Debris flow monitoring equipment for concrete bridge framework area

Abstract

The invention discloses debris flow monitoring equipment for a concrete bridge framework area. Accurate early warning is realized through multi-modal sensing and intelligent decision. The geological vibration sensing module captures debris flow characteristic vibration; the soil humidity module monitors sudden change of water content; the non-contact distance measuring module tracks debris flow movement in real time; and the structural deformation module evaluates bridge safety. The data processing module is used for performing EMD decomposition on the vibration signals to extract energy entropy, and performing Kalman filtering on distance measurement data to predict a track; and fusing multi-source features based on an LSTM / random forest model, and outputting a risk level and confidence. The response module is used for triggering a local sound-light alarm through primary response; the second-level response is used for pushing remote early warning; and the three-level response is linked with the traffic signal and the protective net. And multi-sensor dynamic weight adjustment, data consistency verification, hierarchical response and structural safety linkage are realized. The method is suitable for all-weather debris flow monitoring of mountainous bridges, and significantly improves the early warning timeliness and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of debris flow monitoring, and in particular to a debris flow monitoring device in a concrete bridge structure area. Background Art

[0002] Given the frequent occurrence of concrete bridge disasters in mountainous areas, strengthening structural safety risk prevention and control measures can begin with regular structural inspections and improved structural safety monitoring systems. Traditional debris flow monitoring technology has significant flaws in safety monitoring: geological radar or cameras cannot distinguish between debris flows and interference from landslides, vehicle vibration, and other factors, resulting in a false alarm rate as high as 30%-40%. Relying on manual inspections or offline data analysis makes it difficult to provide timely warnings of high-speed debris flows (speeds > 10 m / s). Optical equipment is affected by rain, fog, and low illumination at night, increasing ranging errors. Vibration sensors are susceptible to interference from wind, rain, and mechanical vibration.

[0003] Therefore, there is a lack of a debris flow monitoring device in the concrete bridge structure area to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a debris flow monitoring device for a concrete bridge structure area.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A debris flow monitoring device for a concrete bridge structure area, comprising:

[0007] a geological vibration sensing module configured to detect geological vibration signals of the bridge foundation and surrounding areas;

[0008] A soil moisture sensing module configured to monitor dynamic changes in soil moisture content;

[0009] A non-contact distance measurement module configured to measure the distance between the debris flow front and the bridge and the speed of movement in real time;

[0010] A data processing module is connected to the geological vibration sensing module, the soil moisture sensing module and the non-contact ranging module, and is used to fuse the vibration signal, the humidity change data and the ranging data, and output the debris flow risk level through a preset risk assessment model;

[0011] The early warning module triggers graded response actions according to the risk level.

[0012] As a further optimization of the present technical solution, the non-contact ranging module includes a first ranging module and a second ranging module, which respectively implement ranging based on different physical principles; the data processing module is configured to dynamically adjust the data weights of the first ranging module and the second ranging module according to the intensity of environmental interference.

[0013] As a further optimization of this technical solution, the data processing module performs modal decomposition on the geological vibration signal and extracts the energy distribution characteristics of the preset frequency band; the risk assessment model is a classification model based on machine learning, and the input parameters include at least vibration energy entropy, humidity change rate and debris flow front speed.

[0014] As a further optimization of the present technical solution, the hierarchical response actions include: first response level: when the soil moisture change rate exceeds the first threshold and the vibration energy exceeds the second threshold, a local alarm is triggered; second response level: when the distance between the debris flow fronts is less than the third threshold and the movement speed exceeds the fourth threshold, an early warning information is sent to the remote management platform; third response level: when the deformation parameter of the bridge structure exceeds the fifth threshold, the traffic signal control system is linked.

[0015] As a further optimization of this technical solution, the first ranging module is a millimeter wave radar, and the second ranging module is a lidar or an ultrasonic ranging device; the environmental interference intensity includes at least one of rain and fog concentration, visibility or electromagnetic noise intensity.

[0016] As a further optimization of this technical solution, it also includes a structural deformation perception module configured to monitor the inclination or displacement of the bridge; the data processing module uses the inclination or displacement as an input parameter of the risk assessment model.

