Cave-in early warning method, device, medium and equipment based on cellular cave-in equipment

Abstract

The application relates to a landslide early warning method and device based on a cellular landslide equipment, a medium and an equipment, and relates to the technical field of landslide early warning. The method comprises the following steps: obtaining deformation monitoring data of each monitoring point on a slope surface of a monitoring section of a tourist highway through a plurality of target cellular cluster version landslide equipment; judging whether a landslide precursor exists in the current monitoring section according to the deformation monitoring data of each monitoring point; if the current monitoring section has a landslide precursor, determining a risk value of the monitoring section appearing landslide, and obtaining auxiliary picture information of the monitoring section through a camera arranged at a first preset distance before and after the monitoring section; identifying a road risk trigger factor according to the auxiliary picture information, determining a final early warning level according to the road risk trigger factor and the risk value, and issuing an early warning reminder through a sound and light alarm arranged at a second preset distance before and after the monitoring section. The application has the effect of improving the timeliness of landslide early warning.

Description

Technical Field

[0001] This application relates to the field of landslide early warning technology, specifically to a landslide early warning method, device, medium, and equipment based on cellular landslide equipment. Background Technology

[0002] The geological conditions along mountainous tourist roads are complex. Under the influence of external factors such as rainfall infiltration, snowmelt softening, and vehicle vibration, the internal stress and seepage fields of the slope soil and rock continuously evolve, easily inducing geological disasters such as landslides and collapses. As vital transportation routes within scenic areas, tourist roads are characterized by high traffic volume, dense tourist populations, and a wide social impact. Once a slope collapse occurs, it not only directly threatens the lives and property of passing vehicles and tourists but may also disrupt the main transportation artery of the scenic area, causing large-scale tourist congestion and significant economic losses. Therefore, conducting 24 / 7 online monitoring and early warning of high-risk slopes along tourist roads has become a crucial aspect of scenic area operation and safety management.

[0003] Currently, the common early warning method for slope collapses along tourist highways involves periodically or in real-time data collection at key locations on the slope, triggering an alarm when the monitored data exceeds a preset threshold. This approach simplifies slope monitoring to a linear method of "single-point sensing + threshold judgment." This method can only capture local deformation information of the soil and rock mass at the installation location. However, slope hazards are often scattered across different elevations and horizontal sections of the slope, and micro-deformation signals in monitoring blind spots cannot be effectively captured. Consequently, the system struggles to obtain a complete sequence of precursory information before the slope enters the accelerated creep stage of instability, resulting in poor timeliness of collapse warnings. Summary of the Invention

[0004] To improve the timeliness of landslide early warning, this application provides a landslide early warning method, device, medium, and equipment based on cellular landslide equipment.

[0005] The first aspect of this application provides a landslide early warning method based on a cellular landslide device, specifically including: By using multiple preset target cellular cluster collapse devices, deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway is obtained. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least tilt sensors, strain sensors and vibration sensors. Based on the deformation monitoring data of each monitoring point, determine whether there are any signs of an impending landslide in the currently monitored road section; If there are signs of an impending landslide in the monitored road section, the risk value of a landslide in the monitored road section is determined, and auxiliary image information of the monitored road section is obtained by cameras deployed at a first preset distance before and after the monitored road section. Based on the auxiliary image information, road risk triggering factors are identified. Based on the road risk triggering factors and the risk value, a final warning level is determined. Based on the final warning level, warnings are issued via audible and visual alarms deployed at second preset distances before and after the monitored road segment. The location information of the monitored road segment, the final warning level, and the auxiliary image information are sent to the management personnel's terminal. The second preset distance is greater than the first preset distance. By using multiple target cellular cluster-type landslide monitoring devices pre-installed at various monitoring points on the slope surface, deformation monitoring data such as tilt angle, strain, and vibration are simultaneously acquired at each monitoring point, forming a continuous monitoring network covering the entire potential hazard area. Any local micro-deformation can be captured nearby, thereby eliminating the perception blind spots caused by single-point deployment and providing a complete data foundation for landslide precursor identification. Based on this, the presence of landslide precursors is determined according to the deformation monitoring data of each monitoring point. This judgment does not rely on the isolated threshold of a single sensor, but is based on a comprehensive judgment based on the spatial correlation and temporal progression of multiple monitoring points. If landslide precursors are found, the risk value is further determined, and cameras deployed at the first preset distance before and after the monitored road section are activated to acquire auxiliary image information. Based on the auxiliary image information, road risk triggering factors are identified, and the final warning level is determined in conjunction with the risk value. This allows the internal mechanical state of the slope reflected by sensor data to be cross-validated with the macroscopic phenomena of the ground surface reflected by visual information, doubly confirming the authenticity of the risk and effectively suppressing false alarms while reducing missed reports. Ultimately, based on the final warning level, the system triggers a remote warning from an audible and visual alarm located at a second preset distance before and after the monitored road section. The alarm also simultaneously sends the location information, warning level, and on-site footage to the management personnel's terminal. This allows for alarm and information synchronization before vehicles and tourists enter the danger zone, transforming traditional post-disaster emergency response into proactive early warning based on precursor signals, significantly improving the timeliness of landslide warnings.

