A slope rockfall intelligent protection net lifting control method based on multi-sensor cooperation

Abstract

The application discloses a kind of slope rockfall intelligent protection net lifting control method based on multi-sensing coordination.The method comprises the following steps: collecting multi-source monitoring data for time synchronization and preprocessing;Extract the rockfall event characteristics and carry out multi-sensing coordination fusion, obtain the spatial position of rockfall, rockfall equivalent size and rockfall movement speed, and output fusion confidence;Risk assessment based on the fusion result, calculate risk score and generate warning level, while calculating the predicted arrival time of rockfall to the protection plane;According to the warning level and the rockfall equivalent size, generate a protection strategy, calculate the target height of the protection net and control the lifting of the protection net;When the rockfall equivalent size meets the preset rock breaking condition, control the rock breaking device and the protection net to execute rock breaking and interception under the condition that manual confirmation condition is established.The application realizes the monitoring, early warning and protection control of slope rockfall event through multi-sensing coordination fusion and the linkage control of protection net lifting and rock breaking.

Description

Technical Field

[0001] This invention relates to the field of slope disaster monitoring technology, specifically to a method for controlling the raising and lowering of an intelligent rockfall protection net based on multi-sensor collaboration. Background Technology

[0002] Rockfall incidents on slopes are sudden, with rocks moving rapidly along the slope surface or structure under gravity, posing a risk of entering tracks, roads, or protected areas. Current methods for managing rockfalls typically involve fixed protective structures, manual inspections, or monitoring with a single type of sensor. Manual inspections are constrained by temporal continuity and real-time requirements, making it difficult to provide continuous data for control actions during an incident. Single-type sensors are susceptible to data loss or false triggering under conditions of line-of-sight obstruction, changes in lighting, environmental interference, or equipment offline, resulting in inconsistent and unreliable outputs regarding the spatial location, size, and movement of rockfalls.

[0003] In existing protective equipment, the height of protective netting is often fixed or manually set, lacking a linkage control process with rockfall risk factors (equivalent size of the rockfall, rockfall speed, predicted arrival time, etc.). This results in an unclear temporal and spatial correspondence between the raising and lowering of the protective netting and rockfall events. For large rocks, relying solely on the protective netting for interception places demands on structural maintenance due to the impact load and the volume of the accumulated material. Furthermore, without a linked rock-breaking device, the timing and interlocking relationship between the rock-breaking action and the protective action is unclear, making it difficult to establish a unified execution logic.

[0004] Furthermore, when rockfall monitoring data, early warning events, and control execution records are stored in a scattered manner or are not archived in a structured manner, it is difficult to conduct statistical analysis on rockfall trends, spatial distribution, and response times within a specified time range, and it is also difficult to trace operations and maintenance based on historical records. Summary of the Invention

[0005] This invention provides a method for controlling the raising and lowering of an intelligent slope rockfall protection net based on multi-sensor collaboration. It is used to form a unified time reference processing flow for multi-source monitoring data, output the spatial location of rockfall events, the equivalent size of the rockfall, and the speed of rockfall movement, and control the raising and lowering of the protection net and the linkage of rock breaking based on the risk assessment results. At the same time, the monitoring data and control data are stored and historical data analysis is performed.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling the raising and lowering of an intelligent slope rockfall protection net based on multi-sensor collaboration, comprising:

[0007] Multi-source monitoring data was collected at multiple monitoring points along the slope. The multi-source monitoring data included image data collected by a camera device, distance echo data collected by an ultrasonic device, and infrared detection data collected by an infrared detection device.

[0008] The multi-source monitoring data is time-synchronized and preprocessed to obtain an effective data stream under a unified time reference;

[0009] Feature extraction is performed on the effective data stream to obtain the rockfall event features;

[0010] The features of the rockfall event are fused using multi-sensor collaborative fusion to obtain the spatial location, equivalent size, and velocity of the rockfall, and the fusion confidence score is output.

[0011] Risk assessment is performed based on the spatial location, the equivalent size of the falling rock, the speed of the falling rock, and the fusion confidence level. A risk score is calculated and an early warning level is generated. At the same time, the predicted arrival time of the falling rock to the protective plane is calculated.

