Dangerous rock monitoring and early warning system based on multi-source information fusion and preparation method

Abstract

The application is a dangerous rock monitoring and early warning system based on multi-source information fusion and a preparation method, belonging to the technical field of geological disaster monitoring. The system comprises a multi-physical field monitoring unit, a multi-source data acquisition and synchronization unit, a collaborative analysis and early warning unit, and an alarm unit. The multi-physical field monitoring unit integrates infrasound, microseismic, mechanical, motion posture and video monitoring modules, and realizes millisecond-level synchronous acquisition through a unified clock source data acquisition device. The preparation method comprises the steps of field survey, waveguide pre-burial, sensor installation, synchronous debugging, background noise acquisition and trial operation. The early warning method is based on the quantitative criterion of time domain sudden increase, frequency domain migration, energy release and rupture mode change constructed by the collaborative response of infrasound-microseismic, realizing the whole process division and hierarchical early warning of dangerous rock instability. The application realizes the accurate synchronization and collaborative analysis of multi-physical field signals, improves the accuracy and reliability of the monitoring and early warning of falling dangerous rocks, and is suitable for unattended monitoring in complex terrain.

Description

Technical Field

[0001] This invention relates to a rockfall monitoring and early warning system based on multi-source information fusion and its preparation method, belonging to the field of geological disaster monitoring and early warning technology, and is particularly applicable to a rockfall monitoring and early warning system based on multi-source information fusion and the method for preparing the physical model of the system. Background Technology

[0002] Falling rockfalls are among the most typical and dangerous types of geological hazards in high-risk areas such as the Three Gorges Reservoir area. Their instability is characterized by suddenness, wide-ranging impact, and concentrated energy. Physical model testing is a crucial method for studying the instability mechanism of falling rockfalls. By simulating geological conditions, loads, rainfall, and other triggering factors, the characteristic evolution process of rockfall deformation and failure can be realistically reproduced. Therefore, conducting physical model testing research on falling rockfalls has significant theoretical value and practical engineering implications for revealing their instability mechanism and establishing effective early warning methods.

[0003] Traditional monitoring methods mainly rely on macroscopic mechanical parameters such as stress and displacement. While these methods can reflect the overall stability of unstable rock masses, they struggle to capture early signals of internal microscopic damage evolution before instability, resulting in insufficient timeliness and accuracy in early warning. In recent years, simulation experiments and monitoring and early warning technologies for rockfall have seen some development. Chinese patent application CN121740736A discloses a simulation test system and method for rockfall degradation and collapse under multi-field coupling. Through a flexible stress loading device, corrosion device, and acoustic emission measurement system, it simulates the rockfall degradation process under the synergistic effects of stress field, seepage field, and chemical field, and calculates the degree of damage based on acoustic emission characteristics. Chinese patent application CN113724480A discloses a monitoring and early warning system for the impact of high-speed rail operation on ultra-high and steep unstable rock masses above tunnel entrances. It uses distributed fiber optic acoustic wave sensing technology to collect vibration and acoustic wave signals, and combines this with radio frequency identification (RFID) technology for displacement monitoring. However, existing technologies still have the following shortcomings: First, most experimental systems can only simulate stress or chemical corrosion in isolation, lacking the coordinated acquisition and analysis of infrasound (<20 Hz) and microseismic (0.1~100 Hz) signals. Infrasound and microseismic signals, as physical signals directly radiated during rock fracture, propagate through air and solid media respectively, and have unique advantages in capturing fracture incubation information. Second, existing data acquisition systems are mostly limited to monitoring single physical quantities (such as deformation and vibration amplitude), lacking a millisecond-level synchronous acquisition architecture for multiple physical fields (infrasound, microseismic, stress, displacement, and attitude), resulting in insufficient time-series alignment accuracy of multi-source data and difficulty in establishing accurate coordinated response relationships. Third, existing early warning criteria are mostly based on single signal sources or simple threshold comparisons, without establishing multi-dimensional quantitative criteria based on the time domain, frequency domain, energy domain, and fracture mode transformation of infrasound-microseismic coordinated response, thus the accuracy and reliability of early warning need to be improved.

[0004] Therefore, there is an urgent need for a rockfall monitoring and early warning system that can simultaneously collect multi-source data such as infrasound, microseismic, stress, displacement and video, and establish a multi-physics field collaborative response relationship. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a rockfall monitoring and early warning system and preparation method based on multi-source information fusion. This system realizes multi-source information collaborative response analysis of the entire process of falling rockfall from stable compression, damage accumulation, critical instability to dynamic instability by constructing a multi-physics field synchronous monitoring architecture, and establishes a highly reliable early warning criterion.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A rockfall monitoring and early warning system based on multi-source information fusion, combined with Figure 1 It is characterized by comprising a multi-physics field monitoring unit, a multi-source data acquisition and synchronization unit, a collaborative analysis and early warning unit, and an alarm unit.

[0008] The multiphysics monitoring unit is installed in the target unstable rock mass and its surrounding area to collect multiphysics response signals in real time throughout the entire process of rock mass instability.

[0009] Furthermore, the multiphysics monitoring unit includes:

[0010] The infrasound monitoring module includes three infrasound sensors and a rigid waveguide. The first sensor is connected to the interior of the unstable rock mass via the rigid waveguide, with the other end of the waveguide extending to the outside and connecting to the sensor. It is used to collect low-frequency sound waves (frequency <20 Hz) radiated into the air during the rock mass's fracturing process. The second and third sensors are positioned at a 45° angle to the horizontal on either side of the unstable rock mass to pick up infrasound signals radiated from the air.

[0011] Microseismic monitoring module: includes a microseismic sensor. The microseismic sensor is installed on the surface of the unstable rock mass or pre-embedded inside the unstable rock mass, and is used to pick up elastic wave signals (frequency 0.1~100 Hz) released by micro-fracture events inside the rock mass and propagated through the solid medium.