[0006] In one implementation, determining the risk value of a landslide in the monitored road section specifically includes: From the deformation monitoring data of each monitoring point in the monitored road section, key monitoring points whose deformation exceeds a preset deformation threshold are selected, and the deformation rate of each key monitoring point is calculated based on the temporal change of the deformation monitoring data of each key monitoring point. Based on the distribution location of each key monitoring point on the slope surface, a minimum convex hull region covering all key monitoring points is constructed, and the area and shape dispersion of the minimum convex hull region are calculated. Using the deformation rate of each key monitoring point as a weight, the area of ​​the minimum convex hull region and the shape dispersion are weighted and fused to obtain a deformation development index that characterizes the development trend of deformation space. Obtain the slope geological type, slope parameters, and historical landslide records of the monitored road section, and determine the basic vulnerability coefficient of the monitored road section based on the matching relationship between the slope geological type, slope parameters, and historical landslide records. Based on the deformation development index and the basic vulnerability coefficient, the risk value of landslides in the monitored road section is obtained. The more concentrated the key monitoring points, the higher the deformation rate, and the larger the basic vulnerability coefficient, the greater the risk value.

[0007] In one implementation, determining the final warning level based on the road risk triggering factors and the risk value specifically includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; The traffic flow and tourist density of the monitored road section at the current moment are extracted from the auxiliary screen information, and the personnel exposure coefficient is determined based on the traffic flow and tourist density. The system acquires real-time meteorological data of the area where the monitored road section is located and rainfall forecast data for a preset period of time in the future. Based on the rainfall intensity and temperature change range in the real-time meteorological data and rainfall forecast data, the system determines the landslide triggering and promoting coefficient. Based on the location of the monitored road section and the distribution of surrounding emergency evacuation routes, calculate the shortest evacuation time required to reach the nearest safe area from the current location, and use the shortest evacuation time as a response tolerance parameter; The initial warning level is corrected by the personnel exposure coefficient, the landslide triggering factor, and the response tolerance parameter to obtain the final warning level. The higher the personnel exposure coefficient or the landslide triggering factor, the greater the upward correction of the initial warning level. The smaller the response tolerance parameter, the greater the upward correction of the initial warning level.

[0008] In one implementation, determining the final warning level based on the road risk triggering factors and the risk value further includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; Based on the auxiliary screen information, the proportion of tourist buses, hazardous materials transport vehicles and other vehicles in the current monitored road segment is determined, and the proportion of tourist buses, hazardous materials transport vehicles and other vehicles is weighted and summed to obtain the vehicle impact factor. The weights corresponding to the proportion of tourist buses and hazardous materials transport vehicles are higher than the weights corresponding to other vehicles. The system obtains the current number of vehicles stranded in the monitored road section and the real-time traffic status data of each available detour route within a preset range before and after the monitored road section. Based on the number of stranded vehicles, the congestion index of the monitored road segment is determined, and the remaining capacity of each available detour route is determined based on all the real-time traffic status data. The ratio of the number of stranded vehicles to the total remaining capacity is calculated, and the evacuation difficulty factor is determined based on the congestion index and the ratio. The initial warning level is revised based on the vehicle impact factor and the evacuation difficulty factor to obtain the final warning level.