[0012] Based on the warning level, the equivalent size of the falling rocks and the predicted arrival time, a protection strategy is generated, the target height of the protective net is calculated and a lifting control command for the protective net is sent to the corresponding base unit to raise or lower the protective net to the target height.

[0013] When the equivalent size of the falling rock is greater than or equal to a preset size threshold, a rock-breaking linkage control command is sent to the base unit to activate the rock-breaking device and coordinate with the protective net to break the rock and intercept it; when the equivalent size of the falling rock is less than the preset size threshold, only the protective net is controlled to intercept it.

[0014] The multi-source monitoring data, the fusion results, the early warning level, the protective net lifting control command, the rock-breaking linkage control command, and the execution feedback of the base unit are stored, and historical data analysis is performed to output the analysis results.

[0015] Furthermore, the present invention also provides an intelligent slope rockfall protection net lifting and lowering control system, including a multi-sensor monitoring unit, a communication unit, a control platform and a storage unit, as well as multiple base units; wherein, the control platform is configured to execute the above-mentioned method, the base units are used to execute the protection net lifting and lowering control commands and the rock breaking linkage control commands and send back execution feedback, and the storage units are used to perform structured storage of multi-source monitoring data, fusion results, early warning information and control logs and provide data sources for historical data analysis.

[0016] Compared with the prior art, the present invention has at least the following technical effects:

[0017] By synchronizing image data, range echo data and infrared detection data and forming a collaborative observation package, the joint processing of multi-source monitoring data under a unified time reference is achieved.

[0018] By using multi-sensor collaborative fusion to output spatial location, equivalent size of falling rocks, speed of falling rocks and fusion confidence, and updating the fusion weights based on online status and consistency verification results, the data from missing or abnormal sensors can be downweighted or zeroed out.

[0019] Risk assessment calculates risk scores and generates warning levels, while predicting arrival times are also calculated to output the time-series parameters required for generating protection strategies.

[0020] By calculating and limiting the target height of the protective net, and combining it with the height feedback of the base unit to form a closed-loop control, a closed-loop management of the protective net lifting control command to execution feedback is realized.

[0021] By issuing rock-breaking linkage control commands based on preset size thresholds and managing linkage timing, the coordinated execution and interlocking control of the rock-breaking device and the protective net can be achieved in the same event.

[0022] By structurally storing multi-source monitoring data, fusion results, early warning information, control commands and execution feedback, and performing historical data analysis based on the stored data, the system can output and trace trend statistics, spatial distribution statistics and response time statistics. Attached Figure Description

[0023] Figure 1 This is a schematic diagram illustrating the operational effect of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention.

[0024] Figure 2 This is a schematic diagram of the system architecture of a method for controlling the lifting and lowering of an intelligent slope rockfall protection net based on multi-sensor collaboration according to the present invention.

[0025] Figure 3 This is a schematic diagram of the rockfall detection and early warning module of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration of the present invention.

[0026] Figure 4 This is a schematic diagram of the lifting control module of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention. Figure 1 ;

[0027] Figure 5 This is a schematic diagram of the lifting control module of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention. Figure 2 ;

[0028] Figure 6 This is a schematic diagram showing the detailed status information of the base unit of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention.

[0029] Figure 7This is a schematic diagram of the historical data analysis module of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration of the present invention;

[0030] Figure 8 This is a flowchart of a method for controlling the raising and lowering of an intelligent slope rockfall protection net based on multi-sensor collaboration, according to the present invention.

[0031] Figure 9 This is a schematic diagram of the early warning process for rockfall-prone sections based on historical data in the intelligent rockfall protection net lifting control method for slopes based on multi-sensor collaboration of the present invention.

[0032] Figure 10 This is a schematic diagram of the base unit of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention.

[0033] Figure 11 This is an example of the fluctuation of the total number of detections and the number of high-risk events in the past 30 days for the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration of the present invention.

[0034] Figure 12 This is a bar chart showing the historical cumulative detection frequency of rockfalls at 12 monitoring points, illustrating the lifting control method for intelligent slope rockfall protection nets based on multi-sensor collaboration according to the present invention.