[0012] The mechanical monitoring module includes at least one field-applicable load sensor, resistance strain gauges deployed on the surface of the unstable rock mass, and fiber Bragg grating (FBG) bare grating strain sensors. The load sensor is installed in a load monitoring hole drilled near the top or rear edge fracture of the unstable rock mass, and connected to the anchored end at the bottom of the hole via a rigid connecting rod. It is used to monitor changes in tensile stress at the rear edge of the unstable rock mass or pressure in the locking section. The resistance strain gauges and FBG bare grating strain sensors are deployed at the tip of the rear edge fracture and near the locking section of the unstable rock mass to acquire strain evolution data of key parts of the unstable rock mass.

[0013] Motion attitude monitoring module: including a dual-axis tilt / accelerometer installed on the top or leading edge of the unstable rock mass, used to simultaneously measure the tilt angle change and triaxial acceleration of the unstable rock mass, in order to capture the attitude anomaly and instantaneous dynamic characteristics before instability.

[0014] Video monitoring module: Includes a high-resolution video camera, which is deployed in a safe area on the opposite side of the unstable rock mass to fully record the macroscopic damage morphology and crack propagation process of the unstable rock mass.

[0015] The multi-source data acquisition and synchronization unit includes a data acquisition device with a unified clock source, which is connected to the multi-physics field monitoring unit to realize the synchronous acquisition and timing alignment of multi-physics field data, with a timing alignment error of less than 1 millisecond.

[0016] Furthermore, the infrasound monitoring module, microseismic monitoring module, mechanical monitoring module, motion posture monitoring module, and video monitoring module are all connected to the data acquisition device of the multi-source data acquisition and synchronization unit through a signal acquisition instrument.

[0017] The collaborative analysis and early warning unit is a computer device that receives multi-source synchronized data output by the multi-source data acquisition and synchronization unit and performs the following analysis: (i) Dividing the entire process of unstable rock mass into a stable stage, a damage accumulation stage, a critical instability stage, and a dynamic instability stage; (ii) Constructing an early warning criterion based on infrasound-microseismic collaborative response, the early warning criterion including: time-domain collaborative sudden increase criterion, frequency-domain collaborative migration criterion, energy release collaborative criterion, and rupture mode transformation criterion; (iii) When one or more of the early warning criteria are triggered, outputting an unstable rock mass early warning signal.

[0018] The alarm unit is connected to the collaborative analysis and early warning unit and is used to receive early warning signals and perform alarm actions, including on-site audible and visual alarms and remote SMS alarms.

[0019] The method for preparing a rockfall monitoring and early warning system based on multi-source information fusion is characterized by the following steps:

[0020] Step S100: On-site Survey and Scheme Design: Conduct a geological survey of the target unstable rock mass to determine its geometric shape, distribution of the main controlling structural planes, location of potential locking sections, and surrounding stable bedrock areas. Based on the survey results, design the deployment locations of each module in the multiphysics monitoring unit, including: the burial locations of infrasound sensors and waveguides, the installation holes of microseismic sensors, the bonding locations of load sensors and strain gauges, the installation locations of dual-axis tilt / accelerometers, and the deployment locations of video cameras.

[0021] Step S200: Waveguide Pre-embedding and Sensor Installation Hole Preparation: During the physical model casting stage of the unstable rock mass, rigid waveguides are pre-cast into key locations inside the unstable rock mass, ensuring tight coupling between the waveguides and the rock mass material. The waveguide outlet extends to the outside of the model. Microseismic sensor installation holes are pre-reserved during model casting.

[0022] Step S300: Sensor installation and integration. Further, step S300 specifically includes:

[0023] S301: Install an infrasound sensor at the waveguide outlet to ensure a reliable seal at the connection and prevent interference from ambient airflow.

[0024] S302: Install the micro-vibration sensor in the reserved mounting hole, and apply coupling agent between the sensor and the hole wall to ensure signal transmission effect;

[0025] S303: Attach strain gauges to the tip of the crack at the rear edge of the unstable rock mass and near the locked section, and connect them to the strain acquisition instrument; attach FBG bare grid strain sensors to the surface of the locked section and key parts of the crack propagation path, and connect them to the fiber optic demodulator.

[0026] S304: Install borehole load sensors (such as anchor bolt axial force gauges) in monitoring holes near the top or rear edge cracks of the unstable rock mass and connect them to the data acquisition system;

[0027] S305: Rigidly install a biaxial tilt / accelerometer on the top or leading edge of the unstable rock mass;

[0028] S306: Deploy high-resolution video cameras in the safe area opposite the unstable rock mass, and adjust the shooting angle to fully cover the unstable rock mass.

[0029] Furthermore, to better install the sensors, the area where the strain gauges are pasted is ground smooth and surface dust is removed; a rigid base or expansion bolt fixing point is prefabricated at the installation location of the tilt / accelerometer to ensure rigid coupling between the sensor and the rock mass.

[0030] Step S400: System Connection and Synchronization Debugging: Connect the infrasound sensor, micro-vibration sensor, load sensor, strain gauge, dual-axis tilt / accelerometer, and video camera to the data acquisition device with a unified clock source via signal cables or a wireless transmission module. Verify the time synchronization accuracy of all monitoring channels using transient impact signals, ensuring that the timing alignment error is less than 1 millisecond. Set the sampling parameters of the data acquisition device, including sampling rate, acquisition duration, and triggering method.

[0031] Step S500: Background Noise Acquisition and Baseline Establishment: Start the monitoring system and acquire background noise and microseismic data for at least 1 minute as a baseline reference for subsequent signal processing. Record the environmental conditions during background acquisition (such as wind speed, temperature, human activities, etc.) for subsequent signal filtering and anomaly removal.

