 degrees, accurately identifying the first three vibration modes of the building.

[0034] Example 3: Offshore Oil Platform Structure Monitoring System In a structural health monitoring project for an offshore oil platform, 45 wireless acquisition nodes were deployed at locations such as the platform deck, key nodes of the jacket, and underwater risers. The main control gateway was installed in the platform control room and transmitted the data back to the land-based monitoring center via satellite communication.

[0035] Step S1: Network Topology Construction The main control gateway sends broadcast beacons. Platform deck nodes have good signal conditions, with RSSI values ​​above -70dBm, and are classified as Level 1 synchronization nodes, with a total of 8 Level 1 synchronization nodes. Due to seawater attenuation, jacket and subsea riser nodes cannot communicate directly with the gateway and are classified as Level 2 synchronization nodes, with a total of 37 Level 2 synchronization nodes. Subsea nodes are connected to surface buoy nodes via watertight cables, and the buoy nodes then communicate wirelessly with the deck nodes, forming a multi-hop synchronization network.

[0036] Step S2: Physical layer hardware timestamp synchronization The gateway obtains the reference time via satellite communication. Each node uses the DW1000 wireless transceiver chip, which features high-precision SFD capture capabilities, enabling nanosecond-level timestamp recording during bidirectional time exchange. Underwater nodes are relayed via buoy nodes, and a fixed delay correction for cable transmission is added to the transmission delay calculation, calibrated to 1.2μs / 100m. After initial synchronization, the synchronization accuracy of the entire network reaches ±1μs.

[0037] Step S3: Temperature-Sensitive Dynamic Compensation of Crystal Oscillator Frequency The underwater node operates at a constant ambient temperature of approximately 10℃ ± 2℃, resulting in minimal crystal oscillator drift. Temperature compensation primarily relies on static correction. The deck node, affected by sunlight and sea breezes, experiences temperature fluctuations between 5℃ and 35℃. Temperature data is collected every second and dynamically adjusted using a temperature-frequency model. Experimental results show that the synchronization accuracy after compensation remains stable within ±0.8μs.

[0038] Step S4: Adaptive Synchronization Period Adjustment Under normal sea conditions, the structural response amplitude is less than 0.02g, and the synchronization period is set to 30 seconds. Under typhoon or extreme wave conditions, the response amplitude exceeds 0.08g, and the system automatically switches to high dynamic mode, shortening the synchronization period to 1 second. When a node detects a temperature change rate exceeding 0.5℃ / min, it actively requests the gateway to temporarily shorten the synchronization period to 2 seconds, restoring it after the temperature stabilizes.

[0039] Step S5: Synchronous Acquisition and Data Feedback All nodes synchronously collect acceleration, strain, and temperature data. The sampling frequency is set to 100Hz on deck and 20Hz underwater, based on monitoring requirements. Data transmission uses a TDMA mechanism, with the gateway allocating time slots to each node. Health status frames are reported every hour, containing information such as synchronization status, link quality, and battery voltage. After 24 months of continuous operation, the system achieved an average node synchronization success rate of 99.5% and a data validity rate of 98.2%.

[0040] Example 4: Construction Monitoring System for Large-Span Spatial Grid Structures During the construction of a 320-meter-span spatial grid structure in a stadium, this invention was used for synchronous stress-strain monitoring during the steel component hoisting and closure stages. Twenty wireless acquisition nodes were deployed at key structural nodes to monitor parameters including strain, temperature, and tilt angle.

[0041] Step S1: Network Topology Construction During construction, the structure was not yet fully formed, and the node distribution dynamically changed with the hoisting progress. The main control gateway was installed on a drone platform, which flew to the top of the structure before the start of each day's construction and sent broadcast beacons. Each node was stratified based on its RSSI value with the drone gateway: 10 nodes with an RSSI higher than -80dBm were designated as Level 1 synchronization nodes, and the remaining 10 nodes forwarded synchronization signals through these Level 1 nodes. During construction, each node automatically performed a link quality assessment every 30 minutes and dynamically adjusted the topology.

[0042] Step S2: Physical layer hardware timestamp synchronization The UAV gateway hovered 30 meters above the structure and transmitted synchronization frames. Each node used physical layer SFD capture technology to record timestamps, completing bidirectional time exchange. Due to strong electromagnetic interference from tower crane operation and welding work in the construction environment, the physical layer synchronization mechanism effectively eliminated the impact of interference on the protocol stack, and the synchronization accuracy was consistently better than ±2μs. During the closure phase, all 20 nodes completed the initial calibration synchronously, with an initial synchronization accuracy of ±1.2μs.

[0043] Step S3: Temperature-Sensitive Dynamic Compensation of Crystal Oscillator Frequency During construction, the ambient temperature varied between 10℃ and 35℃, and the temperature of each node was collected every 2 seconds. Since the nodes were hoisted at different times along with the steel components, each node underwent temperature-frequency model calibration before hoisting. During operation, the nodes calculated compensation values ​​based on real-time temperature and dynamically adjusted the timers. Actual measurements showed that, under daily temperature variations, the crystal oscillator frequency drift was controlled within ±1.5ppm.