[0035] Figure 13 The pie chart above illustrates the average response time (2.39s) from early warning to effective protection, as an example of the intelligent slope rockfall protection net lifting control method based on multi-sensor collaboration according to the present invention. Attached Figure Description

[0036] 1. Base body; 2. Support structure; 3. Protective net assembly; 4. Rock breaking device. Detailed Implementation

[0037] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] In this embodiment, the control platform is a computing device with data processing and control command generation capabilities, including a processor, a memory, and a communication interface for communicating with the multi-sensor monitoring unit and the base unit.

[0039] The memory stores a computer program, which, when executed by the processor, is used to realize time synchronization and preprocessing of multi-source monitoring data, feature extraction, multi-sensor collaborative fusion, risk assessment, early warning level generation, protection strategy calculation, and generation of protective net lifting control commands and rock breaking linkage control commands.

[0040] The control platform also includes a human-machine interface unit, which displays the status of monitoring points, rockfall event information, warning levels, and the execution status of the base unit, and receives control commands or confirmation signals input by operators.

[0041] The control platform and the base unit interact with each other through a communication interface to issue control commands and transmit execution feedback.

[0042] In this embodiment, a multi-sensor collaborative intelligent slope rockfall protection net lifting control method is implemented in a slope rockfall protection scenario. Multiple monitoring points are deployed along the slope, each collecting multi-source monitoring data. Multiple base units are also deployed along the slope or along a track / road, each equipped with a protective net and a rock-breaking device. The base units correspond spatially to the monitoring points, enabling the monitoring points to issue control commands to the corresponding base units and receive execution feedback after detecting a rockfall event.

[0043] To ensure consistency in terminology and symbols, this implementation defines key quantities as follows: the discrete time index is... The corresponding timestamp is The width of the sliding time window is At any moment The set of time windows is The sensor set is , representing the camera device, ultrasonic device, and infrared detection device, respectively. The world coordinate system is . Falling rocks The spatial position vector below is The velocity vector of the falling rock is The equivalent size of the falling rock is The fusion confidence level is... The risk score is The warning level is The predicted arrival time is The target height of the protective net is The current height of the protective net is reported as follows: The preset size threshold is Safety margin is The minimum height and maximum height are respectively and .

[0044] refer to Figures 1 to 9In this embodiment, the control platform includes a rockfall detection and early warning module, a protective net lifting control module, a rock-breaking linkage control module, and a historical data analysis module. The rockfall detection and early warning module receives multi-source monitoring data and outputs early warning levels. The protective net lifting control module generates the target height of the protective net according to the protection strategy and issues lifting control commands. The rock-breaking linkage control module issues rock-breaking linkage control commands when trigger conditions are met and manages the execution status of the rock-breaking device and the protective net according to the linkage sequence. The historical data analysis module performs statistical analysis on historical data in the storage unit and outputs the analysis results. The control platform has manual and automatic modes. In manual mode, the protective net lifting control commands and rock-breaking linkage control commands are generated from operational input. In automatic mode, the control platform generates the protective net lifting control commands and rock-breaking linkage control commands based on the early warning level and the fusion results. Control commands in both modes are sent to the base unit through the communication unit and execution feedback is received.

[0045] In step S1, the multi-source monitoring data is collected in the following manner: the camera device collects image data. The ultrasonic device collects distance echo data. Infrared detection device collects infrared detection data. or thermal imaging data The control platform adds time stamps to the data uploaded from each monitoring point, forming a discrete time sequence. .

[0046] In step S2, time synchronization and preprocessing are performed as follows: The control platform establishes a unified time reference, aligns the timestamps from the camera device, ultrasonic device, and infrared detection device, and follows a sliding time window. Multi-source monitoring data within the same time window are combined into a collaborative observation package. Collaborative observation package It includes image data segments, range echo data segments, and infrared detection data segments, and contains online status indicators for each sensor; when any sensor is in... When testing for internal defects, The corresponding missing test markers are recorded for subsequent multi-sensor collaborative fusion.