[0032] Step S600: System Integration and Trial Operation: Start all monitoring modules, check whether the signals of each channel are normal, and verify the integrity of the data acquisition, transmission, storage, and analysis process. Conduct a trial operation for no less than 24 hours. After confirming that the system stability and data quality meet the requirements, officially put it into monitoring and early warning operation.

[0033] Step S700: Continuous monitoring and data acquisition: The system enters continuous monitoring mode, and collects infrasound, micro-vibration, stress, strain, tilt angle, acceleration and video data in real time, and transmits them to the collaborative analysis and early warning unit.

[0034] A method for monitoring and early warning of dangerous rocks based on multi-source information fusion is characterized by the following steps:

[0035] Step S110: On-site survey and sensor deployment: Conduct geological surveys of the target unstable rock mass to determine its geometry, distribution of the main controlling structural planes, and location of the locking sections. Deploy infrasound sensors, microseismic sensors, load sensors, strain gauges, dual-axis tilt / accelerometers, and video cameras at key locations of the unstable rock mass and surrounding stable areas.

[0036] Step S210: System integration and synchronization debugging: Connect all sensors and cameras to a data acquisition device with a unified clock source, and verify the time synchronization accuracy of all monitoring channels through transient tapping signals to ensure that the timing alignment error is less than 1 millisecond.

[0037] Step S310: Background noise acquisition: In the initial stage of operation of the monitoring system, acquire background noise and micro-vibration data for no less than 1 minute as a baseline reference for subsequent signal processing.

[0038] Step S410: Continuous monitoring and data acquisition: Start the monitoring system to continuously acquire infrasound, microseismic, stress, strain, tilt angle, acceleration and video data, and transmit them to the collaborative analysis and early warning unit in real time.

[0039] Step S510: Multi-source data collaborative analysis and early warning: The collaborative analysis and early warning unit performs a full process division of unstable rock mass on the received multi-source synchronous data, constructs an early warning criterion based on infrasound-microseismic collaborative response, and identifies and outputs unstable rock mass early warning signals;

[0040] Furthermore, step S510 specifically includes:

[0041] S511: Calculate the short-time event rate of the infrasound signal and the event rate of the microseismic signal, and determine whether the time-domain coordinated sudden increase criterion is met;

[0042] S512: Extract the dominant frequency of the microseismic signal by power spectral density estimation, calculate the energy proportion of the infrasound signal in the 0.5~5 Hz frequency band, and determine whether the frequency domain cooperative migration criterion is met;

[0043] S513: Calculate the Pearson correlation coefficient between the square integral energy of the microseismic signal and the square root mean square amplitude of the infrasound signal, and determine whether the energy release coordination criterion is met;

[0044] S514: Analyze the changes in shear failure ratio based on the RA / AF values ​​of microseismic events to determine whether the failure mode transition criterion is met;

[0045] S515: When one or more of the above criteria are triggered, the unstable rock mass is determined to have entered the critical instability stage, and an early warning signal is output to the alarm unit.

[0046] Step S610: Alarm Execution: After receiving the warning signal, the alarm unit activates the on-site audible and visual alarm and sends the warning information to the preset management personnel via the SMS unit.

[0047] The beneficial effects of this invention are as follows: It provides a rockfall monitoring and early warning system and its preparation method based on multi-source information fusion. For the first time, it integrates infrasound, microseismic, mechanical, motion attitude, and video monitoring into one system, constructing a comprehensive monitoring architecture covering air-propagated and solid-conducted signals, which can completely capture the physical response of the entire process of rockfall instability. Through unified clock source and transient impact verification, it achieves millisecond-level time alignment of multi-source data, providing a high-quality data foundation for collaborative analysis in the time domain, frequency domain, and energy domain. Based on the collaborative response of infrasound and microseismic events, it proposes four quantitative criteria: time domain surge, frequency domain migration, energy coordination, and fracture mode transformation, effectively overcoming the limitations of single signal monitoring and significantly improving the accuracy and robustness of early warning. At the same time, this invention provides a complete on-site deployment and integration process, including waveguide pre-embedding, sensor installation, synchronous debugging, and baseline establishment, ensuring the reliability and repeatability of the system in complex on-site environments, and is suitable for long-term unattended monitoring under complex terrain conditions. Attached Figure Description

[0048] To make the objectives and technical solutions of this invention clearer, the following figures are provided for illustration:

[0049] Figure 1 This is a schematic diagram of the overall architecture of the dangerous rock monitoring and early warning system based on multi-source information fusion in this invention;

[0050] Figure 2 This is a flowchart of the preparation method of the dangerous rock monitoring and early warning system based on multi-source information fusion in this invention;

[0051] Figure 3 This is a flowchart of the rockfall monitoring and early warning method based on multi-source information fusion in this invention;

[0052] Figure 4 This is a schematic diagram of the field sensor deployment in Embodiment 1 of the present invention. Detailed Implementation

[0053] To make the objectives and technical solutions of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0054] Example 1: This example uses a typical collapsed rock mass in the Three Gorges Reservoir area as the application object, combined with... Figure 4 The target rock mass is located on a slope on the bank of a reservoir in the Three Gorges Reservoir area. The lithology is thick-layered limestone, with a height of approximately 18m, width of approximately 10m, thickness of approximately 6m, and a volume of approximately 1080m³. It is classified as a medium-sized rock mass prone to collapse. A main control structural plane with a dip angle of 78° and a dip direction consistent with the slope surface is developed at the rear edge of the rock mass. The locking section, located in the lower-middle part of the rear edge of the rock mass, is 2.5m long and 8m wide, and is the key part for controlling the stability of the rock mass. Behind the rock mass is stable parent rock, and both sides are open, exhibiting a typical "cantilever beam" structure. The vertical height difference between the top of the rock mass and the road below is approximately 45m, and the road below is a major traffic artery. Instability of this rock mass would directly threaten the safety of passing vehicles and pedestrians. To monitor and provide early warning for this rock mass, this embodiment provides a "rock mass monitoring and early warning system based on multi-source information fusion".

