A slope displacement monitoring device based on GNSS and microseismic fusion

Abstract

The application discloses a kind of slope displacement monitoring devices based on GNSS and microseism fusion, it is related to slope displacement monitoring field, the device includes: geodetic GNSS receiver obtains absolute displacement data of slope;Integrated in it three-axis acceleration sensor collects vibration acceleration, inclination sensor collects inclination data.Data processing module is filtered and corrected to the above-mentioned data and generates fusion dataset.Video acquisition module real-time acquisition field video image, AI video analysis module is extracted apparent abnormal characteristics by target detection and motion recognition.Fusion analysis module carries out displacement state analysis to fusion dataset and obtains fusion displacement field, and apparent abnormal characteristics and fusion displacement field are analyzed in time and space correlation to identify abnormal state, generate alarm information.Low-orbit communication module is transmitted to remote monitoring center by low-orbit satellite link with fusion displacement field data, video image and alarm information.The application improves the accuracy and sensitivity of slope displacement monitoring.

Description

Technical Field

[0001] This application relates to the field of slope displacement monitoring, and in particular to a slope displacement monitoring device based on the fusion of GNSS and microseismic data. Background Technology

[0002] Slope displacement monitoring is an important means of geological disaster prevention and control and geotechnical engineering safety assessment. Existing slope displacement monitoring technologies mainly include two categories: Global Navigation Satellite System (GNSS) surface displacement monitoring and microseismic (vibration acceleration) monitoring.

[0003] GNSS monitoring technology, based on satellite positioning systems with known spatial locations, employs the fundamental principle of spatial resection to achieve absolute positioning of ground user terminal equipment, providing high-precision surface displacement data. Geodetic GNSS receivers can achieve a static plane accuracy of ±(2.5 + 0.5 × 10⁻⁶). -6 ×D)mm, elevation accuracy can reach ±(5.0+0.5×10 -6 ×D)mm, widely used in geological disasters, geotechnical engineering and other fields.

[0004] Micro-vibration (vibration acceleration) monitoring technology uses a micro-electro-mechanical system (MEMS) triaxial accelerometer installed inside the object under test to accurately measure the vibration response signal of the object under test under different vibration source excitation, so as to understand the stability of the monitored object, monitor the status in real time, and issue early warning information.

[0005] However, in existing technologies, GNSS monitoring and microseismic monitoring are usually deployed and operated as two independent systems, which has the following shortcomings: (1) Data isolation: GNSS displacement data and microseismic vibration data are collected and analyzed separately, failing to achieve effective integration and making it difficult to fully reflect the stability state of the slope.

[0006] (2) Response lag: When relying solely on GNSS displacement monitoring, anomalies can only be detected after significant slope displacement occurs, lacking early detection of microseismic signals that foreshadow slope instability.

[0007] (3) High false alarm rate: Single monitoring methods are easily affected by environmental interference (such as temperature changes, electromagnetic interference, vibration and noise), leading to abnormal data or false alarms.

[0008] (4) High power consumption: Existing GNSS equipment consumes a lot of power and has limited endurance when there is no mains power supply in the field.

[0009] (5) Lack of adaptive adjustment capability: Existing equipment cannot automatically adjust the working mode according to the actual monitoring status, making it difficult to achieve a dynamic balance between monitoring accuracy and power consumption.

[0010] Some existing technologies attempt to combine GNSS with accelerometers for landslide monitoring. For example, deformation data is calculated using GNSS and then monitored in real-time by an IMU, or the ratio of acceleration measurements to displacement measurements is determined to be greater than a preset threshold. However, in these solutions, GNSS and accelerometers still operate relatively independently, failing to achieve deep data fusion and coordinated control. Furthermore, existing technologies typically use separate GNSS receivers and microseismic sensors, resulting in low integration, complex deployment, and a lack of adaptive operating mode adjustment mechanisms based on fused data. Summary of the Invention

[0011] The purpose of this application is to provide a slope displacement monitoring device based on the fusion of GNSS and microseismic data, which can solve the technical problems of GNSS and microseismic monitoring being independent, data not being effectively integrated, and lacking adaptive adjustment capabilities in existing slope displacement monitoring, and achieve high-precision, low-power, all-weather integrated slope displacement monitoring.

[0012] To achieve the above objectives, this application provides a slope displacement monitoring device based on GNSS and microseismic fusion, comprising: Geodetic GNSS receivers are used to receive signals from satellite positioning and navigation systems and acquire absolute displacement data of slope monitoring points; A triaxial accelerometer is integrated inside the geodesic GNSS receiver to collect triaxial vibration acceleration data of slope monitoring points in real time. The tilt sensor, integrated inside the geodetic GNSS receiver, is used to collect tilt data from slope monitoring points in real time. The data processing module is used to perform filtering and correction processing on the absolute displacement data, the triaxial vibration acceleration data and the tilt angle data respectively, and generate a fused dataset; The video acquisition module is used to acquire video image data at slope monitoring points in real time. The AI ​​video analysis module is used to perform target detection and motion recognition processing on the video image data and extract apparent anomaly feature data at the slope monitoring point; The fusion analysis module is used to fuse the absolute displacement data, triaxial vibration acceleration data and tilt angle data in the fusion dataset and perform displacement state analysis to obtain fused displacement field data at the slope monitoring point. The module also performs correlation analysis between the apparent anomaly feature data and the fused displacement field data to identify the abnormal state at the slope monitoring point and generate anomaly alarm information when an abnormal state is identified. The low-Earth orbit communication module is used to transmit fused displacement field data, video image data and / or abnormal alarm information at the slope monitoring point to a remote monitoring center via a low-Earth orbit satellite communication link.

[0013] According to the specific embodiments provided in this application, this application achieves the following technical effects: A geodetic GNSS receiver provides high-precision absolute displacement, while the integrated triaxial accelerometer and tilt sensor simultaneously capture micro-vibrations and attitude changes. The data processing module filters and corrects the three types of raw data, eliminating environmental noise and instrument errors. The fusion analysis module further mines complementary information between multi-source data, effectively reducing the drift error of a single GNSS receiver in dynamic or obstructed environments, and significantly improving the overall accuracy and long-term stability of the slope displacement field. Furthermore, the triaxial acceleration and tilt data can sensitively reflect early deformation precursors such as micro-fractures and increased tilting in the soil and rock mass, while the GNSS data provides macroscopic displacement trends. Fusion analysis can simultaneously acquire high-frequency vibration characteristics and low-frequency displacement patterns, achieving a full-dimensional characterization from static deformation to dynamic instability, providing richer judgment criteria for landslide early warning, and effectively reducing the risk of missed and false alarms. Building upon this foundation, an AI video analytics module is introduced to perform target detection and motion recognition for apparent anomalies. This allows for cross-verification between fused displacement field data and visual appearance features. It enhances alarm reliability by providing video confirmation when sudden displacement changes occur, and offers early warnings when apparent anomalies are first observed before significant displacement changes, significantly reducing the risk of missed and false alarms. Simultaneously, a low-Earth orbit (LEO) communication module is added. Utilizing LEO satellite links, it enables efficient and reliable transmission of fused displacement field data, anomaly alarm information, and video clips in remote mountainous areas with weak or interrupted 4G / 5G network coverage. It also supports remote command issuance and adaptive device control. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of a slope displacement monitoring device based on GNSS and microseismic fusion, provided as an embodiment of this application.

