Inclinometer and inspection robot collaborative slope monitoring method

Abstract

The invention discloses an inclinometer and inspection robot collaborative slope monitoring method, which comprises the following steps: monitoring is carried out through an inclinometer and an inspection robot respectively, the inclinometer continuously collects local angle data, and the inspection robot executes scanning and generates three-dimensional point cloud data; performing space-time alignment and data fusion on the local angle data and the three-dimensional point cloud data, and establishing a fusion stability index of the structural body; a bidirectional triggering mechanism between the inclinometer and the inspection robot is established, when the inclinometer monitors that the angle change is abnormal, the inspection robot is triggered to carry out high-density reinspection, and when the inspection robot recognizes a macroscopic deformation hidden danger, an adjacent inclinometer is triggered to carry out refined monitoring; and performing dynamic evaluation and early warning on the health state of the structural body based on feedback data of the fusion stability index and the bidirectional trigger mechanism. The comprehensive performance, the real-time performance and the reliability of the monitoring system are remarkably improved, and an intelligent solution is provided for safe operation and maintenance of major infrastructures.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of slope safety monitoring, and particularly relates to a slope monitoring method based on cooperation of an inclinometer and a patrol robot. BACKGROUND

[0002] In the long-term operation and maintenance of large civil engineering structures such as slopes, tailings ponds, dams and tunnels, structural stability monitoring is a key link for preventing geological disasters and ensuring the safety of people's lives and property. Traditional monitoring methods mainly rely on manual inspection or laying static sensor networks. Manual inspection is not only inefficient and costly, but also poses a serious safety risk to personnel in harsh or dangerous environments, making it difficult to achieve high-frequency, full-coverage and normalized monitoring. Static sensors such as inclinometers and displacement meters can achieve local continuous monitoring, but their sensing range is limited and can only reflect small changes near the installation point. It is difficult to capture large-scale macro deformation such as overall slip and large-area crack propagation, and there is a "point-to-area" monitoring blind spot.

[0003] In recent years, patrol robots such as unmanned aerial vehicles and tracked robots equipped with sensors such as laser radars and depth cameras have been introduced into the field of structural health monitoring. Such mobile devices can autonomously patrol, generate high-precision three-dimensional point cloud maps, and intuitively present the macro deformation of the structure, effectively compensating for the limited field of view of static sensors. However, their monitoring mode is usually periodic scanning, such as once a week, with obvious "time blind spots". Local rapid deformation or small angle mutations occurring during this period are easily missed, and immediate response to sudden risks cannot be achieved. In addition, existing systems mostly operate static sensors and patrol robots as independent systems, and the data collected by the two are isolated, lacking effective fusion analysis and coordination mechanisms. Static data and dynamic data cannot be cross-validated, resulting in high false alarm rates and difficulty in quantifying the correlation between local micro changes and macro instability.

[0004] In summary, the existing technology has several obvious shortcomings: fixed sensors laid can only monitor small changes near the installation point, but cannot grasp the overall deformation trend; mobile patrol robots can scan the overall appearance of a large area, but are prone to missing local and sudden small changes. Moreover, these two systems usually operate independently, and the data is not shared or fused, leading to slow analysis and judgment and delayed warning response. Therefore, there is an urgent need for a collaborative monitoring method that can achieve "dynamic and static combination, complementary advantages, and intelligent linkage". SUMMARY

[0005] To solve the above technical problems, the present application provides a slope monitoring method based on cooperation of an inclinometer and a patrol robot, which realizes comprehensive, real-time and accurate monitoring of the health status of the structure by constructing a collaborative mechanism of "static sensing-dynamic verification-bidirectional triggering-fusion analysis".

[0006] The application provides a slope monitoring method based on a tiltmeter and a patrol robot.

[0007] The tiltmeter and the patrol robot are used for monitoring respectively, the tiltmeter continuously collects local angle data at a first preset frequency, and the patrol robot performs scanning and generates three-dimensional point cloud data according to a preset period;

[0008] The local angle data and the three-dimensional point cloud data are spatiotemporally aligned and fused to establish a fusion stability index of the structure;

[0009] A bidirectional triggering mechanism is established between the tiltmeter and the patrol robot, the patrol robot is triggered to perform high-density review when the tiltmeter detects abnormal angle change, and the adjacent tiltmeter is triggered to perform fine monitoring when the patrol robot identifies a macroscopic deformation hidden danger;

[0010] Based on the feedback data of the fusion stability index and the bidirectional triggering mechanism, the health state of the structure is dynamically evaluated and warned.