[0055] like Figure 1 As shown, the system described in this embodiment includes: a multi-physics monitoring unit (100), a multi-source data acquisition and synchronization unit (200), a collaborative analysis and early warning unit (300), and an alarm unit (400).

[0056] The multiphysics monitoring unit (100) is installed in the target unstable rock mass and its surrounding area to collect multiphysics response signals in real time throughout the entire process of rock mass instability. Specifically, it includes the following modules:

[0057] 1. The infrasound monitoring module (110) includes three infrasound sensors (111-1, 111-2, 111-3) and a rigid waveguide (112). The waveguide (112) is a galvanized steel pipe with an inner diameter of 20 mm and a wall thickness of 3 mm. One end of the waveguide (112) is buried at the bottom of the main control fracture at the rear edge of the unstable rock mass at a depth of 1.2 m. The waveguide (112) is sealed with cement mortar to ensure effective transmission of the sound wave signal. The other end of the waveguide (112) extends to about 0.3 m outside the fracture and is connected to the first acoustic sensor (111-1). The second acoustic sensor (111-2) and the third acoustic sensor (111-3) are placed at an angle of 45° to the horizontal plane on both sides in front of the unstable rock mass, about 1 m away from the unstable rock mass. The infrasound sensors (111-1~111-3) all use capacitive infrasound sensors with a frequency response range of 0.5~200Hz and a sensitivity of 50mV / Pa. They are used to collect low-frequency sound wave signals (frequency <20Hz) radiated into the air during the fracturing of unstable rock masses.

[0058] 2. Microseismic monitoring module (120), including one microseismic sensor (121). The microseismic sensor (121) adopts a three-component intelligent seismic sensor (such as the SmartSolo series), with a frequency response range of 0.2Hz~1000Hz, a sensitivity of 1.25V / g, and a sampling rate set to 500Hz. The microseismic sensor (121) is installed at the center of the top of the unstable rock mass. Before installation, a mounting hole with a diameter of 50mm and a depth of 80mm is drilled, and coupling agent is applied between the sensor and the hole wall to ensure signal transmission.

[0059] 3. Mechanical monitoring module (130), including a borehole load sensor (131), a resistance strain gauge (132) and an FBG bare grid strain sensor (133).

[0060] The borehole load sensor (131) is a vibrating wire anchor stress gauge, installed at the center of the locking section, with a drilling depth of 2.0m, used to monitor the stress state changes of the locking section in real time.

[0061] Six resistance strain gauges (132) were installed, model BF350-3AA (resistance 350Ω, sensitivity coefficient 2.0). The locations were: one on each side of the tip of the main control fracture (2 in total), one each at the upper, middle, and lower parts of the locking section (3 in total), and one at the root of the unstable rock mass. Before attaching the strain gauges, the area to be attached was ground smooth, surface dust was removed, and 502 glue was used for attachment.

[0062] Three FBG bare grating strain sensors (133) are deployed with center wavelengths of 1528nm, 1538nm and 1548nm respectively. They are deployed on the surface of the locking section and in key parts of the main control crack propagation path. Strain data are collected by an optical fiber demodulator.

[0063] 4. Motion attitude monitoring module (140), including a dual-axis tilt / accelerometer (141), which adopts a MEMS capacitive dual-axis tilt / accelerometer with a tilt range of ±90°, static accuracy of ±0.01°, dynamic accuracy of ±0.05°, and acceleration range of ±16g. The dual-axis tilt / accelerometer (141) is rigidly installed at the center of the top of the unstable rock mass. A rigid base is prefabricated before installation and fixed with expansion bolts. It is used to simultaneously measure the pitch angle, roll angle, and triaxial acceleration of the unstable rock mass.

[0064] The multi-source data acquisition and synchronization unit (200) includes a data acquisition device (201) with a unified clock source. The data acquisition device (201) adopts an industrial-grade data acquisition terminal, with a built-in GPS / BeiDou dual-mode timing module, providing a unified clock source with an accuracy of ±50ns.

[0065] The infrasound sensor (111-1~111-3), micro-vibration sensor (121), load sensor (131), resistance strain gauge (132), FBG bare grid strain sensor (133), dual-axis tilt / accelerometer (141) and video camera (151) are all connected to the data acquisition device (201) through their respective signal acquisition instruments. The specific connection methods are as follows: the infrasound sensor (111-1~111-3) is connected to the data acquisition device (201) through the BNC interface; the micro-vibration sensor (121) is connected to the data acquisition device (201) through the SmartSolo data recovery unit; the load sensor (131) is connected to the data acquisition device (201) through the vibrating wire acquisition instrument; the resistance strain gauge (132) is connected to the data acquisition device (201) through the static / dynamic strain acquisition instrument; the FBG bare grid strain sensor (133) is connected to the data acquisition device (201) through the fiber optic demodulator; the dual-axis tilt / accelerometer (141) is connected to the data acquisition device (201) through the RS485 interface; and the video camera (151) is connected to the data acquisition device (201) through the network switch.

[0066] The data acquisition device (201) is configured with the following sampling rates: infrasound sensor 200Hz, micro-vibration sensor 500Hz, load sensor 50Hz, resistance strain gauge 50Hz, FBG bare grid strain sensor 100Hz, dual-axis tilt / accelerometer 100Hz, and video camera 30fps. All channels achieve synchronous acquisition and timing alignment. Verification using transient impact signals shows that the timing alignment error is less than 0.5ms.