[0016] Figure 2 This is a schematic diagram of a geodesic GNSS receiver in one embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] In one exemplary embodiment, a slope displacement monitoring device based on GNSS and microseismic fusion is provided. This device is applicable to absolute slope displacement monitoring scenarios in fields such as geological disaster monitoring, geotechnical engineering safety monitoring, building deformation monitoring, transportation infrastructure monitoring, water conservancy engineering monitoring, and municipal engineering monitoring. In this embodiment, as shown... Figure 1 As shown, the slope displacement monitoring device based on GNSS and microseismic fusion includes: a geodetic GNSS receiver 101, a triaxial accelerometer 102, a tilt sensor 103, a data processing module 104, a video acquisition module 105, an AI video analysis module 106, a fusion analysis module 107, and a low-Earth orbit communication module 108. The triaxial accelerometer 102 and the tilt sensor 103 are integrated within the geodetic GNSS receiver 101. The data processing module 104, the AI ​​video analysis module 106, the fusion analysis module 107, and the low-Earth orbit communication module 108 are deployed within a computer device, which performs the corresponding functions. The functions of each component are described below.

[0020] (a) The geodetic GNSS receiver 101 is used to receive signals from the satellite positioning and navigation system and obtain the absolute displacement data of the slope monitoring point.

[0021] In a specific application example, the geodetic GNSS receiver 101 is compatible with the four major global satellite positioning and navigation systems, and features high precision, low power consumption, and stable reliability. For example... Figure 2 As shown, the geodetic GNSS receiver 101 includes a GNSS monitoring host 201, a solar panel 202, a solar controller 203, a battery 204, a GNSS antenna 205, and a 4G antenna 206. The main features of the geodetic GNSS receiver 101 are as follows: (1) Supports 11 frequency signals from four systems: BEIDOU B1 / B2 / B3, GPS L1 / L2 / L5, GLONASS G1 / G2, and GALILEO E1 / E5a / E5b. It has built-in high-precision adjustment software to achieve high-precision positioning through multi-system and multi-frequency fusion.

[0022] (2) Integrating tilt angle and acceleration modules to assist in the detection of sudden changes in displacement data and drive the adaptive adjustment of the observation working mode.

[0023] (3) Based on chip-level independent development, the overall power consumption is greatly reduced, making it more suitable for continuous operation in the field.

[0024] (4) Supports data communication transmission modes such as 4G / NB-IoT, wireless bridge, and Ethernet, and supports communication protocols such as MQTT, TCP / IP, NTRIP, and JT808.

[0025] (5) Equipped with a high-performance industrial-grade 32-bit ARM core processor, it has powerful local computing capabilities and a built-in embedded web server, enabling remote two-way interaction.

[0026] (6) Based on the application scenario and monitoring requirements, support dynamic (static) high-precision calculation.

[0027] (7) The front end embeds algorithms such as stellar day filtering, Kalman filtering, and mutation detection, and the cloud integrates models such as big data cleaning, trend analysis, gray prediction, and tangent angle analysis to greatly ensure data reliability.

[0028] (8) The product adopts an aluminum whole milled shell, internal and external sealing protection, and anti-drop soft rubber design, which has excellent protection, anti-drop and shock absorption performance.

[0029] The technical parameters of the geodetic GNSS receiver 101 are shown in Table 1.

[0030] Table 1 Technical parameters of the geodetic GNSS receiver 101

[0031] The working principle of the geodetic GNSS receiver 101 is as follows: based on a satellite positioning system with a known spatial location and employing the basic principle of space rendezvous, it achieves absolute positioning of ground user terminal equipment. The receiver terminal deployed on-site remotely transmits observation data wirelessly to the monitoring center (generally received and stored in the cloud) according to a predetermined acquisition and transmission frequency. The monitoring center analyzes and processes the observation data in real time for use by relevant technical and management departments. From the perspective of data flow and functional division, the entire monitoring system can be divided into three main parts: the space segment, the ground reference station segment, and the monitoring terminal.

[0032] To improve observation accuracy and reliability, the following considerations should be taken into account when selecting observation sites for the geodetic GNSS receiver 101: (1) The field of view is relatively open, and there are no large areas of obstruction within a 15° elevation angle range of the field of view.

[0033] (2) Keep away from high-power radio sources (such as TV stations, radio stations, microwave stations, etc.) at a distance of not less than 200 meters, and keep away from high-voltage transmission lines and microwave radio signal transmission channels at a distance of not less than 50 meters.

[0034] (3) There should be no objects that strongly reflect signals nearby (such as large buildings, large bodies of water, etc.).

[0035] (4) The transportation is relatively convenient, which is conducive to construction, surveying and subsequent maintenance.

[0036] (5) The benchmark point surveying shall be carried out in accordance with the design plan, and its foundation shall be guaranteed to be stable.

[0037] (6) The setting of monitoring points shall be carried out in accordance with the design plan. If fine adjustments are necessary, they shall be determined in full communication with the design unit.

[0038] To ensure real-time transmission, processing, and application analysis of observation data, the geodetic GNSS receiver 101 provides relevant basic information after installation, facilitating cloud-based configuration of the project and equipment. This basic information includes: project name and number, approximate location (meter-level accuracy), serial number, communication card number and International Mobile Equipment Identity (IMEI), acquisition frequency and processing frequency, and real-world photos of the equipment installation.

[0039] After the geodetic GNSS receiver 101 is installed and configured, it immediately enters the continuous observation operation state. The main unit of the equipment is equipped with corresponding indicator lights to determine whether the equipment is normal in the field. The definitions of the equipment status indicator lights are shown in Table 2.

[0040] Table 2 Status Indicator Definitions

[0041] (ii) The triaxial accelerometer 102 is used to collect triaxial vibration acceleration data of the slope monitoring point in real time.

[0042] In a specific application example, the triaxial accelerometer 102 integrates a high-precision triaxial MEMS vibration sensor to accurately measure the vibration response signal of the test object under different vibration source excitations and transmit it back to the vibration signal data acquisition instrument in real time. The vibration signal data acquisition instrument is portable and solar-powered, connected one-to-one with the triaxial accelerometer 102 via wired connection, automatically completing the entire real-time vibration data acquisition process, including sensor self-testing, automatic vibration servo triggering recording, data storage, and data transmission. Vibration signal data analysis software deployed in the cloud provides medium data such as amplitude, dominant frequency, duration, and vibration velocity attenuation coefficient, as well as velocity time-history curves. Technicians can remotely log in to the system at any time to query vibration data curves and charts in real time and analyze and assess potential risks online.

[0043] The triaxial accelerometer 102 measures X, Y, and Z axes, has a frequency range of 0Hz to 1000Hz, a range of ±2g or ±4g, a sensitivity of 0.5mg, an A / D resolution of 20 bits, an amplitude nonlinearity of less than 2%, a temperature drift of less than 0.05mg / ℃, an IP68 package rating, and an operating temperature range of -45℃ to 85℃.