[0011] Optionally, the monitoring by the tiltmeter and the patrol robot respectively comprises the following steps.

[0012] A plurality of tiltmeters with wireless communication function are arranged in key risk areas of the structure;

[0013] The patrol robot is arranged at a preset landing point or track;

[0014] The tiltmeter and the patrol robot are connected to a unified monitoring platform.

[0015] Optionally, the spatiotemporal alignment and data fusion of the local angle data and the three-dimensional point cloud data to establish the fusion stability index of the structure comprises the following steps.

[0016] The local angle data and the three-dimensional point cloud data are time-interpolated and coordinate-system-converted to obtain a multi-source fusion data set with time synchronization and coordinate unification;

[0017] Based on the multi-source fusion data set, a fusion stability index of the structure is established by weighted fusion of a local stability index, a macroscopic stability index and a spatiotemporal correlation coefficient.

[0018] Optionally, the time interpolation and coordinate-system conversion of the local angle data and the three-dimensional point cloud data to obtain a multi-source fusion data set with time synchronization and coordinate unification comprises the following steps.

[0019] For the time series data collected by the tiltmeter, linear or spline interpolation method is used to align to the scanning time of the inspection robot; for the three-dimensional point cloud data, the angle change in the local coordinate system is converted to the global point cloud coordinate system through the known installation pose of the tiltmeter, and a unified multi-source fusion dataset is constructed.

[0020] Optionally, the angle change obtained from the local angle data measured by the tiltmeter is introduced into the deformation analysis algorithm of the point cloud map as a constraint condition during data fusion.

[0021] Optionally, the local angle mutation monitored by the tiltmeter and the macro deformation spatial position recognized by the inspection robot are compared for cross verification, and when they are highly correlated in space and time, it is determined that there is a real hidden danger.

[0022] Optionally, the spatiotemporal correlation coefficient is obtained by calculating the Pearson correlation coefficient of the local change rate of the tiltmeter and the point cloud deformation gradient of the adjacent area, or is obtained by a spatial distance decay model.

[0023] Optionally, in the bidirectional triggering mechanism, when the tiltmeter monitors angle change abnormalities, the inspection robot is triggered to perform high-density review, specifically:

[0024] The angle change rate is calculated according to the angle measurement values of the tiltmeter at adjacent sampling times;

[0025] When the angle change rate exceeds a first preset threshold, a review instruction is sent to the inspection robot;

[0026] The inspection robot responds to the review instruction, interrupts the current task, and navigates to the area where the tiltmeter sending the instruction is located with priority, and performs a review task with higher scanning density than normal.

[0027] Optionally, in the bidirectional triggering mechanism, when the inspection robot recognizes a macro deformation hidden danger, the adjacent tiltmeter is triggered to perform fine monitoring, specifically:

[0028] A macro deformation hidden danger is identified according to the three-dimensional point cloud data generated by the inspection robot;

[0029] The adjacent area range is dynamically determined according to the area or volume of the identified macro deformation hidden danger;

[0030] A monitoring instruction is sent to the tiltmeter located in the adjacent area range;

[0031] The tiltmeter receiving the instruction raises the sampling frequency from the first preset frequency to the second preset frequency, performs local fine monitoring, and continues until the angle change rate monitored by the related tiltmeter falls below the second preset threshold.

[0032] Optionally, based on the fusion stability index and the feedback data of the bidirectional triggering mechanism, the structural health state is dynamically evaluated and early warned, specifically including:

[0033] When the fusion stability index is lower than a safety threshold, or both the local stability score and the macro stability score show a continuous downward trend, a warning information is generated;

[0034] The change rate of the reciprocal of the fusion stability index over time is calculated as a comprehensive risk growth rate;

[0035] When the comprehensive risk growth rate exceeds a third preset threshold, a high-level warning information is generated.

[0036] Compared with the prior art, the present application has the following advantages and technical effects:

[0037] Based on the spatiotemporal alignment and data fusion technical means, the time series angle data collected by the inclinometer and the spatial point cloud data obtained by the inspection robot are time-interpolated, aligned and coordinated, and on this basis, geometric feature fusion and state complementary verification are performed, solving the problem that local data and macro data are isolated from each other and cannot be analyzed in conjunction in traditional monitoring, and improving the data utilization rate and the hazard identification accuracy.