[0009] In one embodiment, the method further includes: When it is determined that there are signs of a landslide in the current monitored road section, the monitoring point with signs of a landslide is identified as the target monitoring point, and the rainfall data of the area where the monitored road section is located in the past preset time period and the soil moisture content data collected by the soil moisture sensor embedded in the slope are obtained. Based on the temporal changes of the rainfall data and the soil moisture content data, the infiltration expansion rate of the slope within the preset past time period is calculated. In the target cellular cluster version of the collapse device deployed based on the target monitoring point, the sampling sequence of the tilt sensor and the strain sensor within the preset past time period is fitted to form the deformation accumulation curve of the target monitoring point, and the deformation acceleration abrupt point and its corresponding occurrence time are extracted from the deformation accumulation curve. Determine whether the occurrence time falls within the time window when the infiltration expansion rate exceeds a preset rate threshold; if it falls within the time window, then confirm that there are signs of a landslide in the currently monitored road section.

[0010] In one implementation, determining the final warning level based on the road risk triggering factors and the risk value specifically includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; Obtain multiple historical time periods during which traffic congestion occurred on the tourist highway during the same period, and select at least one time period of interest from the multiple historical time periods where traffic congestion occurred; The system obtains multiple historical road sections of the tourist highway that have experienced traffic congestion during the specified period of interest, and determines at least one traffic congestion-prone road section based on these multiple historical road sections. Based on the multiple historical time periods, a first occurrence index of vehicle congestion within the period of interest is determined, and based on the multiple historical road segments, a second occurrence index of vehicle congestion within each of the road segments of interest is determined. Based on the risk value, the estimated warning duration for a landslide to occur on the monitored road section relative to the current time is determined, and the landslide warning period is determined based on the estimated warning duration and the current time. The higher the risk value, the longer the estimated warning duration. The initial warning level is corrected based on the landslide warning period, the first occurrence index, and the second occurrence index to obtain the final warning level.

[0011] In one implementation, the step of correcting the initial warning level based on the landslide warning period, the first occurrence index, and the second occurrence index to obtain the final warning level specifically includes: If the monitored road segment is included in the road segment of the period of concern, then the period of concern is determined as the target period, and the road segment of concern that includes the monitored road segment is determined as the target segment. When the target period is included in the landslide warning period, the risk index of traffic congestion in the monitored road segment during the target period is determined based on the first occurrence index of traffic congestion during the target period and the second occurrence index of traffic congestion during the target road segment. Based on the aforementioned risk indices, the overall risk index of traffic congestion on the monitored road section during the landslide warning period is determined, and the initial warning level is corrected according to the overall risk index to obtain the final warning level.

[0012] A second aspect of this application provides a landslide early warning device based on a cellular landslide equipment, specifically comprising: The data acquisition module is used to acquire deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway through multiple preset target cellular cluster collapse devices. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least an angle sensor, a strain sensor and a vibration sensor. The precursor detection module is used to determine whether there are precursors to a landslide in the current monitored road section based on the deformation monitoring data of each monitoring point. The image acquisition module is used to determine the risk value of a landslide in the monitored road section if there are signs of an impending landslide, and to acquire auxiliary image information of the monitored road section through cameras deployed at a first preset distance before and after the monitored road section. The landslide early warning module is used to identify road risk triggering factors based on the auxiliary screen information, determine the final early warning level based on the road risk triggering factors and the risk value, and issue early warning reminders through audible and visual alarms deployed at a second preset distance before and after the monitored road section based on the final early warning level. The module also sends the location information of the monitored road section, the final early warning level, and the auxiliary screen information to the terminal of the management personnel. The second preset distance is greater than the first preset distance.

[0013] A third aspect of this application provides a computer-readable storage medium storing a computer program that, when loaded and executed by a processor, performs the steps of the method described in any one of the first aspects.

[0014] A fourth aspect of this application provides an electronic device, specifically comprising: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, the processor being configured to load and execute the computer program stored in the memory to cause the electronic device to perform the method as described in any one of the first aspects.