[0017] Beneficial effects: The present invention provides a debris flow monitoring device for the concrete bridge structure area. This device achieves the following core advantages through technological innovation: 1. High-precision monitoring and early warning: Through multimodal data fusion (vibration, humidity, ranging, deformation), it accurately identifies the characteristic signals of debris flows and achieves early warning 15-30 minutes in advance, far exceeding the 5-10 minute response capability of traditional single-sensor solutions. 2. The dynamic weight adjustment mechanism (such as the millimeter wave radar weight of 0.7 in rainy and foggy environments) ensures that the ranging accuracy remains at ±0.5m in severe weather, and the false alarm rate is reduced to 4.7% (rainstorm scene). 3. Strong environmental adaptability: millimeter wave radar and lidar complement each other, with a detection rate of >90% in rain and fog / nighttime. Reliability is improved through multi-sensor fusion in vegetation obstruction scenes (millimeter wave weight of 0.9 when vegetation coverage is >70%). The edge-cloud collaborative architecture (90% of data is processed at the edge) ensures low latency (average response time <2 seconds), and the cloud model is continuously optimized. 4. Hierarchical response and linkage control: The three-level response mechanism (local alarm → remote warning → traffic control) covers different risk scenarios, linking the protection net (reaction time < 5 seconds) and the signal light system to improve emergency response efficiency. Data consistency verification (vibration + humidity + ranging double threshold) reduces the false alarm rate by 85.3% (tested in a rainstorm scenario). 5. Flexible deployment and cost optimization: Supports low-cost solutions (ultrasonic ranging + MobileNetV2 model), reducing hardware costs by 60%, suitable for small and medium-sized bridges with limited budgets. The modular design is compatible with multiple sensor types (such as MEMS accelerometers and fiber Bragg grating displacement meters) and adapts to different terrain requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 : System architecture block diagram of the debris flow monitoring equipment in the concrete bridge structure area of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] Please refer to the instruction manual Figure 1 The present invention provides a technical solution: a debris flow monitoring device for a concrete bridge structure area, comprising:

[0021] a geological vibration sensing module configured to detect geological vibration signals of the bridge foundation and surrounding areas;

[0022] A soil moisture sensing module configured to monitor dynamic changes in soil moisture content;

[0023] A non-contact distance measurement module configured to measure the distance between the debris flow front and the bridge and the speed of movement in real time;

[0024] A data processing module is connected to the geological vibration sensing module, the soil moisture sensing module and the non-contact ranging module, and is used to fuse the vibration signal, the humidity change data and the ranging data, and output the debris flow risk level through a preset risk assessment model;

[0025] The early warning module triggers graded response actions according to the risk level.

[0026] More specifically, the hardware configuration of the basic debris flow monitoring equipment is:

[0027] Geological vibration sensing module: Two GeoSIG GBV-358 broadband seismic geophones installed at the bridge foundation have a frequency response range of 0.1-100 Hz and detect the low-frequency vibration characteristics of debris flows (0.5-5 Hz).

[0028] Soil moisture sensing module: A Decagon EC-5 capacitive moisture sensor array (10m spacing, 0.5m depth) is embedded in the soil 50m upstream of the bridge to monitor moisture content changes in real time.

[0029] Non-contact ranging module: Using the TI AWR2243 millimeter-wave radar (77GHz, detection range 600m), deployed on top of the bridge, it measures the distance to the debris flow front in real time.

[0030] Data processing module: NVIDIA Jetson AGX Xavier edge computing unit, running a random forest model. Input parameters include: energy entropy of the vibration signal after EMD decomposition (0.5-5Hz frequency range);

[0031] Soil moisture change rate (% / minute); debris flow front speed measured by millimeter-wave radar (m / s). Warning logic: First response level: When the soil moisture change rate is greater than 5% / minute and the vibration energy is greater than 30J, the bridgehead sound and light alarm (120dB buzzer + red warning light) will be activated.

[0032] In a specific implementation, the non-contact ranging module includes a first ranging module and a second ranging module, which respectively implement ranging based on different physical principles; the data processing module is configured to dynamically adjust the data weights of the first ranging module and the second ranging module according to the intensity of environmental interference.

[0033] More specifically, the enhanced ranging and dynamic weighting adjustments are as follows:

[0034] The first ranging module: Velodyne VLP-16 lidar (detection distance 300m, accuracy ±3cm), used for high-precision ranging; the second ranging module: TI AWR2243 millimeter wave radar is retained as a backup for rainy and foggy environments.

[0035] Data processing optimization: Environmental interference intensity is assessed using the camera visibility analysis module. When visibility exceeds 100 meters, lidar data is weighted 0.8 and millimeter-wave radar data is weighted 0.2. When visibility is ≤50 meters, the weighting switches to 0.7 for millimeter-wave radar and 0.3 for lidar. Warning logic: Level 2 response: When the distance between debris flow fronts is less than 300 meters and the speed is greater than 8 meters / second, an early warning text message containing the location coordinates is sent to the bridge management center via the 4G module.