[0044] The vibration signal data acquisition instrument features a full Chinese LCD screen display, is powered by a rechargeable lithium battery, has 32GB of storage, a sampling rate of 0Hz~1000Hz, and includes timed recording, triggered recording, and a combination of timed and triggered recording modes. It has an IP67 rating and an operating temperature range of -45℃~85℃.

[0045] The triaxial accelerometer 102 is a device that can measure the tilting and vibration of a monitored object during human activities such as industrial production, construction, and transportation. By observing the characteristic data on the triaxial accelerometer 102, the stability of the entire monitored object can be grasped, the status can be monitored in real time, and early warning information can be issued, which is of great significance for disaster prevention and mitigation.

[0046] (iii) Inclination sensor 103 is used to collect inclination data of slope monitoring points in real time.

[0047] (iv) The data processing module 104 is used to perform filtering and correction processing on the absolute displacement data, the triaxial vibration acceleration data and the tilt angle data respectively, and generate a fused dataset.

[0048] In a specific application example, the data processing module 104 incorporates an embedded web server and is equipped with an industrial-grade 32-bit ARM core processor. The data processing module 104 includes the following functional units.

[0049] The time synchronization unit is used to match the time stamps of the absolute displacement data and the tilt angle data with the time stamp of the absolute displacement data as a reference, and to unify the absolute displacement data, the triaxial vibration acceleration data and the tilt angle data to the same time reference by using an interpolation alignment method.

[0050] Specifically, the timestamps of the absolute displacement data are extracted, and the triaxial vibration acceleration data and tilt angle data are linearly interpolated or spline interpolated according to the timestamps, so that the three data are strictly synchronized in time, forming a time-aligned original synchronized dataset.

[0051] The format unification unit is used to unify the format of time-aligned absolute displacement data, triaxial vibration acceleration data, and tilt angle data. Specifically, it converts data from different data sources into a unified preset data format and adds data source identifiers and timestamp tags.

[0052] The displacement processing unit is used to sequentially perform sidereal filtering, Kalman filtering, cycle slip detection and repair processing on the absolute displacement data to obtain preprocessed absolute displacement data.

[0053] Specifically, firstly, sidereal filtering is performed on the absolute displacement data. Based on the sidereal-day repetition characteristic of multipath errors in GNSS measurements, a multipath error model is extracted with a sidereal-day period. Periodic multipath error components are subtracted from the raw GNSS observations, eliminating systematic errors caused by the repetitive cycles of the satellite constellation. Then, Kalman filtering is performed on the sidereal-day filtered absolute displacement data to establish a state-space model of the GNSS displacement. Through predictive-update iterative recursion, the absolute displacement data is optimally estimated, filtering out observation noise and random errors, and outputting a smoothed absolute displacement estimate. Finally, cycle slip detection and repair are performed on the smoothed absolute displacement estimate. Cycle slip detection is performed based on the epoch difference or MW (Melbourne-Wübbena) combination of carrier phase observations, and epochs where cycle slips occur are repaired or marked.

[0054] An acceleration processing unit is used to sequentially perform zero-bias correction, digital filtering, temperature drift compensation, and gravity acceleration separation processing on the triaxial vibration acceleration data to obtain preprocessed triaxial vibration acceleration data.

[0055] Specifically, firstly, zero-bias correction is performed on the triaxial vibration acceleration data. Based on the sensor's factory calibration parameters and the zero-bias measurement values ​​under static conditions, the zero-bias components of each of the three axes are subtracted. Then, digital filtering is performed on the zero-bias-corrected triaxial vibration acceleration data. A bandpass filter is used to remove DC components and high-frequency noise outside the frequency range, retaining the effective vibration signal within the 0Hz~1000Hz frequency band. Next, temperature drift compensation is performed on the filtered triaxial acceleration signal. Based on the temperature drift coefficient of the triaxial accelerometer 102 (less than 0.05mg / ℃) and the real-time monitored ambient temperature, temperature compensation correction is applied to the acceleration measurements. Finally, gravity acceleration separation is performed on the temperature-compensated triaxial acceleration signal. Gravity acceleration components and dynamic vibration acceleration components are separated from the triaxial vibration acceleration data using low-pass filtering or the critical damping method. The gravity acceleration component is used to assist in tilt angle calculation, while the dynamic vibration acceleration component is used for vibration characteristic analysis.

[0056] The tilt angle processing unit is used to perform zero-point correction and low-pass filtering on the tilt angle data in sequence, and to perform cross-validation and tilt angle deviation correction on the low-pass filtered tilt angle data based on the pre-processed triaxial vibration acceleration data to obtain the pre-processed tilt angle data.

[0057] Specifically, firstly, zero-position calibration is performed on the tilt data based on the initial zero-position calibration value of the tilt sensor 103. Then, low-pass filtering is performed on the zero-position calibrated tilt data to filter out high-frequency vibration noise interference with tilt measurement and extract stable tilt change trends. Finally, the filtered tilt data is cross-validated with the gravitational acceleration component. When the deviation between the two calculated tilt values ​​exceeds a preset threshold, the fused weighted value is used as the final tilt output.

[0058] The fusion detection unit is used to generate a fusion dataset based on the preprocessed absolute displacement data, preprocessed triaxial vibration acceleration data, and preprocessed tilt angle data.

[0059] (v) The video acquisition module 105 includes one or more high-definition industrial cameras, installed near the slope monitoring point. The video acquisition module 105 has a built-in infrared supplementary light device and night vision function, used for all-weather real-time acquisition of video image data at the slope monitoring point. The video acquisition module 105 is connected to the AI ​​video analysis module 106 via wired or wireless means, and transmits the acquired video image data to the AI ​​video analysis module 106.

[0060] (vi) The AI ​​video analysis module 106 employs a deep learning model deployed on edge computing devices or cloud servers to perform target detection and motion recognition processing on video image data, extracting apparent anomaly feature data at slope monitoring points. Specifically, the AI ​​video analysis module 106 includes: a target detection unit, a motion recognition unit, and an anomaly event classification unit.

[0061] The target detection unit is used to perform real-time detection and location calibration of targets such as surface cracks, landslide boundaries, rockfall deposits, and abnormal vegetation changes in the slope monitoring area based on target detection algorithms such as YOLO or Faster R-CNN, and generate target detection results.

[0062] The motion recognition unit is used to perform motion vector analysis on video image sequences based on optical flow or inter-frame difference methods, extract the apparent displacement field and deformation rate of the soil and rock mass in the slope monitoring area, and generate apparent motion feature data.

[0063] The abnormal event classification unit is used to input the target detection results and apparent motion feature data into a pre-trained abnormal classification model to identify abnormal events such as slope collapse, rockfall, and slope sliding, and output the abnormal event type and confidence level, generate apparent abnormal feature data, and transmit it to the fusion analysis module 107.

[0064] (vii) The fusion analysis module 107 is used to perform fusion processing and displacement state analysis on the absolute displacement data, triaxial vibration acceleration data and tilt angle data in the fusion dataset to obtain fusion displacement field data at the slope monitoring point, and to perform spatiotemporal correlation analysis on the apparent anomaly feature data and the fusion displacement field data to identify the abnormal state at the slope monitoring point, and generate anomaly alarm information when an abnormal state is identified.

[0065] In a specific application example, the fusion analysis module 107 includes the following functional units.