[0047] Preprocessing is performed separately for each data type. Denoising and illumination compensation are applied to the image data, and the preprocessed image data is output. Range verification and anomaly removal are performed on the range echo data, and filtering and smoothing are applied to the remaining sequence to output preprocessed range echo data. Triggered debouncing is performed on the infrared detection data, and preprocessed infrared detection data is output. When the infrared detection data is thermal imaging data, background suppression and threshold segmentation are performed on the thermal imaging data, and preprocessed thermal imaging data is output. .

[0048] In step S3, the features of the rockfall event are extracted as follows. For the imaging device, the control platform preprocesses the image data. Perform object detection to obtain a set of candidate boxes. The system performs target tracking to obtain the pixel trajectory and pixel scale of the falling rock target. After the camera device completes camera calibration, the control platform converts the pixel trajectory into a spatial position estimate based on the calibration parameters. And convert the pixel scale into the equivalent size estimate of the falling rocks. - Control platform based on Temporal difference calculation for speed estimation of camera device Its calculation formula is

[0049]

[0050] For ultrasonic devices, the control platform preprocesses distance echo data. Calculate radial velocity characteristics Its calculation formula is

[0051]

[0052] The control platform generates the measurement vector of the ultrasonic device.

[0053]

[0054] For infrared detection devices, the control platform obtains infrared triggering characteristics from preprocessed infrared detection data. When the infrared detection device outputs pre-processed thermal imaging data At that time, the control platform Perform connected component extraction to obtain the centroid of the infrared target profile and convert it into a spatial position estimate for the infrared detection device. The equivalent size estimate of the falling rocks for the infrared detection device is obtained from the infrared target contour scale. .

[0055] In step S4, multi-sensor collaborative fusion is performed as follows: The control platform is based on the output of the camera device. Forming the first candidate fusion result The control platform is based on the measurement vector of the ultrasonic device. Forming a second candidate fusion result The second candidate fusion result is used to constrain the proximity and velocity components of the falling rock towards the ultrasonic device. The control platform is based on infrared triggering characteristics. as well as Forming a third candidate fusion result .

[0056] The control platform performs a consensus check on the first and third candidate fusion results. The consensus check includes position residual verification and size residual verification. The position residual vector is defined as follows:

[0057]

[0058] Dimensional residual is defined as

[0059]

[0060] Control platform according to and A consistency assessment result is generated and used to update the health status of each sensor. The control platform manages each sensor. Maintain health The health score is determined by the online status indicator, the missing data indicator, and the consistency judgment result; when a sensor is missing data or offline, the corresponding health score is zero. The control platform calculates adaptive weights based on the health score.

[0061]

[0062] in This is a constant greater than zero. The control platform performs weighted fusion on the candidate fusion results based on adaptive weights to obtain the spatial location of the falling rocks. equivalent size of falling rocks With the speed of falling rocks The fusion confidence level corresponding to the fusion result. Determined by health status and consistency assessment results; the control platform constructs a system based on the summarized health status value and consistency assessment results. and will and —and output.

[0063] In step S5, risk assessment, predicted arrival time calculation, and early warning level generation are performed as follows: The control platform sets up a protective plane. The protective plane in the world coordinate system The following satisfies the plane equation

[0064]

[0065] in For unit normal vector, The value is a constant. The control platform extrapolates the trajectory of falling rocks based on the rockfall fusion results. And calculate the predicted arrival time.

[0066]

[0067] when or At that time, the control platform will Record it as unavailable and do not use this value in the protection policy.

[0068] The control platform is based on the equivalent size of the falling rocks. Modulus of Falling Rock Motion Predicted arrival time With fusion confidence Calculate risk score The risk score is calculated using a weighted summation method:

[0069]

[0070] in Non-negative weighting coefficients: It is a monotonically non-decreasing mapping function. The control platform uses the threshold set... Map risk scores to warning levels:

[0071]

[0072] The control platform generates early warning information records, which must include at least a timestamp. Spatial location equivalent size of falling rocks Falling rock speed Fusion confidence Risk Score Warning Level With predicted arrival time

[0073] In step S6, the protection strategy generation and the control of the protective net raising and lowering are executed as follows: The control platform determines the protection strategy based on the warning level. equivalent size of falling rocks With predicted arrival time Generate a protection strategy and calculate the target height of the protection network. The target height of the protective netting is calculated and limited according to the following formula:

[0074]

[0075] The control platform is positioned relative to the location of the falling rocks. The corresponding base unit issues a lifting control command for the protective netting. This command must include at least the base unit identifier, a timestamp, and the target height of the protective netting. Action indicators. After receiving the lifting control command for the protective net, the base unit drives the lifting mechanism of the protective net to perform lifting and lowering, and collects height feedback to obtain the current height feedback of the protective net. and will Together with the execution status, it forms an execution feedback and control platform.