[0015] In summary, this application includes at least one of the following beneficial technical effects: by using multiple cellular cluster-type landslide monitoring devices deployed at various monitoring points on the slope surface, the tilt angle, strain, and vibration deformation monitoring data of each monitoring point are acquired simultaneously, forming a distributed sensing matrix covering the hazard area, eliminating blind spots in single-point monitoring, and making comprehensive judgments based on multi-point spatial correlation and temporal progression, effectively identifying precursor signals before the slope enters accelerated instability, thereby improving the timeliness of landslide early warning. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of a landslide early warning method based on a cellular landslide device provided in an embodiment of this application; Figure 2 This is a schematic diagram of the system architecture of a landslide early warning method based on a cellular landslide device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the layout of a camera and an audible and visual alarm provided in an embodiment of this application; Figure 4 This is a schematic diagram illustrating the association between a time period of interest and the corresponding road segment of interest, provided in an embodiment of this application. Figure 5 This is a schematic diagram of a landslide early warning device based on a cellular landslide equipment provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0017] Explanation of reference numerals in the attached diagram: 11. Data acquisition module; 12. Warning sign judgment module; 13. Image acquisition module; 14. Landslide early warning module; 201. Processor; 202. Memory; 203. Bus; 204. Network interface. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0019] In the description of the embodiments of this application, words such as "exemplarily," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplarily," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of words such as "exemplarily," "for example," or "for instance" is intended to present the relevant concepts in a specific manner.

[0020] In the description of the embodiments of this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, B existing alone, or A and B existing simultaneously. Furthermore, unless otherwise stated, the term "multiple" means two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and their variations all mean "including but not limited to," unless otherwise specifically emphasized.

[0021] See Figure 1 This application discloses a flowchart illustrating a landslide early warning method based on a cellular landslide device. This method can be implemented using a computer program or run on a landslide early warning device based on a cellular landslide device using a von Neumann architecture. The computer program can be integrated into an application or run as a standalone utility application, specifically including: S101: By using multiple preset target cellular cluster collapse devices, obtain deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least tilt sensors, strain sensors and vibration sensors.

[0022] Specifically, in this embodiment, multiple monitoring points are staggered along the contour lines and perpendicular to the contour lines on the slope surface of the determined monitoring section. The horizontal spacing between adjacent monitoring points is 5 to 10 meters, and the vertical spacing is 5 to 10 meters. The specific spacing is determined based on the area of ​​the potential hazard area and the complexity of the terrain. The smaller the potential hazard area or the more complex the terrain, the smaller the spacing value is used to ensure that the effective monitoring range of adjacent monitoring points covers each other and eliminates monitoring blind spots. In this embodiment, the horizontal spacing is 8 meters and the vertical spacing is 5 meters. One target cellular cluster version landslide device is deployed at each monitoring point. This device uses a cellular cluster version landslide sensor (model GH-LP5000-CE), which integrates at least three types of sensing elements: tilt sensor, strain sensor, and vibration sensor. The tilt sensor is used to collect data on the change of the slope angle at the monitoring point in real time. The strain sensor is used to collect data on the strain force of the soil and rock on the slope surface at the monitoring point in real time. The vibration sensor is used to collect data on the vibration signal of the slope at the monitoring point in real time. The three sensors collect data synchronously with a unified sampling period. Each target cellular trunking-type landslide monitoring device forms a decentralized access layer network via LoRa Mesh or Sub-GHz wireless self-organizing network protocols. Each device node acts as a relay, forming multi-hop transmission links. Data in the access layer network is aggregated by the cellular trunking gateway and then uploaded to the server via 4G or 5G cellular backhaul links for storage and analysis, completing the collection and transmission of deformation monitoring data at each monitoring point. Each target cellular trunking-type landslide monitoring device is powered by a combination of solar panels and high-capacity battery packs. The lower-power sensing devices are equipped with a 130W monocrystalline solar panel and a 160,000mAh battery, while the higher-power camera devices are equipped with a 220W monocrystalline solar panel and a 340,000mAh battery. During peak tourist seasons, they operate at a higher sampling frequency, automatically switching to a low-power inspection mode during off-peak seasons to reduce energy consumption.

[0023] It should be noted that the monitored section refers to a specific section of tourist highway identified through geological survey and assessment as having potential slope collapse hazards and requiring the deployment of monitoring equipment for real-time continuous monitoring. Deformation monitoring data refers to a comprehensive set of monitoring data, including tilt angle time-series data, strain force time-series data, and vibration waveform time-series data, collected by tilt sensors, strain sensors, and vibration sensors in the target cellular cluster-type slope collapse monitoring equipment at each monitoring point and uploaded to the server via a cellular network. The beneficial effect of this step is that by staggering multiple cellular cluster-type slope collapse monitoring devices along the slope surface to form a distributed sensing matrix, the monitoring blind spots of the traditional single-point monitoring mode are eliminated, enabling precise capture of minute deformations in various areas of the slope.