[0067] The collaborative analysis and early warning unit (300) is a computer device (301) equipped with collaborative analysis and early warning software. It is connected to the data acquisition device (201) via a network and receives multi-source synchronous data. The hardware configuration of the computer device (301) is as follows: Intel Core i7 processor, 32GB memory, 1TB SSD storage, and NVIDIA RTX 3060 graphics card (for video image processing).

[0068] The collaborative analysis and early warning unit (300) performs the following analysis: (i) it divides the entire process of unstable rock mass into a stable stage, a damage accumulation stage, a critical instability stage, and a dynamic instability stage; (ii) it constructs early warning criteria based on infrasound-microseismic collaborative response, including time-domain collaborative surge criteria, frequency-domain collaborative migration criteria, energy release collaborative criteria, and rupture mode transformation criteria; (iii) when one or more of the early warning criteria are triggered, it outputs an early warning signal for unstable rock mass.

[0069] The alarm unit (400) is connected to the collaborative analysis and early warning unit (300) and includes a field audible and visual alarm (401) and an SMS unit (402). The field audible and visual alarm (401) is installed at the highway entrance below the unstable rock mass, has a power of 100W, and features a red flashing light and a 120dB alarm sound. The SMS unit (402) has a built-in 4G communication module and pre-stores the mobile phone numbers of 5 management personnel.

[0070] Example 2: Based on the scenario and system of Example 1, this example provides a "method for preparing a rockfall monitoring and early warning system based on multi-source information fusion." For example... Figure 2 The process shown includes the following steps:

[0071] Step S100: On-site survey and scheme design.

[0072] A geological survey was conducted on the target unstable rock mass using a combination of UAV aerial photography and manual reconnaissance. The survey determined that the rock mass is approximately 18m high, 10m wide, and 6m thick, with a main controlling structural plane dip angle of 78°. The potential locking section is located in the lower-middle rear edge of the unstable rock mass, measuring 2.5m in length and 8m in width. The surrounding stable bedrock is located approximately 5m behind the unstable rock mass. Based on the survey results, the deployment locations of each module in the multiphysics monitoring unit were designed as follows: the waveguide tube of the infrasound sensor was buried at the bottom of the main controlling fracture at a depth of 1.2m; two other infrasound sensors were placed at a 45° angle on both sides in front of the unstable rock mass; the microseismic sensor was deployed at the center of the top of the unstable rock mass; six resistance strain gauges and three FBG bare grid strain sensors were deployed at the tip of the main controlling fracture and the locking section; a dual-axis tilt / accelerometer was installed at the center of the top of the unstable rock mass; and a video camera was deployed 60m to the opposite side of the unstable rock mass.

[0073] Step S200: Waveguide pre-embedding and sensor mounting hole preparation.

[0074] A hole with a diameter of 25 mm and a depth of 1.2 m was drilled at the bottom of the main control fracture of the target unstable rock mass. A rigid waveguide (112) with an inner diameter of 20 mm was embedded in the hole. The space between the waveguide (112) and the hole wall was filled with cement mortar to ensure a tight seal. The outlet of the waveguide extended to about 0.3 m outside the fracture. A microseismic sensor mounting hole with a diameter of 50 mm and a depth of 80 mm was drilled at the center of the top of the unstable rock mass.

[0075] Step S300: Sensor installation and integration.

[0076] S301: Install the first acoustic sensor (111-1) at the outlet of the waveguide; install the second acoustic sensor (111-2) and the third acoustic sensor (111-3) at a 45° angle on both sides of the rock mass in front of the rock mass, about 1m away from the rock mass. The sensors should be pointed in the direction of the rock mass. Apply sealant to the interface to ensure a reliable seal.

[0077] S302: Install the micro-vibration sensor (121) in the reserved mounting hole, and apply coupling agent between the sensor and the hole wall to ensure signal transmission effect.

[0078] S303: Six resistance strain gauges (132) are attached to both sides of the tip of the main control crack and the surface of the locking section using 502 glue. Before attaching, the area to be attached is ground smooth and the surface dust is removed. Three FBG bare grid strain sensors (133) are attached to the surface of the locking section and the path of the main control crack and connected to the fiber optic demodulator.

[0079] S304: Install a borehole load sensor (131) in the monitoring hole behind the crack at the rear edge of the top of the unstable rock mass and connect it to the data acquisition device (201) to monitor the change of tensile stress at the rear edge of the unstable rock mass in real time.

[0080] S305: A biaxial tilt / accelerometer (141) is rigidly installed at the center of the top of the unstable rock mass. A rigid base is prefabricated before installation and fixed with expansion bolts.

[0081] S306: Deploy a high-resolution video camera (151) in a safe area 60m away from the dangerous rock mass, and adjust the focus and shooting angle to ensure complete coverage of the dangerous rock mass.

[0082] Step S400: System connection and synchronization debugging.

[0083] The infrasound sensors (111-1~111-3), micro-vibration sensors (121), load sensors (131), resistance strain gauges (132), FBG bare grid strain sensors (133), dual-axis tilt / accelerometers (141), and video cameras (151) are connected to a data acquisition device (201) with GPS / BeiDou dual-mode timing via signal lines or wireless transmission modules. The time synchronization accuracy of all monitoring channels is verified by transient impact signals. The timing alignment error is less than 0.5ms, meeting the requirement of less than 1ms. The sampling parameters of the data acquisition device are set as follows: infrasound 200Hz, micro-vibration 500Hz, load sensor 50Hz, strain gauge 50Hz, FBG 100Hz, tilt 100Hz, and video 30fps. The acquisition mode is continuous acquisition.

[0084] Step S500: Background noise acquisition and baseline establishment.