[0076] The protective netting lifting mechanism of the base unit uses closed-loop control to achieve target height tracking. The base unit calculates the height error. And calculate the actuator input according to the discrete PID control law.

[0077]

[0078] Where K p K i K d For control parameters, the base unit is equipped with limit signals and fault monitoring signals; when a limit signal or fault monitoring signal is triggered, the base unit stops the actuator in the corresponding direction and sends a fault code back to the control platform.

[0079] In step S7, the rock-breaking linkage control is executed as follows: The control platform performs a trigger determination based on the equivalent size of the falling rock and a preset size threshold: when d k ≥d th At that time, the control platform generates a rock-breaking linkage control command and sends it to the corresponding base unit; when d k <d th At this time, the control platform does not send rock-breaking linkage control commands. Rock-breaking linkage control commands must include at least the base unit identifier, a timestamp, and a rock-breaking action identifier.

[0080] After receiving the rock-breaking linkage control command, the base unit executes the rock-breaking linkage sequence. The rock-breaking linkage sequence includes: rock-breaking device extension, raising and lowering the protective net to the target height, rock-breaking device activation, rock-breaking device shutdown, rock-breaking device retraction, and protective net retraction. The base unit sets position detection signals for the rock-breaking device's extension and retraction, and executes the next action after the position detection signal is satisfied. The base unit sets interlock conditions for the rock-breaking device and the protective net lifting mechanism; interlock conditions include emergency stop signals, limit signals, and communication status signals. If any interlock condition is not satisfied, the base unit stops the rock-breaking device and the protective net lifting mechanism and sends back a fault code.

[0081] In this implementation, the determination of whether the rockfall event has ended and the risk has been eliminated is executed by the control platform. The control platform uses a collaborative observation package. The results of consecutive time windows are used as the basis for judgment; when the fusion results of multiple consecutive time windows do not output the spatial location of the falling rocks. And the warning level is At that time, the control platform generates a contraction and recovery command and sends it to the base unit, causing the protective net to contract to its initial state and the rock-breaking device to retract to its initial state.

[0082] In this embodiment, communication, execution feedback, and status management are performed as follows: When the control platform issues control commands to the base unit, it assigns a command identifier and a timestamp to each control command and records them in the control log. The base unit sends back execution feedback for each control command. The execution feedback includes at least the command identifier, timestamp, and the current height h of the protective net. k The system displays the status, execution status, and fault codes of the rock-breaking device. The control platform stores execution feedback and corresponding control commands in a linked storage unit. A heartbeat message is established between the control platform and the base unit; if the control platform does not receive a heartbeat message or execution feedback within a timeout period, it sets the corresponding base unit status to offline and stops sending new control commands to that base unit.

[0083] In step S8, data storage and historical data analysis are performed as follows: The storage unit stores multi-source monitoring data, fusion results, early warning information, control commands, and execution feedback in a structured manner. Multi-source monitoring data includes at least timestamps, monitoring point identifiers, image data indexes, distance echo data sequences, infrared detection data sequences, and online status identifiers. The fusion results include at least timestamps. Spatial location equivalent size of falling rocks Speed ​​of falling rocks With fusion confidence The warning information should include at least a risk score. Warning Level With predicted arrival time The control log must include at least the target height of the protective netting. Control command identifier, issuance timestamp, execution feedback timestamp, execution status and fault code.

[0084] The historical data analysis module performs statistical analysis on historical data in the storage unit and outputs the analysis results. The historical data analysis module operates within a specified time range. The number of rockfall events is counted internally. The sum of the rockfall events is defined by the indicator function.