[0024] Furthermore, the landslide early warning method based on cellular landslide equipment disclosed in this application is executed by a server. The server obtains deformation monitoring data collected by the target cellular cluster landslide equipment, acquires auxiliary image information through cameras before and after the monitored road section, identifies landslide precursors and the final warning level accordingly, and sends the relevant landslide early warning information to the management personnel's terminal (personal computer or smartphone). The warning is then issued through audible and visual alarms before and after the monitored road section. For details, please refer to [link to relevant documentation]. Figure 2 .

[0025] S102: Based on the deformation monitoring data of each monitoring point, determine whether there are signs of an impending landslide in the currently monitored road section.

[0026] Specifically, after receiving the deformation monitoring data uploaded by each monitoring point, the server performs real-time analysis on the tilt sensor data, strain sensor data, and vibration sensor data of each monitoring point. First, the server calculates the rate of change of tilt angle per unit time for the tilt angle time-series data of each monitoring point and compares this rate of change with a preset tilt angle change rate threshold. Simultaneously, it calculates the strain increment per unit time for the strain force time-series data of each monitoring point and compares this strain increment with a preset strain increment threshold. Furthermore, it extracts vibration amplitude and dominant frequency characteristics from the vibration waveform time-series data of each monitoring point to determine whether low-frequency, high-amplitude vibration characteristics related to geological slippage are present. Low frequency refers to a dominant frequency below a preset lower frequency threshold (this threshold is calibrated according to the geological type of the slope; 5 to 15 Hz for rock slopes and 1 to 5 Hz for soil slopes). High amplitude refers to a vibration amplitude exceeding a preset multiple of the baseline vibration amplitude. The aforementioned preset thresholds for the rate of change of tilt angle and the threshold for the increase in strain are determined as follows: During the stable operation period after the initial deployment of the target cellular cluster version of the landslide equipment, deformation monitoring data from each monitoring point are continuously collected as baseline data. This stable operation period is the time period after the equipment deployment is completed and the slope is confirmed to be in a stable state by on-site geological assessment, lasting no less than 30 days. The mean and standard deviation of the rate of change of tilt angle and the increase in strain in the baseline data are statistically analyzed. The mean plus three times the standard deviation is used as the preset threshold for the corresponding parameters. This method of value selection is based on the assumption of normal distribution, which can ensure that false alarms are not triggered within the normal fluctuation range while maintaining sufficient detection sensitivity for abnormal changes exceeding the normal fluctuation range. When any of the following conditions are met: the rate of change of tilt angle at a monitoring point exceeds the corresponding preset threshold, or the increase in strain exceeds the corresponding preset threshold, or the vibration characteristics show low-frequency high-amplitude characteristics, the monitoring point is marked as an abnormal monitoring point. The server further analyzes the number of all abnormal monitoring points in the current monitoring section and their spatial distribution on the slope surface. A landslide precursor is determined to exist in the current monitoring section if one of the following two conditions is met: The first condition is that two or more spatially adjacent abnormal monitoring points simultaneously exceed their respective preset thresholds within the same time window. The length of this time window is set according to the slope geological type: 5 to 15 minutes for rock slopes and 15 to 30 minutes for soil slopes, because the collapse development rate of rock slopes is usually faster than the slip development rate of soil slopes. The second condition is that the deformation monitoring data of a single monitoring point shows a progressive growth trend, meaning that the rate of change of tilt angle or strain increment of the monitoring point continuously increases over three or more consecutive sampling periods, and the increase in each sampling period is not less than the increase in the previous sampling period. If neither of the above conditions is met in the current monitoring section, it is determined that there is no landslide precursor in the current monitoring section, and each target cellular cluster landslide monitoring device continues to operate at the normal sampling frequency.

[0027] S103: If there are signs of an impending landslide in the current monitored road section, determine the risk value of a landslide in the monitored road section, and obtain auxiliary image information of the monitored road section through cameras deployed at the first preset distance before and after the monitored road section.