[0044] Step S4: Adaptive Synchronization Period Adjustment During non-closure periods, the structure remains in a static state with a response amplitude below 0.01g, a synchronization period of 60 seconds, and the nodes enter low-power mode. During the closure phase, the gateway detects the welding process parameter start signal, automatically switches to high-dynamic mode, shortens the synchronization period to 2 seconds, and continuously acquires strain and temperature data. After welding is completed, the system returns to static monitoring mode.

[0045] Step S5: Synchronous Acquisition and Data Feedback All nodes synchronously initiate strain and temperature acquisition based on the gateway's global trigger command, with the strain sampling frequency set to 10Hz and the temperature sampling frequency set to 1Hz. After the acquired data is timestamped globally, it is transmitted back to the drone gateway via TDMA mechanism, and the gateway uploads the data to the cloud monitoring platform via the 4G network. During the 6-month construction period, the system accumulated more than 5 million sets of valid data, with a data integrity rate of 97.5%, successfully capturing the temperature field distribution and stress redistribution patterns during the closure weld process.

[0046] Example 5: High-speed railway bridge vibration monitoring system In a high-speed railway bridge with a span of 128 meters and a design speed of 350 km / h, 24 wireless acquisition nodes were deployed at the mid-span, supports, and piers of the bridge. The main control gateway was installed in the power distribution box on the top of the pier and connected to the railway maintenance section monitoring center via fiber optic cable.

[0047] Step S1: Network Topology Construction The main control gateway sends a broadcast beacon, and each node measures its RSSI value. Nodes at the top of the piers, with no obstructions between them and the gateway, have RSSI values ​​higher than -70dBm and are classified as Level 1 synchronization nodes; there are 6 of these. Nodes at mid-span and inside the beam structure, due to obstruction by the steel beams, have RSSI values ​​between -80dBm and -90dBm and are classified as Level 2 synchronization nodes; there are 18 of these, and they forward synchronization signals through the nearest Level 1 node. After node registration, the gateway generates and stores a synchronization tree topology.

[0048] Step S2: Physical layer hardware timestamp synchronization In open areas along the railway line with good GNSS signal, the gateway obtains the reference time through the GNSS timing module. Each node completes its initial bidirectional time exchange within 30 seconds: primary nodes exchange directly with the gateway, while secondary nodes relay through primary nodes. When calculating transmission delay, the delay of secondary nodes includes the wireless relay forwarding time, which is calibrated to 50μs ± 5μs through actual measurements. After initial synchronization, the synchronization accuracy of primary nodes is ±300 nanoseconds, and the synchronization accuracy of secondary nodes is ±800 nanoseconds.

[0049] Step S3: Temperature-Sensitive Dynamic Compensation of Crystal Oscillator Frequency The bridge structure is affected by sunlight and seasons, with beam temperature varying from -20℃ to 50℃. Each node underwent temperature-frequency model calibration before deployment, covering temperature points of -20℃, 0℃, 25℃, 40℃, and 60℃. During operation, the node's temperature was collected every second, compensation values ​​were calculated, and timers were adjusted. During winter operation, the crystal oscillator drift reached -20ppm at -15℃, which was controlled within ±1ppm after compensation.

[0050] Step S4: Adaptive Synchronization Period Adjustment When no train is passing, the structural response amplitude is below 0.02g, the synchronization period is set to 90 seconds, the node enters deep sleep mode, and power consumption drops to 0.5mW. When the track circuit detects an approaching train, a trigger signal is transmitted to the gateway via optical fiber. The gateway broadcasts a switching command within 0.3 seconds, and the synchronization period of the entire network switches to 0.2 seconds, meeting the high-frequency vibration sampling requirements at a speed of 350km / h. After the train departs, the system returns to static monitoring mode.

[0051] Step S5: Synchronous Acquisition and Data Feedback All nodes synchronously collect acceleration and dynamic strain data during train passage, with a sampling frequency set to 500Hz. The collected data packets include the theoretical trigger time, actual trigger time, clock offset, and temperature compensation value. Data transmission uses a TDMA mechanism, with all nodes uploading data within 20 seconds of the train's passage. The main control gateway performs timestamp verification on each data packet; if... This is then marked as a synchronization quality warning. The system has been running for 12 months, recording over 1500 train passing events, with 99.2% of the data being of satisfactory synchronization. Modal analysis accurately identified the bridge's dynamic amplification factor at different train speeds.