[0066] The displacement mutation screening unit is used to calculate the rate of change of absolute displacement data within the current window. A dynamic adaptive threshold is used as the mutation discrimination threshold, and epoch segments that continuously exceed the dynamic adaptive threshold are marked as candidate displacement mutation time periods, generating candidate mutation time points. For the first j GNSS displacement values ​​of an epoch. For the first j GNSS displacement value at epoch -1 This represents the rate of change of displacement.

[0067] Specifically, the formula for calculating the dynamic adaptive threshold is as follows: ;in, For dynamic adaptive threshold, This represents the mean rate of change of displacement within the current window. The standard deviation of the rate of change of displacement within the current window. k Preset coefficient (preferred) k =3~5).

[0068] The microseismic feature parameter extraction unit is used to construct an analysis time window centered on the candidate abrupt change time point, and extract the vibration feature parameters of the triaxial vibration acceleration data within the analysis time window.

[0069] Specifically, using the candidate mutation time points t m Centered on the data, extend the data forward and backward by a preset time length (preferably 30 seconds forward and 10 seconds backward) as the analysis time window.

[0070] The vibration characteristic parameters include short-time vibration energy, dominant vibration frequency, and vibration velocity attenuation coefficient.

[0071] The short-time vibration energy is ;in, It is short-term vibrational energy. Sampling points i Place x The acceleration value of the shaft. Sampling points i Place y The acceleration value of the shaft. Sampling points i Place z The acceleration value of the shaft. The sampling time interval, p To determine the starting sampling point of the analysis time window, q This is the end sampling point of the analysis time window.

[0072] The dominant vibration frequency f 0 is obtained by performing a fast Fourier transform on the triaxial vibration acceleration data within the analysis time window.

[0073] The vibration velocity attenuation coefficient α is obtained by exponentially fitting the triaxial vibration acceleration data within the analysis time window. .in, v ( t Within the analysis time window t The acceleration value at time t. v 0 represents the initial acceleration value within the analysis time window.

[0074] The effective mutation determination unit is used to confirm the effective mutation of the candidate mutation time point based on the relationship between the vibration characteristic parameters and the preset threshold, and obtain the effective displacement mutation point.

[0075] Specifically, the effective mutation determination unit uses a comprehensive voting decision rule to confirm the effective mutation of the candidate mutation time point. The comprehensive voting decision rule is as follows: when the ratio of short-term vibration energy to the average vibration energy during the preceding stable period exceeds a first energy multiple threshold, and the drift of the dominant vibration frequency exceeds a preset dominant frequency drift threshold or the vibration velocity attenuation coefficient is lower than a preset attenuation threshold, the candidate mutation time point is determined as an effective displacement mutation point. When the ratio of short-term vibration energy to the average vibration energy during the preceding stable period exceeds the first energy multiple threshold, but the drift of the dominant vibration frequency does not exceed the preset dominant frequency drift threshold and the vibration velocity attenuation coefficient is lower than the preset attenuation threshold, the candidate mutation time point is marked as a suspected interference mutation point.

[0076] The comprehensive voting decision-making rules include the following three judgment conditions: Judgment condition A (energy mutation criterion): When short-time vibrational energy E short Compared to the average vibrational energy during the preceding stationary period (600 seconds to 30 seconds before the candidate mutation moment). When the ratio exceeds the first energy multiple threshold (preferably 3 times), that is... This indicates that a significant vibration event occurred at the monitoring point, making it a candidate time point for abrupt changes. t m Marked as a valid mutation candidate.

[0077] Judgment condition B (dominant frequency drift criterion): When the dominant vibration frequency... f 0 compared to the average frequency during the preceding stable period When the drift exceeds the preset main frequency drift threshold (preferably 20%), that is... This indicates a significant change in the vibration frequency caused by internal rupture or slippage of the slope, further confirming the possibility of an effective abrupt change.

[0078] Judgment condition C (attenuation anomaly criterion): When the vibration velocity attenuation coefficient α is lower than the preset attenuation threshold (preferably α<2), it indicates that the vibration energy attenuates slowly, which is consistent with the vibration characteristics of slow rupture of large blocks before the slope is about to slide, further supporting the determination of effective abrupt change.

[0079] When a candidate mutation time point is satisfied with at least one of the following conditions: A, B, or C, it is determined as an effective displacement mutation point. When a candidate mutation time point is satisfied with only condition A but not conditions B or C, it is marked as a suspected interference mutation point.

[0080] The spatial consistency verification unit is used to perform spatial consistency verification on absolute displacement data using tilt angle data, and to determine the confidence level of effective displacement abrupt change points.

[0081] Specifically, the spatial consistency verification unit includes a similarity calculation subunit and a confidence leveling subunit.

[0082] The similarity calculation subunit is used to extract the tilt angle change vector within a preset time window before and after the effective displacement abrupt point. and absolute displacement change vector And calculate the cosine similarity between the tilt angle change vector and the absolute displacement change vector. .in, The vector representing the change in tilt angle. for x The change in the tilt angle of the axis. for y The change in the tilt angle of the axis. for z The change in the tilt angle of the axis. It is the absolute displacement vector. for x The absolute change in displacement of the shaft. for y The absolute change in displacement of the shaft. for z The absolute change in displacement of the shaft. Sim The cosine similarity between the tilt angle change vector and the absolute displacement change vector.

[0083] The confidence leveling subunit is used when the cosine similarity is greater than a first preset similarity threshold (preferred). Sim When the inclination angle change direction is greater than 0.7 (indicating that the inclination angle change direction is consistent with the GNSS displacement direction), the confidence level of the effective displacement abrupt change point is set to the first confidence level (high confidence). The cosine similarity is preferably between the first preset similarity threshold and the second preset similarity threshold. Sim When the cosine similarity is between 0.4 and 0.7, the confidence level of the effective displacement abrupt change point is set to the second confidence level (medium confidence). When the cosine similarity is less than or equal to the second preset similarity threshold (preferably...),... Sim When the inclination angle change direction is ≤0.4 (indicating that the inclination angle change direction is inconsistent with the GNSS displacement direction), the confidence level of the effective displacement abrupt change point is set to the third confidence level (low confidence) and marked as pending review. The first preset similarity threshold is greater than the second preset similarity threshold.

[0084] Furthermore, this application further evaluates the mutation level of effective displacement mutation points, and the evaluation criteria are as follows: Level I sudden change (micro-change level): the rate of change of displacement at the effective displacement sudden change point. Located at the dynamic adaptive threshold Within the range of 1 to 2 times, and short-term vibrational energy E short This level of abrupt change corresponds to micro-fractures or localized small-scale adjustments in the shallow surface layer of the slope, and is 3 to 5 times the average energy of the preceding stable period.

[0085] Level II abrupt change (intermediate level): Rate of displacement change at the effective displacement abrupt change point Located at the dynamic adaptive threshold Within the range of 2 to 4 times, and short-term vibrational energy E short This level of abrupt change corresponds to an intensification of activity in the deep rupture zone of the slope or an acceleration of local slippage, and is 5 to 10 times the average energy of the preceding stable period.