[0085]

[0086] in For the first The start timestamp of a rockfall event. This is an indicator function. The historical data analysis module performs spatial distribution statistics on rockfall events based on monitoring point identifiers or spatial locations, and outputs the spatial distribution statistics results. The historical data analysis module calculates the response time based on the control command issuance timestamp and the execution feedback arrival timestamp. The response time for a single event is defined as...

[0087]

[0088] in For the first The warning timestamp for each event. This is the timestamp of the protective netting being raised and lowered in connection with the event. The historical data analysis module... The system performs statistical analysis and outputs response time statistics. The historical data analysis module supports filtering by time range and exporting the analysis results, then writes the exported records to the storage unit.

[0089] In step S8, the control platform stores the multi-source monitoring data, fusion results, early warning information, protective net lifting control commands, rock breaking linkage control commands, and execution feedback from the base unit in a structured manner.

[0090] Structured storage includes at least the time dimension, monitoring point identifier, base unit identifier, rockfall spatial location, rockfall equivalent size, warning level, protective net target height, and execution status field.

[0091] The historical data analysis module performs multi-dimensional statistical analysis on rockfall events based on structured data. This multi-dimensional statistical analysis includes at least the following:

[0092] Statistical analysis of the number of rockfall events within a specified time range is conducted to obtain statistical results on the changes in rockfall events over time.

[0093] Statistical analysis was conducted on the distribution of rockfall events at different monitoring points or different spatial locations to obtain the spatial distribution results of rockfall events;

[0094] Statistical analysis was conducted on the issuance time of the control command for raising and lowering the protective net and the arrival time of the corresponding execution feedback to obtain statistical results of the system response time.

[0095] The historical data analysis module outputs the statistical analysis results to the human-computer interaction unit of the control platform or exports them as data files for querying or subsequent processing.

[0096] refer to Figure 11 Analysis: The blue broken line (total number of tests) reflects the overall activity level of rockfall on the slope.

[0097] Red broken line (high-risk event): Marks high-risk periods that require special attention and handling.

[0098] Pattern identification: By comparing the two curves, it can be seen that high-risk events are often accompanied by a peak in the total number of detections. However, local peaks in high-risk events may also occur during periods of lower total detections (such as mid-November), indicating the need for continuous monitoring.

[0099] refer to Figure 12Note: Red bars (>20 times): Represent high-risk areas (e.g., Points 2, 6, 7, 9, 10, 11). These areas experience frequent rockfalls and are key areas for protection and reinforcement. It is recommended to prioritize the deployment of rockbreaking devices in these locations and increase the frequency of inspections.

[0100] Yellow bars (10-20 times): Represent medium-risk areas. Regular monitoring is required, and attention should be paid to the increasing trend of their frequency.

[0101] Green bars (<10 times): Represent low-risk areas. Although the current risk is low, basic monitoring coverage still needs to be maintained.

[0102] This chart uses color coding to visualize risk levels, helping managers quickly identify key areas and optimize the allocation of protective resources.

[0103] refer to Figure 13 The system's response distribution over different time periods is shown. Most responses (approximately 70%) are completed within 2 seconds, with an average response time of 2.39 seconds, demonstrating the system's real-time performance in responding to rockfall risks.

[0104] Statistical data explanation: <1s (25.7%): Ultra-fast response range; 1-2s (44.5%): Main response range; 2-3s (18.3%): Normal response range; >3s (11.5%): Long-tailed response range.

[0105] refer to Figure 10 In this embodiment, the base unit is an execution device for executing the protective net lifting control command and the rock breaking linkage control command.

[0106] The base unit includes a base body 1, a support structure 2, a protective net assembly 3, and a rock-breaking device 4.

[0107] The base body 1 is a fixed structure installed on the slope or protected area, used to support the support structure 2, the protective net assembly 3, and the rock-breaking device 4. The base body 1 is equipped with a drive unit and a communication unit. The drive unit is used to drive the protective net assembly 3 and the rock-breaking device 4 to perform corresponding actions, and the communication unit is used to communicate with the control platform to receive control commands and send back execution feedback.