[0052] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for synchronous data acquisition using a wireless sensor network for structural health monitoring, characterized in that, Includes the following steps: Step S1: Network topology construction. The main control gateway sends a broadcast beacon, and multiple wireless acquisition nodes are automatically divided into primary synchronization nodes and secondary synchronization nodes based on the received signal strength and hop count, forming a tree-like synchronization network. Step S2: Physical layer hardware timestamp synchronization. The main control gateway and the wireless acquisition node perform bidirectional time exchange. The wireless acquisition node uses the physical layer frame start delimiter capture function of the wireless transceiver chip to automatically latch the local crystal oscillator count value as the receiving time at the rising edge of the frame start delimiter of the received synchronization frame, eliminating the interrupt delay between the protocol stack and the processor. It also calculates the transmission delay and clock offset based on the four timestamps of the bidirectional time exchange and adjusts the local virtual clock. Step S3: Temperature-sensing dynamic compensation of crystal oscillator frequency. Each wireless acquisition node monitors the temperature in real time through a built-in temperature sensor, calculates the current frequency error according to the preset temperature-frequency offset model, dynamically adjusts the timer reload value, and corrects the crystal oscillator frequency drift within the two synchronization intervals. Step S4: Adaptive synchronization cycle adjustment. The main control gateway dynamically adjusts the network synchronization cycle according to the working conditions of structural health monitoring: when the structural response amplitude exceeds the preset threshold, it switches to high dynamic mode and shortens the synchronization cycle; when the structural response amplitude is lower than the preset threshold and the temperature change rate is lower than the preset threshold, it switches to static monitoring mode and extends the synchronization cycle. The wireless acquisition node enters sleep mode within the synchronization interval. Step S5: Synchronous data acquisition and data transmission. All wireless acquisition nodes, under a unified virtual clock domain, initiate analog-to-digital conversion according to the global trigger command or collaborative acquisition timetable of the main control gateway. After the acquired data is stamped with a global timestamp, it is transmitted back to the main control gateway through the time division multiple access mechanism.

2. The method according to claim 1, characterized in that, The wireless acquisition node includes a microcontroller, a wireless transceiver chip with physical layer frame start delimiter capture function, a temperature-compensated crystal oscillator, a temperature sensor, an analog-to-digital conversion acquisition unit, and a power management unit; the capture output terminal of the wireless transceiver chip is connected to the capture input pin of the microcontroller. The analog-to-digital conversion acquisition unit is triggered by a microcontroller timer and is linked to the synchronization signal of the wireless transceiver chip.

3. The method according to claim 1, characterized in that, The bidirectional time exchange mentioned in step S2 specifically includes: The main control gateway at all times Send synchronization frames; Wireless acquisition node records reception time ; Wireless data acquisition nodes at any time A response frame is provided, the response frame containing and ; The main control gateway at all times Receive response frame; Calculate transmission delay and clock offset The wireless acquisition node adjusts the timer counter frequency via a software phase-locked loop based on the clock offset.

4. The method according to claim 1, characterized in that, In step S2, adjusting the local virtual clock uses a software phase-locked loop to adjust the timer counter frequency, rather than directly modifying the system time, to avoid time jumps affecting the continuity of data acquisition.

5. The method according to claim 1, characterized in that, The temperature-frequency offset model described in step S3 is obtained as follows: the wireless acquisition node is placed in a temperature-controlled chamber, the actual frequency of the crystal oscillator is measured at different temperature points, and a quadratic function is obtained by fitting the data using the least squares method. ,in The current temperature. For frequency error, , , represents the fitting coefficient.

6. The method according to claim 1, characterized in that, The adaptive synchronization cycle adjustment in step S4 also includes a temperature transient mode: when the wireless acquisition node detects that the temperature change rate exceeds the second preset threshold, the node actively requests the main control gateway to temporarily shorten the synchronization cycle, and restores it after the temperature stabilizes.

7. The method according to claim 1, characterized in that, It also includes anomaly handling and self-healing steps: If a wireless acquisition node fails to complete bidirectional time exchange for a preset number of consecutive times, the node is marked as out of sync. A stray node actively initiates a fast resynchronization process with its neighboring synchronized nodes, using the local time of the neighboring nodes as a reference to recalibrate its clock. If fast resynchronization fails, the out-of-sync node enters sleep-retry mode and attempts to rejoin the network at random backoff intervals. When a secondary synchronization node detects that the link quality with its parent node is lower than the link threshold, it automatically searches for other primary or secondary nodes in the vicinity and reconstructs the synchronization tree.

8. The method according to claim 7, characterized in that, It also includes a health status reporting step: the wireless acquisition node packages the synchronization status, link quality, temperature compensation amount, and crystal oscillator frequency deviation information into a health status frame and periodically sends it back to the main control gateway.

9. The method according to claim 1, characterized in that, It also includes data quality verification steps: Each wireless acquisition node contains the theoretical trigger time, the actual trigger time, the clock offset of the last synchronization time, and the temperature compensation correction value in each data packet; After receiving the data, the main control gateway calculates the deviation between the actual trigger time and the theoretical trigger time. If the deviation exceeds the preset accuracy threshold, the data packet is marked as a synchronization quality warning. If multiple data packets consecutively trigger synchronization quality warnings, the master gateway will trigger a forced resynchronization command for that node.

10. The method according to claim 1, characterized in that, It also includes: the main control gateway has a built-in global navigation satellite system timing module and a temperature-controlled crystal oscillator. When the global navigation satellite system signal is valid, the main control gateway uses the Kalman filter algorithm to tame the temperature-controlled crystal oscillator and form a reference time domain; when the global navigation satellite system signal fails, the main control gateway uses the temperature-controlled crystal oscillator as the reference time source.

Images