[0086] Level III Sudden Change (Rapid Transformation): Rate of displacement change at the point of effective displacement change. Exceeding the dynamic adaptive threshold More than four times that of short-term vibrational energyE short The energy level is more than 10 times the average energy of the preceding stable period, and the vibration velocity attenuation coefficient α is less than 1.5. This sudden change corresponds to the precursor or rapid sliding stage of the overall slope instability.

[0087] Furthermore, this application also performs the following interference removal operation on candidate mutation points identified as suspected interference mutation points, and valid displacement mutation points with a low confidence level: (a) Extract the GNSS satellite geometric accuracy factor (PDOP value) within the time period. When the PDOP value exceeds the preset PDOP threshold (preferably PDOP>4), the sudden change is attributed to the positioning jump caused by poor satellite geometric distribution and is removed.

[0088] (b) Extract the rate of change of ambient temperature during the period. When the rate of change of ambient temperature exceeds the preset temperature change rate threshold (preferably 2℃ / min), the sudden change is attributed to the thermal expansion and contraction effect of the geodesic GNSS receiver 101 caused by the sudden temperature change or the residual temperature drift of the triaxial accelerometer 102, and is discarded.

[0089] (c) For suspected interference mutation points that cannot be attributed after being excluded by (a) and (b), they are marked as "interference of unknown origin" and stored. When multiple interferences with the same characteristics appear in succession in subsequent periods, the self-check process of data processing module 104 is triggered.

[0090] The displacement field determination unit is used to generate fused displacement field data at the slope monitoring point based on the fused dataset, the effective displacement abrupt change point, and the confidence level of the effective displacement abrupt change point.

[0091] Specifically, the displacement field determination unit includes: an initial weight determination subunit, a weight normalization subunit, an acceleration calculation subunit, an inclination calculation subunit, a filtering and fusion subunit, a weighted summation subunit, and a secondary fusion subunit.

[0092] The initial weight determination subunit is used to evaluate the confidence of the absolute displacement data, triaxial vibration acceleration data and tilt angle data in the fused dataset, and determine the initial values ​​of the displacement confidence weight, acceleration confidence weight and tilt angle confidence weight.

[0093] (1) The confidence assessment process for absolute displacement data is as follows: obtain the satellite geometric precision factor PDOP value, carrier phase lock status, cycle slip repair mark and differential age of the geodesic GNSS receiver 101 at the current time, and calculate the initial value of the displacement confidence weight. W GNSS : ;in, and Preset empirical coefficients (preferred) , ), The standard deviation of the GNSS positioning solution (output by the geodesic GNSS receiver 101). PDOP When the value is larger or the positioning accuracy is worse, W GNSS The smaller the value, the better. Additionally, if the current epoch is marked as a cycle slip repair epoch, it is multiplied by an additional attenuation factor of 0.8; if it is marked as a differential signal interruption epoch, it is multiplied by an additional attenuation factor of 0.5.

[0094] (2) The confidence assessment of triaxial vibration acceleration data is as follows: obtain the signal amplitude, vibration main frequency stability and temperature change rate of the triaxial accelerometer 102 at the current moment, and calculate the initial value of the acceleration confidence weight. : ;in, The temperature change rate is expressed as °C / s. It is the instantaneous coefficient of variation of the dominant vibration frequency (the ratio of the standard deviation to the mean of the dominant frequency within a preset time window). and Preset empirical coefficients (preferred) , The greater the rate of temperature change or the more unstable the dominant frequency, the greater the influence of environmental interference or nonlinear vibration on the acceleration signal, and the lower the confidence level.

[0095] (3) The confidence evaluation process for tilt angle data is as follows: obtain the tilt angle change rate of tilt angle sensor 103 at the current moment and the deviation from the tilt angle calculated by acceleration and gravity components, and calculate the initial value of tilt angle confidence weight. W tilt : ;in, The rate of change of tilt angle (° / s) This represents the deviation between the tilt sensor reading and the tilt angle calculated based on the acceleration-gravity component. and Preset empirical coefficients (preferred) , When the rate of change is too large or the tilt angle of the two paths deviates significantly, the confidence level decreases.

[0096] The weight normalization subunit is used to dynamically adjust and normalize the initial values ​​of the displacement confidence weight, the initial value of the acceleration confidence weight, and the initial value of the tilt angle confidence weight according to the effective displacement abrupt change point and the confidence level of the effective displacement abrupt change point, respectively, to obtain the normalized displacement confidence weight, the normalized acceleration confidence weight, and the normalized tilt angle confidence weight.

[0097] As an alternative implementation, if the current moment is marked as an effective displacement abrupt change point (any level), the initial value of the displacement confidence weight is multiplied by a first enhancement factor (preferably 1.5), because the displacement change at the abrupt change moment is the most direct evidence of the real displacement response and should be given a higher fusion weight.

[0098] If the current moment is marked as an effective displacement abrupt change point and the abrupt change level is level II or III, then the initial value of the acceleration confidence weight is multiplied by the second enhancement factor (preferably 1.3), because abrupt changes are accompanied by strong microseismic signals, and acceleration data helps to capture rapid dynamic responses.

[0099] If the current moment is marked as a suspected interference mutation point, the initial value of the displacement confidence weight is multiplied by the first attenuation factor (preferably 0.6), and the initial value of the tilt confidence weight is multiplied by the second enhancement factor (preferably 1.2). Since the tilt data is relatively stable, it can provide auxiliary constraints when the GNSS is interfered with.

[0100] After adjustment, the weights are normalized to ensure that the sum of the three is 1. ; ; ; in, To normalize the displacement confidence weights, To normalize the acceleration confidence weights, To normalize the confidence weights of the tilt angle, The adjusted displacement confidence weights, The adjusted acceleration confidence weights, This is the adjusted tilt confidence weight.

[0101] The acceleration estimation subunit is used to perform a second integration on the dynamic acceleration components of the triaxial vibration acceleration data to obtain the acceleration integral displacement. The tilt angle data is used to adaptively compensate for the integral displacement of acceleration to obtain the acceleration-calculated displacement.

[0102] Specifically, to avoid integral drift, the following constraint correction strategy is adopted to adaptively compensate for the drift of the acceleration integral displacement: (1) For dynamic acceleration components Bandpass filtering (cutoff frequency 0.01Hz~100Hz) is performed to remove DC and very low frequency drift components.

[0103] (2) Velocity integration is performed using the trapezoidal integral method: The velocity sequence is then mean-reduced to eliminate zero drift; among which, It is a velocity sequence.

[0104] (3) Perform displacement integration: ;in, The integral of acceleration and displacement.

[0105] (4) Use dip angle data to constrain the trend of acceleration integral displacement: When the rate of change of dip angle is continuously negative (or positive), it indicates that the slope is continuously tilted in a certain direction. At this time, the acceleration integral displacement should follow the same monotonic trend. If reverse drift occurs, the linear trend correction term constructed by the dip angle data is used. : Adaptive drift compensation is applied to the integral displacement of acceleration, and the compensated acceleration-calculated displacement is denoted as... .in, L The installation height of the tilt sensor. This represents the change in tilt angle.

[0106] The tilt angle calculation subunit is used to calculate the tilt angle displacement based on the tilt angle data. : .