[0108] The support structure 2 is installed on the base body 1 to support the protective net assembly 3 and the rock-breaking device 4. The support structure 2 is set to be adjustable, which is used to change the installation angle of the protective net assembly 3 and the rock-breaking device 4 relative to the base body 1 to adapt to different slope gradients or rockfall trajectories.

[0109] The protective net assembly 3 is connected to the support structure 2. When not in operation, the protective net assembly 3 is in a retracted state. After receiving the protective net lifting control command, it unfolds along the support structure 2 and lifts to the target height of the protective net to intercept falling rocks.

[0110] The rock-breaking device 4 is installed on the support structure 2 and is located in front of or above the protective net assembly 3. After receiving the rock-breaking linkage control command, the rock-breaking device 4 switches from the storage state to the working state to perform rock-breaking processing on falling rocks that meet the rock-breaking trigger conditions.

[0111] The base unit is also equipped with a position detection unit and a status detection unit, which are used to detect the current status of the protective net assembly 3 and the rock breaking device 4, and transmit the detection results back to the control platform as execution feedback.

[0112] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for controlling the lifting and lowering of an intelligent slope rockfall protection net based on multi-sensor collaboration, characterized in that, Includes the following steps: S1. Collect multi-source monitoring data at multiple monitoring points along the slope. The multi-source monitoring data includes image data collected by a camera device, distance echo data collected by an ultrasonic device, and infrared detection data collected by an infrared detection device. S2. Perform time synchronization and preprocessing on the multi-source monitoring data to obtain an effective data stream under a unified time reference; S3. Perform feature extraction on the effective data stream to obtain the features of the rockfall event; S4. Perform multi-sensor collaborative fusion on the features of the rockfall event to obtain the spatial location, equivalent size, and velocity of the rockfall, and output the fusion confidence level. S5. Based on the spatial location, the equivalent size of the falling rock, the speed of the falling rock movement, and the fusion confidence level, a risk assessment is performed, a risk score is calculated, and an early warning level is generated. At the same time, the predicted arrival time of the falling rock to the protective plane is calculated. S6. The control platform generates a protection strategy based on the warning level, the equivalent size of the falling rock and the predicted arrival time, and calculates the target height of the protective net. The control platform sends a lifting control command to the base unit corresponding to the spatial position of the falling rock, so that the protective net is raised or lowered to the target height. The control platform supports both automatic control mode and manual intervention mode. In manual intervention mode, the control platform allows operators to adjust the target height of the protective net while keeping the calculation logic of the target height of the protective net unchanged. The adjusted target height of the protective net serves as the target for the execution of the lifting and lowering control command of the protective net. S7. The control platform performs a rock-breaking trigger determination based on the equivalent size of the falling rock and a preset size threshold. When the equivalent size of the falling rock is greater than or equal to the preset size threshold, the control platform generates a rock-breaking linkage control request signal. The rock-breaking linkage control request signal must simultaneously meet the conditions of a valid manual confirmation signal before the control platform can issue a rock-breaking linkage control command to the corresponding base unit to start the rock-breaking device and coordinate with the protective net to perform rock breaking and interception. When the manual confirmation signal is not valid, the control platform will not issue a rock-breaking linkage control command, but will only execute the lifting and lowering control of the protective net; S8. Store the multi-source monitoring data, the fusion result, the early warning level, the protective net lifting control command, the rock breaking linkage control command, and the execution feedback of the base unit, and perform historical data analysis to output the analysis result.

2. The method according to claim 1, characterized in that, The time synchronization includes: establishing a unified time reference for the camera device, ultrasonic device and infrared detection device; adding timestamps to the multi-source monitoring data and aligning them in time; forming a collaborative observation package by combining image data, distance echo data and infrared detection data within the same time window according to a sliding time window; and using the collaborative observation package as the input for subsequent feature extraction and multi-sensor collaborative fusion.

3. The method according to claim 1, characterized in that, The preprocessing includes: denoising and illumination compensation of the image data to obtain preprocessed image data; anomaly removal and filtering of the distance echo data to obtain preprocessed distance echo data; shaking removal of the infrared detection data to obtain preprocessed infrared detection data; and marking the missing state in the collaborative observation package when any sensor is missing or offline, so that the multi-sensor collaborative fusion can perform adaptive processing.