[0028] Specifically, after identifying signs of impending landslides in the current monitored road section, the risk value of a landslide in that section is calculated, while simultaneously controlling the camera to acquire auxiliary image information. The calculation process for the risk value is as follows: First, from the deformation monitoring data of all monitoring points in the monitored road section, monitoring points whose deformation exceeds a preset deformation threshold are selected as key monitoring points. Here, the deformation is a comprehensive measure of the cumulative change in tilt angle and the cumulative change in strain at each monitoring point. The weighted sum of the two after normalization can be taken as the comprehensive deformation. The preset deformation threshold is determined in the same way as the threshold of each parameter in step S102, that is, it is determined based on the statistical characteristics of the baseline data during the stable operation period. The mean of the comprehensive deformation in the baseline data is taken plus the standard deviation of a preset multiple. The preset multiple is a value between 2 and 5 depending on the slope geological type and the required monitoring sensitivity. The more unstable the geological conditions or the higher the sensitivity requirement, the smaller the value should be.

[0029] Then, the deformation rate of each key monitoring point is calculated based on the temporal changes in deformation monitoring data of each key monitoring point. Specifically, the comprehensive deformation change value of each key monitoring point within the most recent preset time period is divided by the preset time period to obtain the deformation change per unit time, which is the deformation rate of the key monitoring point. The preset time period is determined according to the geological type of the slope. For rock slopes, the most recent 15 to 30 minutes are taken, and for soil slopes, the most recent 30 minutes to 1 hour are taken. Next, based on the planar coordinates of each key monitoring point on the slope surface, a minimum convex hull region covering all key monitoring points is constructed using a convex hull algorithm (such as the Graham scan algorithm or the Andrew monotonic chain algorithm). The area and perimeter of this minimum convex hull region are then calculated, and the shape dispersion of the minimum convex hull region is calculated. The shape dispersion is calculated as follows: the square of the perimeter of the minimum convex hull region is divided by the product of the area of ​​the region and 4π. The absolute value of the quotient minus 1 is the shape dispersion. This calculation principle is based on the isocircular inequality. When the minimum convex hull region is close to a perfect circle, the ratio of the square of its perimeter to its area is close to 4π, so the shape dispersion is close to 0, indicating that the key monitoring points are concentrated in a compact area. When the minimum convex hull region is a long strip or an irregular polygon, the ratio of the square of its perimeter to its area is much greater than 4π, and the shape dispersion is significantly greater than 0, indicating that the distribution of key monitoring points shows obvious directionality or dispersion, that is, the deformation extends along a specific direction.

[0030] Furthermore, the arithmetic mean of the deformation rates of all key monitoring points is calculated as the average deformation rate. The product of the average deformation rate, the area of ​​the minimum convex hull, and (1 plus the shape dispersion) is used as the deformation development index. The larger the deformation development index, the wider the spatial range of slope deformation, the more irregular the shape, and the faster the deformation rate, i.e., the more severe the deformation development trend. Subsequently, the server obtains the geological type, slope parameters, and historical landslide records of the slope where the monitored road section is located to determine the foundation vulnerability coefficient. Among them, the geological type comes from the rock and soil classification information recorded in the geological survey report of the road section; the slope parameters include the natural slope angle and slope height of the slope, which come from the topographic data measured on site; the historical landslide records include the number of landslide events that occurred in the road section in the past 10 years and the maximum landslide volume of each landslide, which come from the disaster archives of the highway management department. The server pre-stores a basic vulnerability coefficient scoring rule table based on the above parameters. This scoring rule table divides geological types into 3 to 5 levels according to the strength of the soil and rock mass (e.g., hard rock slope, soft rock slope, loess slope, etc.), divides the slope angle into 3 to 5 intervals according to the steepness (e.g., less than 30 degrees, 30 to 45 degrees, 45 to 60 degrees, greater than 60 degrees), and divides the frequency and scale of historical landslides into 3 to 5 levels. Each level corresponds to a sub-item score value between 0 and 1. The three sub-item score values ​​are weighted and summed with preset weights (the weights are determined according to the geological disaster risk assessment specifications of the area, and the weights of slope parameters and geological types are usually higher than the weights of historical records) to obtain the basic vulnerability coefficient. The coefficient ranges from 0 to 1, and the larger the value, the more likely the slope of the road section is to collapse. Finally, the risk value of landslides in the monitored road section is obtained by multiplying the deformation development index and the basic vulnerability coefficient. The deformation development index reflects the current dynamic development trend of deformation, and the basic vulnerability coefficient reflects the inherent landslide susceptibility of the slope. The product of the two can comprehensively characterize the current landslide risk level.