[0107] The filtering and fusion subunit is used to process the absolute displacement data. Displacement calculated from the acceleration Complementary filtering fusion is performed to obtain the complementary filtering result. : .in, For complementary filter coefficients (preferred) This fusion method ensures that the fused displacement converges to the GNSS absolute reference in the long term and reflects the high-frequency dynamic changes of acceleration capture in the short term.

[0108] The weighted summation subunit is used to perform a weighted summation of the absolute displacement data, the acceleration-estimated displacement, and the tilt-estimated displacement based on the normalized displacement confidence weight, the normalized acceleration confidence weight, and the normalized tilt confidence weight, to obtain a preliminary fused displacement field. : .

[0109] The secondary fusion subunit is used to perform secondary fusion of the preliminary fused displacement field and the complementary filtering result to obtain the fused displacement field data at the slope monitoring point. : Wherein, λ is the fusion coefficient, which is adaptively adjusted according to the slope deformation rate: when the deformation rate is less than the preset slow threshold (e.g., 0.01 mm / min), λ takes a larger value (preferably 0.8), and the final fusion value is mainly based on the weighted average absolute benchmark to ensure long-term stability; when the deformation rate is greater than the preset fast threshold (e.g., 0.1 mm / min), λ takes a smaller value (preferably 0.4), and the final fusion value is mainly based on the complementary filter dynamic response to ensure fast tracking capability.

[0110] Further integrate displacement field data Decomposed into horizontal northward components Horizontal eastward component and vertical components And calculate the rate of change of the horizontal northward component respectively. The rate of change of the horizontal eastward component Vertical components Horizontal northward component of acceleration Horizontal eastward component of acceleration and the change in acceleration of the vertical component .

[0111] The spatiotemporal analysis unit is used to perform spatiotemporal correlation analysis on the apparent anomaly feature data and the fused displacement field data to identify abnormal states at the slope monitoring points and generate anomaly alarm information when an anomaly state is identified. Specifically, the spatial location and time label of the apparent anomaly feature data are first extracted and matched with the effective displacement mutation points in the fused displacement field data in terms of spatial proximity and temporal consistency. When the match is successful, the corresponding displacement mutation event in the fused displacement field data is marked as "video-related confirmed anomaly" and its anomaly alarm confidence is increased. When the AI ​​video analysis module 106 detects an anomaly event but the fused displacement field data does not undergo a significant mutation, the video analysis result is output as independent anomaly evidence, and the data processing module adjusts the subsequent data acquisition strategy, such as triggering an encrypted monitoring mode or activating additional sensors.

[0112] In another exemplary embodiment, the fusion analysis module 107 is further configured to use a sliding window method (preferably with a window length of 300 epochs) to calculate the residual sequence between the fused displacement field data and the original GNSS displacement, acceleration integral displacement, and tilt-estimated displacement, thereby estimating the noise variance and displacement covariance of each data source in real time. Acceleration covariance Inclination covariance Based on the minimum variance criterion, the empirical coefficients are dynamically updated every 100 epochs. , , This allows the fusion weights to be automatically optimized over long-term operation. ;in, J The cost function for fusing displacement and the sum of squared residuals from each original data set. η The learning rate is set to 0.01 (preferably). Through online adaptive learning, the fusion algorithm can adapt to different slope monitoring scenarios and environmental changes, continuously maintaining optimal fusion performance.

[0113] The adaptive adjustment unit is used to adaptively adjust the observation working mode of the geodesic GNSS receiver 101 according to the effective displacement abrupt change point and the confidence level of the effective displacement abrupt change point.

[0114] Specifically, the observation working modes of the geodetic GNSS receiver 101 include conventional monitoring mode, encrypted monitoring mode, and low-power standby mode.

[0115] The conventional monitoring mode is as follows: when an effective displacement mutation point is detected and the confidence level of the effective displacement mutation point is within a safe range and continues for a preset time, the geodetic GNSS receiver 101 operates at a first sampling frequency, and the triaxial accelerometer 102 and the tilt sensor 103 operate at a second sampling frequency.

[0116] The encrypted monitoring mode is as follows: when a valid displacement mutation point is detected and the confidence level of the valid displacement mutation point is not within the safe range, the geodetic GNSS receiver 101 switches to the third sampling frequency, and the triaxial accelerometer 102 and the tilt sensor 103 switch to the fourth sampling frequency.

[0117] The low-power standby mode is as follows: when no effective displacement abrupt change point is detected, the geodesic GNSS receiver 101 switches to the fifth sampling frequency.

[0118] Wherein, the third sampling frequency is greater than the first sampling frequency, the fourth sampling frequency is greater than the second sampling frequency, and the fifth sampling frequency is less than the first sampling frequency.

[0119] In another exemplary embodiment, the fusion analysis module 107 is also used to periodically (preferably every 24 hours) statistically analyze the data distribution of all effective displacement mutation points and suspected interference mutation points for the day, and dynamically update the dynamic adaptive threshold using an adaptive learning mechanism. coefficients in k Specifically, when the proportion of suspected interfering mutation points exceeds 30% of the total candidate mutation points for several consecutive days (preferably 5 consecutive days), the coefficient is automatically adjusted upwards. k The value is adjusted (increased by 0.2 each time, with a maximum of 6) to reduce the false alarm rate; when the missed detection rate of effective displacement abrupt change points exceeds the preset missed detection rate threshold (preferably 5%) for several consecutive days, the coefficient is automatically reduced. kThe value is adjusted (decreasing by 0.2 each time, with a lower limit of 2.5) to improve detection sensitivity.

[0120] In yet another exemplary embodiment, the fusion analysis module 107 is further configured to collaboratively analyze the apparent anomaly feature data identified by the AI ​​video analysis module 106 with the fused displacement field data to generate multi-level anomaly alarm information. Specifically, the fusion analysis module 107 performs the following steps: (a) When the AI ​​video analysis module 106 detects an abnormal event and there is a valid displacement mutation point in the fused displacement field data, the fusion analysis module 107 generates a first-level alarm and sends the alarm information, fused displacement field data and video clips of the abnormal event to the remote monitoring center via the low-orbit communication module 108.

[0121] (b) When the AI ​​video analysis module 106 detects an abnormal event but no effective displacement mutation point is detected in the fused displacement field data, the fusion analysis module 107 generates a secondary warning, and at the same time controls the data processing module to trigger the encrypted monitoring mode, and sends the secondary warning information and video analysis results to the remote monitoring center through the low-orbit communication module 108.

[0122] (c) When the fused displacement field data detects a valid displacement abrupt change point but the AI ​​video analysis module 106 does not detect any obvious apparent anomalies, the fusion analysis module 107 generates a three-level prompt and controls the video acquisition module 105 to increase the frame rate or adjust the shooting angle to verify potential anomalies.

[0123] Through the above settings, this application organically combines GNSS and microseismic fusion displacement monitoring with AI video recognition and low-orbit communication technology, realizing comprehensive and multi-dimensional anomaly perception and efficient and reliable transmission from deep underground displacement, surface vibration to apparent deformation. This significantly improves the intelligence level and environmental adaptability of the slope monitoring system, and further reduces the false alarm rate and missed alarm rate.

[0124] (viii) The low-orbit communication module 108 is used to transmit the fused displacement field data, video image data and / or abnormal alarm information at the slope monitoring point to the remote monitoring center via a low-orbit satellite communication link.