[0038] Based on the bidirectional triggering and cooperative response technical means, a bidirectional linkage mechanism of inclinometer abnormal triggering robot rechecking and robot identified hazard triggering inclinometer intensive monitoring is established, overcoming the inherent defects of limited spatial coverage of static sensors and large time sampling interval of mobile robots, which can expand the monitoring spatiotemporal coverage range, realize real-time response to abnormalities and fine monitoring of key areas.

[0039] Based on the evaluation model technical means of multi-source information fusion, a quantitative evaluation index integrating local stability, macro stability and spatiotemporal correlation indicators is constructed, and a comprehensive risk growth rate is introduced for trend judgment, changing the previous evaluation mode relying on a single data source and subjective experience judgment, which can dynamically, quantitatively and prospectively evaluate and grade the risk of the structural health state. BRIEF DESCRIPTION OF DRAWINGS

[0040] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The embodiments of this application and its description together with the drawings make an aid in understanding the application. In the drawings:

[0041] Fig. 1 The method flowchart of the embodiment of the present application;

[0042] Fig. 2 The data fusion flowchart of the embodiment of the present application. DETAILED DESCRIPTION

[0043] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0044] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in a different order.

[0045] Embodiment one

[0046] As Figs. 1-2 shown, the present embodiment provides a tilt angle instrument and a coordinated monitoring method of a patrol robot, comprising the following steps:

[0047] The tilt angle instrument and the movable patrol robot are monitored respectively by being deployed on the structure, wherein the tilt angle instrument continuously collects local angle data at a first preset frequency, and the patrol robot performs scanning according to a preset period and generates three-dimensional point cloud data;

[0048] The local angle data and the three-dimensional point cloud data are spatiotemporally aligned and data fused to establish a fusion stability index of the structure;

[0049] A bidirectional triggering mechanism is established between the tilt angle instrument and the patrol robot, wherein the patrol robot is triggered to perform high-density review when the tilt angle instrument monitors angle change abnormalities, and the adjacent tilt angle instrument is triggered to perform fine monitoring when the patrol robot identifies macroscopic deformation hazards;

[0050] Based on the feedback data of the fusion stability index and the bidirectional triggering mechanism, the health status of the structure is dynamically evaluated and warned.

[0051] As a feasible implementation manner, the method proposed in the present embodiment specifically comprises the following steps:

[0052] Step 1, deploy multiple tilt angle instruments with wireless communication function in the key risk areas of the structure, and at the same time, configure the patrol robot at the fixed take-off and landing point or track, both of which are connected to a unified monitoring platform;

[0053] Step 2, the tilt angle instrument continuously collects local angle change data at a preset frequency, and the patrol robot performs global scanning task according to a preset period to generate a three-dimensional point cloud map;

[0054] Step 3, the monitoring platform spatiotemporally aligns and features fuses the tilt angle instrument data and the patrol robot point cloud data to construct a unified fusion stability index of the structure;

[0055] Step 4, establish a two-way triggering mechanism: when the inclination instrument monitors the data rate of change exceeds the first preset threshold, immediately send a high frequency review instruction to the inspection robot; when the inspection robot identifies macroscopic deformation or new hidden danger, send a monitoring instruction to the inclination instrument in the adjacent area;

[0056] Step 5, based on the fusion data and the two-way triggering mechanism, realize the dynamic evaluation and early warning of the health status of the structure.

[0057] It can be implemented that the process of step 3, in which the monitoring platform aligns the inclination instrument data and the inspection robot point cloud data in time and space and fuses the features, includes:

[0058] Time interpolation and coordinate system conversion are performed on the local angle data and the three-dimensional point cloud data to obtain a time-synchronized and coordinate-unified multi-source fusion data set; based on the multi-source fusion data set, the local stability index, the macro stability index and the time-space correlation coefficient are fused by weighting to establish the fusion stability index of the structure.

[0059] Further, the time-space alignment in step 3 is achieved by time interpolation algorithm and space coordinate conversion: for the time series data collected by the inclination instrument, linear or spline interpolation method is used to align to the scanning time of the inspection robot; for the point cloud data, the angle change in the local coordinate system is converted to the global point cloud coordinate system by the known installation pose of the inclination instrument, and a unified time-space reference frame is constructed.

[0060] Further, the feature fusion in step 3 includes geometric feature fusion and state complementary verification: the geometric feature fusion takes the angle change measured by the inclination instrument as a constraint condition and introduces it into the deformation analysis algorithm of the point cloud map to optimize the calculation accuracy of geometric parameters such as crack propagation direction and dangerous rock body inclination angle; the state complementary verification compares the local angle mutation monitored by the inclination instrument with the macroscopic deformation spatial position identified by the inspection robot for cross verification, and when they are highly correlated in time and space, it is determined as a real hidden danger, reducing the false alarm rate.