[0036] Signal preprocessing unit:

[0037] Vibration signal → EMD decomposition → energy entropy extraction;

[0038] Humidity data → gradient calculation → mutation rate analysis;

[0039] Odometry data → Kalman filter → trajectory prediction.

[0040] During specific implementation, the data processing module performs modal decomposition on the geological vibration signal and extracts the energy distribution characteristics of the preset frequency band; the risk assessment model is a classification model based on machine learning, and the input parameters include at least vibration energy entropy, humidity change rate and debris flow front speed.

[0041] In specific implementation, the graded response actions include: first response level: when the soil moisture change rate exceeds the first threshold and the vibration energy exceeds the second threshold, a local alarm is triggered; second response level: when the distance between the debris flow fronts is less than the third threshold and the movement speed exceeds the fourth threshold, an early warning information is sent to the remote management platform; third response level: when the deformation parameter of the bridge structure exceeds the fifth threshold, the traffic signal control system is linked.

[0042] In a specific implementation, the first ranging module is a millimeter wave radar, and the second ranging module is a laser radar or an ultrasonic ranging device; the environmental interference intensity includes at least one of rain and fog concentration, visibility or electromagnetic noise intensity.

[0043] More specifically, the low-cost ultrasonic ranging solution is as follows:

[0044] The first ranging module uses a MaxBotix MB7366 ultrasonic sensor (with a detection range of 150m and a 60% cost reduction). The second ranging module omits the lidar and retains only the millimeter-wave radar. Data processing is simplified by using a lightweight MobileNetV2 model (less than 5MB) running on a Raspberry Pi 4B. Vibration signal analysis uses a fast Fourier transform (FFT) instead of an EMD to reduce computational overhead. Warning logic: Level 3 response: When the inclinometer detects a bridge pier tilt greater than 0.8°, the roadside LED screen displays "Danger Bridge, No Passing."

[0045] More specifically, the valley bridge monitoring with multi-sensor fusion includes the following:

[0046] Geological vibration sensing module: Three ADXL357 MEMS accelerometers (±40g range) are added to the slopes on both sides of the bridge to detect vibrations that could indicate the onset of mountain slip.

[0047] Non-contact ranging module: The first ranging module is the Aeva 4D LiDAR (FMCW laser radar), which measures both distance (500m) and velocity (accuracy ±0.1m / s). The second ranging module is the TI IWR6843 millimeter-wave radar (60GHz), which has enhanced vegetation penetration capabilities. Dynamic weighting strategy: When vegetation coverage exceeds 70%, the millimeter-wave radar weight is increased to 0.9. In heavy rain (rainfall >50mm / h), the LiDAR weight is reduced to 0.4, and the millimeter-wave radar weight is 0.6.

[0048] During specific implementation, it also includes a structural deformation perception module configured to monitor the inclination or displacement of the bridge; the data processing module uses the inclination or displacement as an input parameter of the risk assessment model.

[0049] More specifically, the contents of the coordinated monitoring of bridge structure deformation are as follows:

[0050] Structural deformation sensing module: Inclinometer: Murata SCL3300 (range ±30°, accuracy 0.01°), installed on the top of the bridge pier; displacement sensor: Kikon BGK-FBG-4000 fiber Bragg grating sensor, monitoring the lateral displacement of the bridge deck (accuracy ±1mm). New parameters added to the risk assessment model input: pier inclination change rate (° / hour); bridge deck displacement acceleration (mm / s 2 Warning logic: Third response level: When the pier inclination angle is greater than 1° or the bridge deck displacement speed is greater than 5mm / s, the emergency braking system is triggered and the bridge access gate is closed.

[0051] More specifically, the all-weather emergency response system includes the following: Traffic signal control system: Linked to the traffic lights at both ends of the bridge via the Modbus protocol, the lights are forced to switch to red when a Level 3 warning is reached; Protective device: A hydraulically driven rockfall protection net (response time <5 seconds) is deployed 200 meters upstream of the bridge; Data verification mechanism: Multi-source data consistency verification: A high-level alarm is triggered only when at least two of the vibration, humidity, and ranging data exceed thresholds simultaneously, reducing the false alarm rate to <5%.

[0052] The data verification mechanism is as follows:

[0053] Multi-source data consistency verification logic:

[0054] Trigger condition: When at least two types of data among vibration, humidity, and ranging data exceed the threshold at the same time and the time synchronization error is less than 5 seconds, it is determined to be a valid debris flow event.