[0125] The low-orbit communication module 108 enables reliable data transmission between the monitoring site and the remote monitoring center, especially in remote mountainous areas where terrestrial 4G / 5G network coverage is insufficient or communication is interrupted.

[0126] In a specific application example, the low-Earth orbit (LEO) communication module 108 includes a LEO satellite communication chip, a power amplifier, and a low-profile antenna. The LEO satellite communication chip is compatible with Starlink, OneWeb, or domestic LEO satellite communication protocols and supports bidirectional data transmission. The LEO communication module 108 is connected to the data processing module via a serial interface (such as UART or SPI).

[0127] Under normal operating conditions, the low-Earth orbit communication module 108 prioritizes data transmission via terrestrial networks (4G / NB-IoT / Ethernet); when the monitoring device detects that the terrestrial network signal strength is lower than a preset threshold or data transmission is interrupted, it automatically switches to the low-Earth orbit satellite communication link and sends at least one of the following data to the remote monitoring center: (1) The fused displacement field data at the slope monitoring point and its corresponding timestamp.

[0128] (2) Abnormal event types, confidence levels, and video clips of abnormal events identified by the AI ​​video analysis module 106.

[0129] (3) The key frame images or compressed video streams acquired by the video acquisition module 105.

[0130] (4) The device’s own operating status information (such as power, temperature, sensor self-test status, etc.).

[0131] Meanwhile, the low-orbit communication module 108 is also used to receive instructions from the remote monitoring center, including but not limited to: adjusting the data sampling frequency, switching the working mode, remotely upgrading the firmware, and requesting real-time video transmission. The data processing module responds to the instructions and executes the corresponding operations. The low-orbit communication module 108 ensures that the slope displacement monitoring device can still achieve real-time data transmission and remote control in areas without ground network coverage, significantly improving the environmental adaptability and reliability of the monitoring system.

[0132] This application effectively overcomes the technical shortcomings of single monitoring methods, such as data isolation, response lag, and high false alarm rate, by deeply integrating GNSS high-precision absolute displacement, three-axis microseismic acceleration, and tilt attitude sensing. It achieves a full-dimensional characterization of the process from static deformation to dynamic instability. Based on this, an AI video analysis module is introduced to perform target detection and motion recognition for apparent anomalies such as surface cracks, rockfalls, and slope slippage. This allows for cross-verification between the fused displacement field data and visual appearance features. It can enhance alarm reliability through video confirmation when sudden displacement changes occur, and provide early warnings when apparent anomalies first appear before significant displacement changes, significantly reducing the risk of missed alarms and false alarms. Meanwhile, a low-Earth orbit communication module was added to enable efficient and reliable transmission of fused displacement field data, abnormal alarm information, and video clips in remote mountainous areas with weak or interrupted 4G / 5G network coverage via low-Earth orbit satellite links. It also supports remote command issuance and adaptive equipment control. Combined with the device's built-in edge intelligent processing and dynamic switching capabilities of working modes, a highly reliable, all-weather slope displacement monitoring system integrating multi-source sensing, visual verification, intelligent decision-making, and full-domain communication was finally built, which greatly improved the environmental adaptability and intelligent operation and maintenance level of the monitoring system.

[0133] In summary, compared with the prior art, this application has at least the following beneficial effects: (1) Deep data fusion: GNSS absolute displacement monitoring is integrated with triaxial vibration acceleration monitoring and tilt angle monitoring to achieve deep fusion analysis of multi-source data, which overcomes the defect of data isolation in traditional single monitoring methods and comprehensively reflects the stability status of slope.

[0134] (2) Precursor sensing capability: The three-axis accelerometer 102 can sense the micro-seismic signals of the slope in real time, and can capture the instability precursor information before the slope undergoes significant displacement, thus achieving early warning.

[0135] (3) Adaptive working mode: Based on microseismic data, the absolute displacement data is subjected to sudden change detection, and the working mode of the geodetic GNSS receiver 101 is adaptively adjusted according to the detection results. While ensuring monitoring accuracy, the overall power consumption is effectively reduced, making it more suitable for long-term continuous operation in the field.

[0136] (4) Low false alarm rate: Through joint verification and fusion analysis of absolute displacement data and micro-vibration data, abnormal data caused by environmental interference are effectively filtered out, and the false alarm rate is greatly reduced.

[0137] (5) High integration: The triaxial accelerometer 102 and the tilt sensor 103 are integrated into the geodetic GNSS receiver 101, realizing an integrated design, compact structure, convenient installation, and reducing the workload of on-site deployment.

[0138] (6) Front-end intelligent processing: Equipped with an industrial-grade 32-bit ARM core processor, with built-in algorithms such as stellar-sun filtering, Kalman filtering, and mutation detection, preprocessing can be completed at the data acquisition front end, reducing the computing pressure on the cloud.

[0139] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0140] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A slope displacement monitoring device based on GNSS and microseismic fusion, characterized in that, include: Geodetic GNSS receivers are used to receive signals from satellite positioning and navigation systems and acquire absolute displacement data of slope monitoring points; A triaxial accelerometer is integrated inside the geodesic GNSS receiver to collect triaxial vibration acceleration data of slope monitoring points in real time. The tilt sensor, integrated inside the geodetic GNSS receiver, is used to collect tilt data from slope monitoring points in real time. The data processing module is used to perform filtering and correction processing on the absolute displacement data, the triaxial vibration acceleration data and the tilt angle data respectively, and generate a fused dataset; The video acquisition module is used to acquire video image data at slope monitoring points in real time. The AI ​​video analysis module is used to perform target detection and motion recognition processing on the video image data and extract apparent anomaly feature data at the slope monitoring point; The fusion analysis module is used to fuse the absolute displacement data, triaxial vibration acceleration data and tilt angle data in the fusion dataset and perform displacement state analysis to obtain fused displacement field data at the slope monitoring point. The module also performs spatiotemporal correlation analysis between the apparent anomaly feature data and the fused displacement field data to identify the abnormal state at the slope monitoring point and generate anomaly alarm information when an abnormal state is identified. The low-Earth orbit (LEO) communication module is used to transmit fused displacement field data, video image data, and / or abnormal alarm information from the slope monitoring point to a remote monitoring center via a LEO satellite communication link; when the ground network signal strength is detected to be lower than a preset threshold or data transmission is interrupted, it automatically switches to the LEO satellite communication link.

2. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 1, characterized in that, The data processing module includes: The displacement processing unit is used to sequentially perform sidereal filtering, Kalman filtering, cycle slip detection and repair processing on the absolute displacement data to obtain preprocessed absolute displacement data. An acceleration processing unit is used to sequentially perform zero bias correction, digital filtering, temperature drift compensation and gravity acceleration separation processing on the triaxial vibration acceleration data to obtain pre-processed triaxial vibration acceleration data. The tilt angle processing unit is used to perform zero-point correction and low-pass filtering on the tilt angle data in sequence, and to perform cross-validation and tilt angle deviation correction on the low-pass filtered tilt angle data based on the pre-processed triaxial vibration acceleration data to obtain the pre-processed tilt angle data. The fusion detection unit is used to generate a fusion dataset based on the preprocessed absolute displacement data, preprocessed triaxial vibration acceleration data, and preprocessed tilt angle data.

3. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 2, characterized in that, The data processing module further includes: The time synchronization unit is used to match the time stamps of the absolute displacement data and the tilt angle data with the time stamp of the absolute displacement data as a reference, and to unify the absolute displacement data, the triaxial vibration acceleration data and the tilt angle data to the same time reference by using an interpolation alignment method. The format unification unit is used to unify the format of time-aligned absolute displacement data, triaxial vibration acceleration data, and tilt angle data.

4. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 1, characterized in that, The fusion analysis module includes: The displacement mutation screening unit is used to calculate the displacement change rate of the absolute displacement data within the current window, and uses a dynamic adaptive threshold as the mutation discrimination threshold. The epoch segment that continuously exceeds the dynamic adaptive threshold is marked as a candidate displacement mutation time period, and candidate mutation time points are generated. The microseismic feature parameter extraction unit is used to construct an analysis time window centered on the candidate abrupt change time point, and extract the vibration feature parameters of the triaxial vibration acceleration data within the analysis time window. An effective mutation determination unit is used to confirm the effective mutation of the candidate mutation time point based on the relationship between the vibration characteristic parameters and the preset threshold, and obtain the effective displacement mutation point. The spatial consistency verification unit is used to perform spatial consistency verification on absolute displacement data using tilt angle data, and to determine the confidence level of valid displacement abrupt points. The displacement field determination unit is used to generate fused displacement field data at the slope monitoring point based on the fused dataset, the effective displacement abrupt point and the confidence level of the effective displacement abrupt point; The spatiotemporal analysis unit is used to perform spatiotemporal correlation analysis on the apparent anomaly feature data and the fused displacement field data to identify abnormal states at the slope monitoring points and generate anomaly alarm information when an abnormal state is identified.

5. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 4, characterized in that, The vibration characteristic parameters include short-time vibration energy, dominant vibration frequency, and vibration velocity attenuation coefficient. The short-time vibration energy is ;in, It is short-term vibrational energy. Sampling points i Place x The acceleration value of the shaft. Sampling points i Place y The acceleration value of the shaft. Sampling points i Place z The acceleration value of the shaft. The sampling time interval, p To determine the starting sampling point of the analysis time window, q This is the termination sampling point of the analysis time window; The dominant vibration frequency is obtained by performing a fast Fourier transform on the triaxial vibration acceleration data within the analysis time window; The vibration velocity attenuation coefficient is obtained by exponentially fitting the triaxial vibration acceleration data within the analysis time window.

6. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 5, characterized in that, The effective mutation determination unit uses a comprehensive voting decision rule to confirm the effective mutations at the candidate mutation time points; The comprehensive voting decision rule is as follows: when the ratio of short-term vibration energy to the average vibration energy in the preceding stable period exceeds the first energy multiple threshold, and the drift of the vibration main frequency exceeds the preset main frequency drift threshold or the vibration velocity attenuation coefficient is lower than the preset attenuation threshold, the candidate mutation time point is determined as an effective displacement mutation point. When the ratio of short-term vibration energy to the average vibration energy during the preceding stable period exceeds the first energy multiple threshold, but the drift of the dominant vibration frequency exceeds the preset dominant frequency drift threshold and the vibration velocity attenuation coefficient is lower than the preset attenuation threshold, the candidate mutation time point is marked as a suspected interference mutation point.

7. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 4, characterized in that, The spatial consistency verification unit includes: The similarity calculation subunit is used to extract the tilt angle change vector and the absolute displacement change vector within a preset time window before and after the effective displacement abrupt point, and to calculate the cosine similarity between the tilt angle change vector and the absolute displacement change vector. The confidence level subunit is used to set the confidence level of the effective displacement abrupt change point to a first confidence level when the cosine similarity is greater than a first preset similarity threshold, set the confidence level of the effective displacement abrupt change point to a second confidence level when the cosine similarity is between the first preset similarity threshold and a second preset similarity threshold, and set the confidence level of the effective displacement abrupt change point to a third confidence level when the cosine similarity is less than or equal to the second preset similarity threshold; wherein, the first preset similarity threshold is greater than the second preset similarity threshold.

8. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 4, characterized in that, The displacement field determination unit includes: The initial weight determination subunit is used to evaluate the confidence of the absolute displacement data, triaxial vibration acceleration data and tilt angle data in the fused dataset, and determine the initial values ​​of the displacement confidence weight, acceleration confidence weight and tilt angle confidence weight. The weight normalization subunit is used to dynamically adjust and normalize the initial values ​​of the displacement confidence weight, the initial value of the acceleration confidence weight, and the initial value of the tilt angle confidence weight according to the effective displacement abrupt change point and the confidence level of the effective displacement abrupt change point, respectively, to obtain normalized displacement confidence weight, normalized acceleration confidence weight, and normalized tilt angle confidence weight. An acceleration estimation subunit is used to perform a second integration on the dynamic acceleration components of the triaxial vibration acceleration data to obtain the acceleration integral displacement, and to use the tilt angle data to perform adaptive drift compensation on the acceleration integral displacement to obtain the acceleration estimation displacement. Inclination calculation subunit, used to calculate the inclination displacement based on the inclination data; The filtering and fusion subunit is used to perform complementary filtering and fusion of the absolute displacement data and the acceleration-calculated displacement to obtain a complementary filtering result. The weighted summation subunit is used to perform a weighted summation of the absolute displacement data, the acceleration-calculated displacement, and the tilt-calculated displacement based on the normalized displacement confidence weight, the normalized acceleration confidence weight, and the normalized tilt confidence weight, to obtain a preliminary fused displacement field. The secondary fusion subunit is used to perform secondary fusion of the preliminary fused displacement field and the complementary filtering result to obtain the fused displacement field data at the slope monitoring point.

9. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 4, characterized in that, The fusion analysis module also includes: The adaptive adjustment unit is used to adaptively adjust the observation working mode of the geodesic GNSS receiver based on the effective displacement abrupt change point and the confidence level of the effective displacement abrupt change point.

10. The slope displacement monitoring device based on GNSS and microseismic fusion according to claim 9, characterized in that, The observation modes of the geodetic GNSS receiver include conventional monitoring mode, encrypted monitoring mode, and low-power standby mode. The conventional monitoring mode is as follows: when an effective displacement mutation point is detected and the confidence level of the effective displacement mutation point is within a safe range and continues for a preset time, the geodetic GNSS receiver operates at a first sampling frequency, and the triaxial accelerometer and the tilt sensor operate at a second sampling frequency. The encrypted monitoring mode is as follows: when a valid displacement mutation point is detected and the confidence level of the valid displacement mutation point is not within the safe range, the geodetic GNSS receiver switches to the third sampling frequency, and the triaxial accelerometer and the tilt sensor switch to the fourth sampling frequency. The low-power standby mode is as follows: when no effective displacement change point is detected, the geodesic GNSS receiver switches to the fifth sampling frequency to operate. Wherein, the third sampling frequency is greater than the first sampling frequency, the fourth sampling frequency is greater than the second sampling frequency, and the fifth sampling frequency is less than the first sampling frequency.

Images