[0061] Further, the overall stability index of the structure after fusion can be expressed as:

[0062] ;

[0063] In the formula: is the local stability score calculated based on the inclination instrument data, is the macro stability score based on the point cloud deformation analysis, is the time-space correlation coefficient of local and macro data, is the adjustable fusion weight, and =1.

[0064] Furthermore, the spatiotemporal correlation coefficient It can be obtained by calculating the Pearson correlation coefficient between the local rate of change of the inclinometer and the deformation gradient of the point cloud in the neighboring region, or by calculating it through the spatial distance attenuation model:

[0065] ;

[0066] In the formula: A is the set of inclinometers, and D is the set of point cloud deformation regions. Let be the Euclidean distance between the i-th inclinometer and the center of the j-th deformation region. For distance attenuation scale parameters, Rate of change of angle With point cloud deformation gradient The correlation, This is a weighting factor used to distinguish the correlation strength of different types of potential hazards.

[0067] The feasible bidirectional triggering mechanism in step 4 is specifically as follows:

[0068] Static-to-Dynamic Trigger: When the rate of change of any inclinometer exceeds a first preset threshold for multiple consecutive time periods, or the cumulative change reaches a warning value, the monitoring platform immediately sends an emergency re-inspection command to the inspection robot. The command includes the ID and location information of the target inclinometer. The inspection robot interrupts its original plan and prioritizes a high-density scan of the area to verify whether local anomalies have triggered changes in the macroscopic structure. The rate of change of angle is... Defined as:

[0069] ;

[0070] In the formula: Let be the measurement value of the i-th inclinometer at time t. The sampling interval is denoted as .

[0071] Dynamic-to-static triggering: When the inspection robot identifies macroscopic hazards such as new cracks, displacement of unstable rock masses, or fallen trees in point cloud analysis, the monitoring platform sends monitoring instructions to all inclinometers within a preset range around the hazard. The instructions include the coordinates and frequency parameters of the target area. After receiving the instructions, the inclinometers automatically increase the sampling frequency and continuously capture minute angle changes in the area to achieve localized and refined monitoring.

[0072] Furthermore, the "preset surrounding range" is not a fixed value, but a trigger radius dynamically determined based on the type and severity of the hazard. :

[0073] ;

[0074] In the formula: As the reference trigger radius, The area or volume of the identified macroscopic hazards. This is the area-radius conversion factor, used to reflect the characteristic that large-scale hidden dangers have a wider impact range.

[0075] Furthermore, the execution duration of the monitoring command in step 4 is a preset time window. It can continue until the inclinometer data change rate falls below the second preset threshold, after which it will automatically resume the normal monitoring frequency, balancing data accuracy and device power consumption.

[0076] In practice, the dynamic evaluation and early warning in step 5 are achieved based on the fused structural stability index: by combining the local dynamic response data of the inclinometer with the macroscopic structural deformation data of the inspection robot, the overall stability index of the structure is calculated. ;when Below the safety threshold ,or and When all values ​​show a continuous downward trend, the system generates a tiered early warning message.

[0077] Furthermore, to quantify the evolution trend of potential risks, a comprehensive risk growth rate is defined. for:

[0078] ;

[0079] when When the preset threshold is exceeded, a high-level warning is triggered.

[0080] In practice, the inclinometer features a low-power design and a wireless communication module, supporting remote configuration of the working mode; the inspection robot integrates a lidar, a depth camera, and an inertial navigation system, possessing autonomous navigation and 3D reconstruction capabilities; both communicate with the monitoring platform via NB-IoT, LoRa, or Mesh networks to ensure communication reliability in complex terrain.

[0081] The method is feasible and applicable to civil engineering structures that require long-term health monitoring, such as slopes, tailings ponds, dams, tunnels, and culverts.

[0082] Example 2

[0083] The specific procedure for implementing the method of this embodiment in a certain high and steep slope health monitoring project is as follows:

[0084] Step 1: System Deployment and Initialization. Multiple low-power inclinometers with NB-IoT communication capabilities are deployed in key risk areas such as potential slip surfaces and crack development zones on the slope. Each inclinometer records its precise installation pose (position and orientation in the global coordinate system). Simultaneously, a landing platform for an inspection robot is set up in the safe area of ​​the slope, and a base station is configured to receive its Wi-Fi or 5G signal. All devices are connected to a unified cloud monitoring platform to complete network registration and time synchronization.