[0085] The monitoring system was activated, and background data was continuously collected for 2 minutes under undisturbed conditions (nighttime, no wind, no construction). The results showed: the root mean square value of infrasound background was 0.03 Pa, with the main source of ambient infrasound being a light breeze; the root mean square value of microseismic background was 0.01 m / s², with the main source of ambient vibration being distant highway traffic. Recording environmental conditions: wind speed 1.2 m / s, temperature 18℃, no rainfall. The above background data served as a baseline reference for subsequent signal processing.

[0086] Step S600: System integration and trial operation.

[0087] Start all monitoring modules, check the signal strength of each channel, and verify the integrity of the data acquisition, transmission, storage, and analysis process. After 72 hours of continuous trial operation, the data acquisition integrity rate reached 99.5%, the signal quality met the analysis requirements, and the system was confirmed to be stable before being put into formal monitoring and early warning operation.

[0088] Step S700: Continuous monitoring and data acquisition.

[0089] The system enters continuous monitoring mode, collecting infrasound, microseismic, stress, strain, tilt angle, acceleration and video data in real time, and transmitting them to the collaborative analysis and early warning unit (300) via 4G network.

[0090] Example 3: For the scenario described in Example 1, this example provides a "method for monitoring and early warning of dangerous rocks based on multi-source information fusion." (The example is repeated in the original text.) Figure 3 The process shown includes the following steps:

[0091] Step S110: On-site survey and sensor deployment.

[0092] The on-site survey and sensor deployment are completed according to steps S100 to S300 of Example 2, which will not be repeated here.

[0093] Step S210: System integration and synchronous debugging.

[0094] The system integration and synchronous debugging are completed according to step S400 of Example 2, which will not be repeated here.

[0095] Step S310: Background noise acquisition.

[0096] The background noise acquisition and baseline establishment are completed according to step S500 of Example 2, which will not be repeated here.

[0097] Step S410: Continuous monitoring and data acquisition.

[0098] The system enters continuous monitoring mode, continuously collecting infrasound, microseismic, stress, strain, tilt angle, acceleration, and video data, and transmitting them in real time to the collaborative analysis and early warning unit (300).

[0099] Step S510: Collaborative analysis and early warning of multi-source data.

[0100] The computer device (301) in the collaborative analysis and early warning unit (300) runs collaborative analysis and early warning software and performs the following analysis process on the received multi-source synchronous data:

[0101] (I) Division of the entire process of unstable rock mass:

[0102] The collaborative analysis and early warning software has a built-in four-stage division module. Based on the comprehensive variation characteristics of mechanical parameters (stress, strain) and rupture signals (infrasound event rate, microseismic event rate), it dynamically divides the entire process of unstable rock mass into the following four stages:

[0103] Stable phase: Stress increases slowly with monitoring time, strain increases linearly, and the infrasound and microseismic event rates are at background noise levels (infrasound event rate < 5 times / minute, microseismic event rate < 2 times / minute). During this phase, the rock mass is in a state of linear elastic deformation.

[0104] Damage accumulation stage: The stress-strain curve begins to deviate from linearity, the stiffness begins to deteriorate initially, the infrasound event rate rises to 10~30 times / minute, the microseismic event rate rises to 5~15 times / minute, and the dominant frequency of microseismic events begins to shift from the initial 80~120Hz to 50~80Hz.

[0105] Critical instability stage: Stress approaches its peak value, strain increases rapidly, the infrasound event rate surges to over 50 events / minute, the microseismic event rate surges to over 25 events / minute, the dominant frequency of microseismic events drops to below 70% of its initial value (i.e., <56Hz), and the proportion of low-frequency infrasound energy (0.5~5Hz) exceeds 50%. This stage is the critical time window for issuing early warnings.

[0106] Dynamic instability stage: Stress drops sharply, displacement changes abruptly, infrasound and microseismic signals burst into high-energy pulses, and the unstable rock mass falls macroscopically.

[0107] (II) Early warning criteria based on infrasound-microseismic coordinated response

[0108] The collaborative analysis and early warning software has four built-in early warning criterion modules. The specific algorithms and thresholds for each criterion are as follows:

[0109] Criterion 1: Temporal Co-increase Criterion:

[0110] Calculate time period Short-time event rate of infrasound signals within A sliding time window (60 seconds window length, 10-second step) is used to count the number of valid events exceeding the threshold (3 times the background root mean square value, i.e., 0.09 Pa) within the window. If the mean value within the sliding time window exceeds three times the standard deviation of the background mean value (5 times / minute) during the stable phase (i.e., ≥15 times / minute), and the duration of this state is not less than 5 seconds, it is judged as a sudden increase in the infrasound event rate.

[0111] Meanwhile, the event rate of microseismic signals was examined within a ±10-second time window of the sudden increase in infrasound event rate. The microseismic event rate must simultaneously exhibit a significant increase exceeding twice the standard deviation of its background value (2 events / minute), i.e., ≥6 events / minute. When the two are highly overlapping in time (time difference ≤2 seconds), the temporal domain co-increase criterion is triggered.

[0112] Criterion 2: Frequency Domain Co-transfer Criterion:

[0113] The instantaneous dominant frequency of the microseismic signal was extracted by power spectral density estimation (using the Welch method, with a window length of 1024 points and an overlap rate of 50%). .when When the stiffness of the rock mass continues to decrease to below 70% of its initial characteristic value (measured at 95 Hz in the stable phase) (i.e., <66.5 Hz), it indicates a significant decrease in the overall stiffness of the rock mass.

[0114] Simultaneously, the energy of the infrasound signal in the 0.5~5Hz frequency band was calculated. With total signal energy ratio .when If the increase exceeds 20% within 60 seconds and its absolute value exceeds 50%, it is determined to be a sudden increase in the proportion of infrasound low-frequency energy.

[0115] When the two frequency domain metrics mentioned above are triggered simultaneously within 1 minute, the frequency domain collaborative migration criterion is triggered.