[0055] Verification rules:

[0056] Rule 1 (vibration-humidity linkage verification): If the vibration energy is greater than 30 J and the soil moisture change rate is greater than 5% / minute, but the ranging module does not detect debris flow movement, the manual verification mode is activated (calling the camera to capture the scene).

[0057] Rule 2 (Ranging-Deformation Cross-Validation): When the ranging module detects that the debris flow front is less than 200 m from the bridge, but the bridge inclination changes by less than 0.3°, it is considered a false detection (possibly due to the movement of a herd of animals) and only a log is recorded without triggering an alarm.

[0058] Sensor health monitoring: Heartbeat packet detection: Send status query commands to each sensor every 10 minutes. If there is no response for three consecutive times, the sensor is marked as faulty and its weight is reduced to 0.1 during data fusion.

[0059] Data rationality check: When the vibration signal amplitude exceeds the measuring range (such as the ADXL357 range ±40g), the self-calibration procedure is started; when the humidity sensor value jumps by more than 50% within 1 minute, the soil sampler retest is triggered (optional accessory).

[0060] Test scenario Verification not enabled Enable Verification False alarm rate decreased Heavy rain (50 mm / h) 32％ 4.7％ 85.3％ Vehicles passing through the bridge (5 tons) 28％ 3.1％ 88.9％ Mountain blasting construction (1km away) 41％ 6.2％ 84.9％

[0061] Table 1 (false alarm rate suppression test)

[0062] The implementation details are as follows:

[0063] Add a soil sampler (optional): collect soil samples when humidity changes suddenly, and analyze mud content through image recognition; backup communication link: when the main 4G network is interrupted, switch to LoRa wireless transmission to verify instructions.

[0064] Time series alignment algorithm: Dynamic time warping (DTW) is used to compensate for the clock deviation of each sensor; abnormal data filter: outliers are identified based on the isolation forest model.

[0065] Scenario: Heavy rain causes the humidity sensor to report a false alarm (humidity suddenly rises from 30% to 80%), but the vibration and ranging data are normal.

[0066] Verification process:

[0067] Step 1: Humidity data triggers a preliminary warning and starts Rule 1 verification;

[0068] Step 2: The vibration energy detected is 15J (<30J threshold), and the ranging module has no debris flow signal;

[0069] Step 3: Use the camera to take pictures. Image analysis shows no debris flow traces.

[0070] Step 4: After the confidence level is calculated, it is determined to be a false alarm and only a log is recorded.

[0071] Working principle: This device realizes debris flow monitoring through a closed-loop process of multimodal perception-dynamic fusion-intelligent decision-making-graded response. The specific working principle is as follows:

[0072] Vibration sensors (such as seismometers and MEMS accelerometers) deployed in the bridge foundation and surrounding mountains collect low-frequency vibration signals (0.1-100Hz) in real time. The characteristic vibrations (0.5-5Hz) generated by rock collisions and fluid impacts during debris flows are captured, amplified, filtered, and transmitted to the data processing module. Technical Benefit: Unlike ordinary landslides, debris flows have higher vibration energy and a more concentrated frequency band, providing a basis for early warning.

[0073] An array of capacitive humidity sensors embedded in the soil monitors moisture content. When rainfall or groundwater infiltration causes a sudden increase in soil moisture (e.g., a >30% increase within an hour), a debris flow risk assessment is triggered. The technical benefit: Humidity data is linked to vibration signals to distinguish between common geological activity and debris flow precursors.

[0074] It emits 77GHz electromagnetic waves and calculates the real-time distance and speed of the debris flow front (accuracy ±0.5m) through the echo time difference and Doppler effect. LiDAR emits laser pulses and measures distance through time of flight (ToF), complementing millimeter-wave radar data in rainy and foggy environments. Dynamic weight adjustment automatically switches the main ranging module according to environmental visibility and rain and fog concentration (for example, in heavy rain, the millimeter wave weight is 0.7 and the lidar weight is 0.3).

[0075] Vibration signals are analyzed using empirical mode decomposition (EMD) to extract energy entropy in the 0.5-5 Hz frequency band. Humidity data is used to calculate the gradient rate of change (Δhumidity / Δtime). Distance measurement data is then filtered using a Kalman filter to predict debris flow trajectories. Machine learning model decision-making uses an LSTM or random forest model. Input features include: vibration energy entropy (to quantify debris flow intensity); humidity gradient (to reflect soil saturation); debris flow front velocity and acceleration (to determine impact risk); and bridge inclination or displacement (to assess structural safety). Model outputs include a risk level (low / medium / high) and a confidence score (0-1).