[0085] Step 2: Regular periodic monitoring; the system enters normalized operation mode. Each inclinometer collects local tilt angle data at a first preset frequency (e.g., once per hour) and uploads it to the monitoring platform via the NB-IoT network. The inspection robot starts according to a preset cycle (e.g., once per week), autonomously cruises along a preset path, and uses an onboard LiDAR and depth camera to perform a 3D scan of the entire slope, generating a high-resolution point cloud map, which is then transmitted back to the platform.

[0086] Step 3: Data fusion and state modeling. The monitoring platform processes multi-source data. First, spatiotemporal alignment is performed: for each scan time of the inspection robot... Using spline interpolation algorithm, the inclinometer is... Time-series data before and after a given moment are aligned to that moment to obtain synchronized local angle values. Simultaneously, based on the inclinometer's installation pose, the measured angle changes are transformed from the local coordinate system to the global coordinate system of the point cloud map, constructing a unified spatiotemporal reference framework. The geometric feature fusion module uses the local tilt data provided by the inclinometer as constraints, inputting them into the point cloud deformation analysis algorithm to optimize the calculation accuracy of macroscopic parameters such as the rock mass tilt angle and crack opening rate. The state complementarity verification module compares and analyzes the correlation between local and macroscopic data. For example, if an inclinometer experiences a sudden angle change within a short period, and the point cloud data of the adjacent area shows a small displacement at the corresponding location, it is determined to be a true response.

[0087] Furthermore, the system calculates the fusion stability index. ,in Reflecting local stability, Reflects the integrity of the macro structure. The correlation between the two is quantified. This index serves as a core indicator for assessing the overall health of the slope.

[0088] Step 4, Bidirectional Triggering and Dynamic Response: The system establishes an intelligent bidirectional triggering mechanism. When any inclinometer detects an angle change rate... If the change rate exceeds the first preset threshold (e.g., a change rate greater than 0.1° / h for two consecutive hours), it indicates that rapid deformation may occur in a local area. The monitoring platform immediately sends a high-density re-inspection command to the inspection robot. The robot interrupts its original task and prioritizes navigation to the area where the target inclinometer is located, performing a point cloud scan with a higher density than usual (e.g., doubling the resolution) to verify whether the local anomaly has caused macroscopic structural instability.

[0089] Conversely, when the inspection robot identifies macroscopic hazards such as newly added cracks or displacement of unstable rock masses in point cloud analysis, the system initiates a "dynamic-to-static" trigger. In this case, the trigger range is not a fixed value, but rather the trigger radius is dynamically calculated based on the severity of the hazard. ,in The area of ​​the potential hazard. The system directs traffic to areas centered on the hazard's location. A monitoring command is sent to all inclinometers within a circular area of ​​radius. The command includes the target area coordinates and the sampling frequency (e.g., increasing from once per hour to once every ten minutes). Upon receiving the command, the inclinometers automatically increase their operating frequency, continuously capturing minute angle changes in the area to achieve localized, refined monitoring, until the rate of angle change falls below a second preset threshold or reaches a preset time window. After that, the normal frequency was restored.

[0090] Step 5: Dynamic Assessment and Early Warning. The system conducts dynamic risk assessment based on fused data and two-way trigger feedback. Fusion Stability Index Updated in real time, when it falls below a safe threshold. ,or and When all values ​​show a continuous downward trend, the system generates a preliminary early warning message.

[0091] To further enhance the foresight of early warning systems, the system calculates the comprehensive risk growth rate. , that is, the rate of change of the reciprocal of the stability index. When If the threshold exceeds the third preset threshold, it indicates that the risk is accumulating rapidly. The system will immediately generate a high-level warning message (such as a red warning) and push it to the management personnel's terminal via SMS, email, etc., prompting them to take emergency response measures.

[0092] In summary, this embodiment achieves closed-loop monitoring of "static perception - dynamic verification - bidirectional triggering - fusion evaluation" through efficient collaboration between the inclinometer and the inspection robot, significantly improving the comprehensiveness, real-time performance and reliability of the monitoring system, and providing an intelligent solution for the safe operation and maintenance of critical infrastructure.