[0116] Criterion 3: Energy Release Coordination Criterion:

[0117] Calculate microseismic signals square integral energy and infrasound signals root mean square amplitude The sliding time window length ΔT = 10 seconds.

[0118] Calculate the Pearson correlation coefficient between two energy series within a sliding window. When the correlation coefficient If the state lasts for more than 10 seconds, it is determined that the energy release phase has been entered, triggering the energy release coordination criterion.

[0119] Criterion 4: Criterion for Change in Failure Mode:

[0120] Based on the waveform parameters of microseismic events, the RA value (rise time / amplitude) and AF value (average frequency = ring count / duration) are calculated for each event. In the RA-AF two-dimensional scatter plot, the K-means clustering algorithm is used. The events are divided into tension-dominated (high AF, low RA) and shear-dominated (low AF, high RA).

[0121] Define the shear fracture ratio This represents the proportion of microseismic events classified as shear-type ruptures within the analysis period (sliding window length of 2 minutes). When the rate rapidly increases from below 30% to above 60% within 90 seconds, it indicates a shift in the rupture mechanism from tension-dominated to shear-dominated, triggering the rupture mode transition criterion.

[0122] (III) Outputting early warning signals for unstable rock formations

[0123] The collaborative analysis and early warning software adopts a multi-criteria fusion decision-making strategy: when any one of the above four criteria is triggered for the first time, it is determined that the unstable rock mass has entered the critical instability stage and outputs a yellow warning signal (attention level); when two or more criteria are triggered at the same time and the duration exceeds 30 seconds, it is determined that the unstable rock mass is about to experience dynamic instability and outputs a red warning signal (alarm level).

[0124] Step S610: Alarm execution.

[0125] After receiving the warning signal, the alarm unit (400) performs the following actions respectively:

[0126] Yellow Alert: SMS unit (402) sends an alert SMS to management personnel, which reads: "[Yellow Alert for Dangerous Rock Mass] The XX dangerous rock mass in the Three Gorges Reservoir area triggered the temporal collaborative surge criterion at 10:23:15 on April 15, 2026. Please strengthen patrols and pay attention."

[0127] Red Alert: The SMS unit (402) sends an alert SMS to the management personnel. At the same time, the on-site audible and visual alarm (401) is activated, emitting a red flashing light and a 120dB alarm sound. The SMS content is: "[Dangerous Rock Red Alert] The XX dangerous rock mass in the Three Gorges Reservoir area triggered a multi-criteria fusion warning (time domain + frequency domain) at 10:25:30 on April 15, 2026. It is about to fall. Please close the road immediately!"

[0128] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. A rockfall monitoring and early warning system based on multi-source information fusion, characterized in that, include: A multiphysics monitoring unit (100) is installed in the target unstable rock mass and its surrounding area to collect multiphysics response signals in real time throughout the entire process of the unstable rock mass instability. The multi-source data acquisition and synchronization unit (200) includes a data acquisition device (201) with a unified clock source, which is connected to the multi-physics monitoring unit (100) to realize the synchronous acquisition and timing alignment of multi-physics data, with a timing alignment error of less than 1 millisecond; The collaborative analysis and early warning unit (300) is a computer device (301) that receives multi-source synchronized data output by the multi-source data acquisition and synchronization unit (200) and performs the following analysis: (i) dividing the entire process of unstable rock mass into a stable stage, a damage accumulation stage, a critical instability stage and a dynamic instability stage; (ii) Construct an early warning criterion based on infrasound-microseismic coordinated response, wherein the early warning criterion includes: time-domain coordinated sudden increase criterion, frequency-domain coordinated migration criterion, energy release coordinated criterion and rupture mode transformation criterion; (iii) When one or more of the early warning criteria are triggered, output a rock instability early warning signal; The alarm unit (400) is connected to the collaborative analysis and early warning unit (300) and is used to receive early warning signals and perform alarm actions.

2. The rockfall monitoring and early warning system based on multi-source information fusion according to claim 1, characterized in that, The multiphysics monitoring unit (100) includes: The infrasound monitoring module (110) includes three infrasound sensors (111-1, 111-2, 111-3) and a rigid waveguide (112). The first infrasound sensor (111-1) is connected to the interior of the unstable rock mass through the rigid waveguide (112) and is used to collect low-frequency sound wave signals radiated into the air during the fracturing of the unstable rock mass. The second and third infrasound sensors (111-2, 111-3) are placed at an angle of 45° to the horizontal plane on both sides in front of the unstable rock mass. The microseismic monitoring module (120) includes a microseismic sensor (121), which is installed on the surface of the unstable rock mass or pre-embedded inside the unstable rock mass, for picking up elastic wave signals released by micro-fracture events inside the rock mass; The mechanical monitoring module (130) includes at least one load sensor (131), a resistance strain gauge (132) deployed on the surface of the unstable rock mass, and an FBG bare grid strain sensor (133). The motion attitude monitoring module (140) includes a biaxial tilt / accelerometer (141) installed on the top or leading edge of the unstable rock mass, for synchronously measuring the tilt angle change and triaxial acceleration of the unstable rock mass; The video monitoring module (150) includes a high-resolution video camera (151) deployed in a safe area opposite the dangerous rock mass to fully record the macroscopic damage morphology and crack propagation process of the dangerous rock mass.

3. The rockfall monitoring and early warning system based on multi-source information fusion according to claim 1, characterized in that, The alarm unit (400) includes a local sound and light alarm (401) and an SMS unit (402).