[0076] An event is considered valid only when at least two types of data, including vibration, humidity, and ranging, exceed the threshold and are synchronized in time (error < 5 seconds). Confidence calculation: Dynamically adjust the alarm threshold based on historical data and real-time deviations to suppress false alarms (such as false triggering of humidity caused by heavy rain).

[0077] Level 1 Response (Potential Risk): Trigger conditions: Soil moisture change rate > 5% / min and vibration energy > 30J; Action: Activate the bridgehead's audible and visual alarm (120dB buzzer + red LED flashing) to alert nearby personnel to take shelter. Level 2 Response (Imminent Risk): Trigger conditions: The debris flow front is < 500m from the bridge and its speed is > 5m / s; Action: Send an early warning message (including GPS coordinates and estimated arrival time) to the bridge management center and launch a drone inspection to confirm the scene. Level 3 Response (Emergency Risk Evasion): Trigger conditions: Bridge inclination > 0.8° or structural displacement speed > 5mm / s; Action: Interact with the traffic signal system: Force the lights at both ends of the bridge to red, closing them to traffic; Activate protective devices: such as hydraulically driven rockfall protection nets and bridge deck isolation gates; Send emergency instructions: Notify fire, medical, and other rescue units to their locations.

[0078] Sensor health monitoring: Heartbeat packets are sent every 10 minutes to monitor sensor status. Faulty sensor data is automatically downgraded (weight ≤ 0.1) to ensure continuous system operation. Dynamic threshold adjustment: Alarm conditions are optimized based on meteorological data (e.g., when rainfall exceeds 50 mm / h, the humidity threshold is lowered to 30%) and historical statistical models. Edge-cloud collaboration: The edge (such as NVIDIA Jetson) processes 90% of data in real time to ensure low-latency response; the cloud performs big data analysis, updates machine learning model parameters, and synchronizes them to the terminal.

[0079] Working principle flow chart:

[0080] Vibration signal → EMD decomposition → feature extraction;

[0081] Humidity data → gradient calculation → feature extraction → data fusion → machine learning model → risk level → graded response;

[0082] Odometry data → Kalman filter → trajectory prediction.

[0083] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A debris flow monitoring device for a concrete bridge structure area, characterized in that: include: a geological vibration sensing module configured to detect geological vibration signals of the bridge foundation and surrounding areas; A soil moisture sensing module configured to monitor dynamic changes in soil moisture content; A non-contact distance measurement module configured to measure the distance between the debris flow front and the bridge and the speed of movement in real time; A data processing module is connected to the geological vibration sensing module, the soil moisture sensing module and the non-contact ranging module, and is used to fuse the vibration signal, the humidity change data and the ranging data, and output the debris flow risk level through a preset risk assessment model; The early warning module triggers graded response actions according to the risk level.

2. The debris flow monitoring device for concrete bridge structure area according to claim 1, characterized in that: The non-contact ranging module includes a first ranging module and a second ranging module, each of which implements ranging based on different physical principles; the data processing module is configured to dynamically adjust the data weights of the first ranging module and the second ranging module according to the intensity of environmental interference.

3. The debris flow monitoring device for concrete bridge structure area according to claim 2, characterized in that: The data processing module performs modal decomposition on the geological vibration signal and extracts the energy distribution characteristics of the preset frequency band; the risk assessment model is a classification model based on machine learning, and the input parameters include at least vibration energy entropy, humidity change rate and debris flow front speed.

4. The debris flow monitoring device for concrete bridge structure area according to claim 3, characterized in that: The hierarchical response actions include: first response level: when the soil moisture change rate exceeds the first threshold and the vibration energy exceeds the second threshold, a local alarm is triggered; second response level: when the distance between the debris flow fronts is less than the third threshold and the movement speed exceeds the fourth threshold, an early warning information is sent to the remote management platform; third response level: when the bridge structure deformation parameter exceeds the fifth threshold, the traffic signal control system is linked.

5. The debris flow monitoring device for concrete bridge structure area according to claim 4, characterized in that: The first ranging module is a millimeter wave radar, and the second ranging module is a laser radar or an ultrasonic ranging device; the environmental interference intensity includes at least one of rain and fog concentration, visibility or electromagnetic noise intensity.

6. The debris flow monitoring device for concrete bridge structure area according to claim 5, characterized in that: It also includes a structural deformation perception module configured to monitor the inclination or displacement of the bridge; the data processing module uses the inclination or displacement as an input parameter of the risk assessment model.

Images