[0093] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for slope monitoring using a tiltmeter and an inspection robot in a coordinated manner, characterized in that, Includes the following steps: Monitoring is performed by an inclinometer deployed on the structure and a mobile inspection robot. The inclinometer continuously collects local angle data at a first preset frequency, and the inspection robot performs scanning and generates three-dimensional point cloud data according to a preset cycle. The local angle data and the three-dimensional point cloud data are spatiotemporally aligned and fused to establish a fusion stability index for the structure. A two-way triggering mechanism is established between the inclinometer and the inspection robot. When the inclinometer detects an abnormal angle change, the inspection robot is triggered to perform a high-density re-inspection. When the inspection robot identifies a potential macroscopic deformation, a nearby inclinometer is triggered to perform fine monitoring. Based on the feedback data from the fusion stability index and the bidirectional triggering mechanism, the health status of the structure is dynamically assessed and an early warning is issued.

2. The method according to claim 1, characterized in that, The monitoring is carried out using inclinometers deployed on the structure and mobile inspection robots, specifically including: Multiple inclinometers with wireless communication capabilities are deployed in the critical risk areas of the structure. The inspection robot is positioned at a preset take-off and landing point or track; The inclinometer and the inspection robot are connected to a unified monitoring platform.

3. The method according to claim 1, characterized in that, The local angle data and the 3D point cloud data are spatiotemporally aligned and fused to establish a fusion stability index for the structure, specifically including: The local angle data and the 3D point cloud data are interpolated over time and transformed in coordinate system to obtain a multi-source fusion dataset that is time-synchronized and coordinate-unified. Based on the multi-source fusion dataset, a fusion stability index for the structure is established by weighting the local stability index, macro-stability index, and spatiotemporal correlation coefficient.

4. The method according to claim 3, characterized in that, The local angle data and the 3D point cloud data are interpolated over time and transformed in coordinate system to obtain a multi-source fusion dataset that is time-synchronized and coordinate-unified. Specifically, this includes: For time-series data collected by the inclinometer, linear or spline interpolation is used to align it to the scanning time of the inspection robot; for 3D point cloud data, the angle changes in the local coordinate system are transformed to the global point cloud coordinate system based on the known installation pose of the inclinometer, thus constructing a unified multi-source fusion dataset.

5. The method according to claim 4, characterized in that, Also includes: During data fusion, the angle change obtained from the local angle data measured by the inclinometer is used as a constraint and introduced into the deformation analysis algorithm of the point cloud map.

6. The method according to claim 3, characterized in that, Also includes: By comparing the local angle mutations monitored by the inclinometer with the macroscopic deformation spatial locations identified by the inspection robot, cross-validation is performed. When the two are highly correlated in time and space, it is determined that there is a real hidden danger.

7. The method according to claim 3, characterized in that, The spatiotemporal correlation coefficient is obtained by calculating the Pearson correlation coefficient between the local rate of change of the inclinometer and the deformation gradient of the point cloud in the neighboring region, or by calculating it through the spatial distance attenuation model.

8. The method according to claim 1, characterized in that, In the bidirectional triggering mechanism, when the inclinometer detects an abnormal angle change, it triggers the inspection robot to perform a high-density re-inspection, specifically as follows: The rate of change of angle is calculated based on the angle measurements taken by the inclinometer at adjacent sampling times. When the angle change rate exceeds the first preset threshold, a re-inspection instruction is sent to the inspection robot; In response to the re-inspection command, the inspection robot interrupts its current task and prioritizes navigation to the area where the inclinometer that sent the command is located, and performs a re-inspection task with a higher scanning density than usual.

9. The method according to claim 8, characterized in that, In the aforementioned two-way triggering mechanism, when the inspection robot identifies a potential macroscopic deformation hazard, it triggers a nearby inclinometer for refined monitoring. Specifically: Identify potential macroscopic deformation hazards based on the 3D point cloud data generated by the inspection robot; The range of the adjacent area is dynamically determined based on the area or volume of the identified macroscopic deformation hazards; Send monitoring commands to inclinometers located within the adjacent area; Upon receiving the instruction, the inclinometer increases the sampling frequency from the first preset frequency to the second preset frequency, performs local fine monitoring, and continues until the rate of change of angle detected by the relevant inclinometer falls below the second preset threshold.

10. The method according to claim 1, characterized in that, Based on the feedback data from the aforementioned fusion stability index and bidirectional triggering mechanism, the health status of the structure is dynamically assessed and an early warning is issued, specifically including: When the fusion stability index is lower than the safety threshold, or when both the local stability score and the macroscopic stability score show a continuous downward trend, an early warning message is generated. The rate of change of the reciprocal of the fusion stability index over time is calculated as the comprehensive risk growth rate. When the overall risk growth rate exceeds a third preset threshold, a high-level early warning message is generated.

Images