4. A method for preparing a rockfall monitoring and early warning system based on multi-source information fusion, characterized in that, Includes the following steps: Step S100: On-site survey and scheme design: Conduct geological surveys of the target unstable rock mass to determine its geometric shape, distribution of the main control structural planes, location of potential locking sections and surrounding stable bedrock areas, and design the deployment points of each module in the multi-physics monitoring unit accordingly. Step S200: Waveguide pre-embedding and sensor installation hole preparation: The rigid waveguide (112) is pre-cast in the key position inside the dangerous rock mass to make the waveguide tightly coupled with the rock mass material, and micro-seismic sensor installation holes are reserved during model casting; Step S300: Sensor installation and integration: Install the infrasound sensor, micro-vibration sensor, strain gauge, FBG bare grid strain sensor, load sensor, dual-axis tilt / accelerometer and video camera in sequence; Step S400: System connection and synchronization debugging: Connect each sensor and camera to the data acquisition device (201) with a unified clock source, verify the time synchronization accuracy of all monitoring channels through transient knocking signals, ensure that the timing alignment error is less than 1 millisecond, and set the sampling parameters of the data acquisition device; Step S500: Background noise acquisition and baseline establishment: Start the monitoring system and acquire background noise and microseismic data for no less than 1 minute as a baseline reference for subsequent signal processing; Step S600: System integration and trial operation: Start all monitoring modules and conduct a trial operation for no less than 24 hours. After confirming that the system stability and data quality meet the requirements, the system will be officially put into monitoring and early warning operation. Step S700: Continuous monitoring and data acquisition: The system enters continuous monitoring mode, collects multi-physics field data in real time and transmits it to the collaborative analysis and early warning unit (300).

5. The method for preparing a rockfall monitoring and early warning system based on multi-source information fusion according to claim 4, characterized in that, Step S300 specifically includes: S301: Install an infrasound sensor (111-1) at the waveguide outlet and ensure a reliable seal at the connection point; S302: Install the micro-vibration sensor (121) in the reserved mounting hole, and apply coupling agent between the sensor and the hole wall; S303: Attach strain gauges (132) to the tip of the crack at the rear edge of the unstable rock mass and near the locking section, and attach FBG bare grid strain sensors (133) to the surface of the locking section and key parts of the crack propagation path. S304: Install borehole load sensors (131) in monitoring holes near the top or rear edge cracks of the unstable rock mass. S305: Rigidly install a biaxial tilt / accelerometer (141) on the top or leading edge of the unstable rock mass. S306: Deploy high-resolution video cameras (151) in the safe area opposite the unstable rock mass.

6. A method for monitoring and early warning of dangerous rocks based on multi-source information fusion, characterized in that, Includes the following steps: Step S110: On-site survey and sensor deployment: Conduct geological surveys of the target unstable rock mass, and deploy infrasound sensors, micro-seismic sensors, load sensors, strain gauges, dual-axis tilt / accelerometers and video cameras in key parts of the unstable rock mass and surrounding stable areas. Step S210: System integration and synchronization debugging: Connect all sensors and cameras to the data acquisition device (201) with a unified clock source to ensure that the timing alignment error is less than 1 millisecond; Step S310: Background noise acquisition: Acquire background noise and microseismic data for no less than 1 minute as a baseline reference for subsequent signal processing; Step S410: Continuous monitoring and data acquisition: Continuously acquire multi-source data and transmit it in real time to the collaborative analysis and early warning unit (300). Step S510: Multi-source data collaborative analysis and early warning: The collaborative analysis and early warning unit (300) performs a full process division of unstable rock mass on the received multi-source synchronous data, constructs an early warning criterion based on infrasound-microseismic collaborative response, and identifies and outputs unstable rock mass early warning signals; Step S610: Alarm Execution: After receiving the warning signal, the alarm unit (400) activates the on-site audible and visual alarm (401) and / or sends the warning information via the SMS unit (402).

7. The method for monitoring and early warning of dangerous rocks based on multi-source information fusion according to claim 6, characterized in that, Step S510 specifically includes: S511: Calculate the short-time event rate of the infrasound signal and the event rate of the microseismic signal, and determine whether the time-domain coordinated sudden increase criterion is met; S512: Extract the dominant frequency of the microseismic signal by power spectral density estimation, calculate the energy proportion of the infrasound signal in the 0.5~5Hz frequency band, and determine whether the frequency domain cooperative migration criterion is met; S513: Calculate the Pearson correlation coefficient between the square integral energy of the microseismic signal and the root mean square amplitude of the infrasound signal, and determine whether the energy release coordination criterion is met; S514: Analyze the changes in shear failure ratio based on the RA / AF values ​​of microseismic events to determine whether the failure mode transition criterion is met; S515: When one or more of the above criteria are triggered, the unstable rock mass is determined to have entered the critical instability stage, and an early warning signal is output to the alarm unit (400).

8. The method for monitoring and early warning of dangerous rocks based on multi-source information fusion according to claim 6, characterized in that, The temporal-domain coordinated surge criterion is triggered when the short-term event rate of the infrasound signal exceeds three times the standard deviation of the background mean during the stable phase, and the event rate of the microseismic signal simultaneously shows a significant increase exceeding two times the standard deviation of its background value, and the two are highly overlapping in time.

9. The method for monitoring and early warning of dangerous rocks based on multi-source information fusion according to claim 6, characterized in that, The frequency domain cooperative migration criterion is triggered when the instantaneous dominant frequency of the microseismic signal continuously decreases to below 70% of its initial characteristic value, and the energy proportion of the infrasound signal in the 0.5~5Hz frequency band exceeds 50%.

10. The method for monitoring and early warning of dangerous rocks based on multi-source information fusion according to claim 6, characterized in that, The energy release criterion is triggered when the Pearson correlation coefficient between the square integral energy of the microseismic signal and the root mean square amplitude of the infrasound signal exceeds 0.8 and lasts for more than 10 seconds.

11. The method for monitoring and early warning of dangerous rocks based on multi-source information fusion according to claim 6, characterized in that, The criterion for the change in fracture mode is triggered when the shear fracture ratio increases from less than 30% to more than 60% within 90 seconds.

Images