4. The method according to claim 1, characterized in that, The feature extraction includes: Image detection features of the falling rock target are extracted from the preprocessed image data and target tracking is performed to obtain the image position and image scale of the falling rock target. Based on the calibration parameters, the image position and image scale are converted into the spatial position estimate of the camera device and the equivalent size estimate of the falling rock of the camera device. The distance variation features are extracted from the preprocessed distance echo data and the radial velocity features are calculated to obtain the distance measurement and velocity measurement of the ultrasonic device. Infrared trigger features are extracted from the preprocessed infrared detection data, and infrared target contour features are further extracted when the infrared detection data is thermal imaging data to obtain the spatial position estimate of the infrared detection device and the equivalent size estimate of the falling rocks of the infrared detection device.

5. The method according to claim 1, characterized in that, The multi-sensor collaborative fusion includes: A first candidate fusion result is obtained based on the spatial position estimation of the camera device, the equivalent size estimation of the falling rock of the camera device, and the tracking result of the falling rock target; a second candidate fusion result is obtained based on the distance measurement and velocity measurement of the ultrasonic device; a third candidate fusion result is obtained based on the infrared triggering feature, the spatial position estimation of the infrared detection device, and the equivalent size estimation of the falling rock of the infrared detection device. The first candidate fusion result, the second candidate fusion result, and the third candidate fusion result are subjected to a collaborative consistency verification. The collaborative consistency verification includes at least the residual verification of the spatial position estimation of the camera device and the spatial position estimation of the infrared detection device, and the residual verification of the equivalent size estimation of the falling rocks of the camera device and the equivalent size estimation of the falling rocks of the infrared detection device. Based on the successful completion of the collaborative consistency verification, the first candidate fusion result, the second candidate fusion result, and the third candidate fusion result are weighted and fused to output the spatial location of the falling rock, the equivalent size of the falling rock, and the speed of the falling rock.

6. The method according to claim 5, characterized in that, The weights in the weighted fusion are adaptive weights, which are determined based on the online status, noise level, and historical accuracy of each sensor. When any sensor is missing, offline, or fails the collaborative consistency verification, the adaptive weight of the corresponding sensor is reduced or set to zero, so that the multi-sensor collaborative fusion can still output the fusion result and the fusion confidence under abnormal sensor conditions.

7. The method according to claim 1, characterized in that, The risk assessment includes: calculating the risk score based on the equivalent size of the falling rock, the speed of the falling rock, the relative relationship between the spatial location and the protective plane, the predicted arrival time, and the fusion confidence level, and mapping the risk score to the warning level; wherein the predicted arrival time is obtained by extrapolating the time of the falling rock to the protective plane from the spatial location and the speed of the falling rock.

8. The method according to claim 1, characterized in that, The target height of the protective net is calculated from the equivalent size of the falling rocks and the safety margin, and height limiting processing is performed to limit it between a preset minimum height and a preset maximum height; the base unit performs closed-loop control on the protective net according to the height feedback signal to achieve the target height of the protective net, and sends the execution status back to the control platform as the execution feedback.

9. The method according to claim 1, characterized in that, The issuance and execution of the rock-breaking linkage control command includes: when the equivalent size of the falling rock is greater than or equal to the preset size threshold, the control platform issues a rock-breaking device extension command, a protective net lifting control command, and a rock-breaking device start command to the base unit according to the preset linkage sequence, so that the rock-breaking device is in the working position and performs rock breaking, while the protective net is raised and lowered to the target height of the protective net for interception; when the rockfall event ends and the risk is eliminated, the control platform issues a rock-breaking device retraction command and a protective net retraction command to the base unit, so that the rock-breaking device is retracted and the protective net is retracted to the initial state.

10. The method according to claim 1, characterized in that, The historical data analysis includes: structured storage of the multi-source monitoring data, the fusion results, the early warning levels, and the execution feedback; generating trend statistics of rockfall events according to the time dimension; generating spatial distribution statistics of rockfall events according to the spatial location of monitoring points; calculating response time statistics based on the issuance time of the protective net lifting control command and the arrival time of the execution feedback; and supporting querying and exporting the analysis results by time range.

Images