[0031] While determining the risk value, the server sends image acquisition commands to two cameras positioned at a first preset distance, one in front of and one behind the monitored road section. The first preset distance is 3 to 10 meters; in this embodiment, it is 5 meters. The two cameras are 4G high-definition cameras with a resolution of 4 megapixels or higher, installed 5 meters in front of the starting point and 5 meters behind the ending point of the monitored road section, respectively, arranged in a facing-to-facing manner. This ensures that their respective fields of view cover the entire monitored road section and the slope area above it. The captured video footage and image frames are uploaded to the server as auxiliary image information for subsequent steps. For details, please refer to [link to relevant documentation]. Figure 3 , Figure 3This diagram illustrates the deployment of cameras and audible / visual alarms. The deformation rate reflects the speed of slope deformation development at the monitoring point; the minimum convex hull region characterizes the spatial distribution range of areas experiencing significant deformation; shape dispersion refers to the degree to which the shape of the minimum convex hull region deviates from a perfect circle, calculated based on the isoperimetric inequality principle; and the deformation development index reflects the comprehensive development trend of slope deformation in terms of spatial range, expansion morphology, and rate. The foundation vulnerability coefficient reflects the inherent susceptibility to landslides on the slope of this road section. This step quantifies the spatial distribution range of deformation by selecting key monitoring points and constructing the minimum convex hull region. Simultaneously, it introduces shape dispersion to characterize the morphological features of the deformation area. This ensures that risk assessment considers not only the amplitude and rate of deformation but also the spatial expansion range and directional characteristics of deformation, improving the accuracy of the risk value in representing the actual landslide development trend.

[0032] S104: Based on the auxiliary screen information, identify the road risk triggering factors, determine the final warning level based on the road risk triggering factors and risk values, and issue warning reminders through audible and visual alarms deployed at the second preset distance before and after the monitored road section according to the final warning level. The location information of the monitored road section, the final warning level, and the auxiliary screen information are sent to the terminal of the management personnel. The second preset distance is greater than the first preset distance.

[0033] Specifically, in this embodiment, the process of identifying road risk triggering factors is as follows: The server uses a pre-trained deep learning visual model to analyze each frame of the auxiliary image information in real time, identifying the following five types of road risk triggering factors. The first type is suspended rocks, which means that there are rock blocks on the slope surface that have separated from the parent rock due to cracks or weathering and erosion, forming a suspended state. The rock block area is located by a target detection model (such as YOLOv8) and the rock mass contour is extracted by an edge detection algorithm. The contact area ratio between the rock block and the bedrock, the suspended length, and the offset of the rock block's center of gravity projection position relative to the supporting boundary are calculated to determine the severity of the suspension. The second type is falling rocks blocking the road, which means that rocks roll down from the slope onto the road surface, causing the lane to be partially or completely blocked. The falling rock area and the road surface area are distinguished by an instance segmentation model (such as Mask R-CNN). The proportion of the falling rock area to the road surface area, the distribution location of the falling rocks, and the remaining passable width of the road are calculated. The third category is road surface cracking and subsidence, which refers to irregular cracks or local depressions in the road surface caused by slope slippage or roadbed erosion. Pixel-level crack segmentation models (such as U-Net or DeepCrack) are used to detect the length, width, and direction of the cracks, and temporal frame difference or optical flow methods are combined to detect local displacement and subsidence area. The fourth category is dense fog, which is a foggy weather condition that significantly reduces visibility and affects driving safety. Visibility estimation algorithms based on dark channel priors or deep learning defogging networks (such as DehazeNet) are used to infer the current visibility distance and fog concentration index. The fifth category is dust, which is a dust phenomenon caused by mountain weathering or slight slope slippage. Optical flow methods are used to track the movement trajectory of dust particles and analyze the rate of change and duration of dust area changes to determine the source and severity of dust.

[0034] Further, based on the identified road risk triggering factors and risk values, an initial warning level is determined. The specific process is as follows: First, according to a preset risk level mapping table (containing different risk intervals and corresponding warning levels), the calculated risk values ​​are divided into corresponding risk intervals, with each interval mapping to a basic warning level (such as green, yellow, orange, and red). Then, the basic warning level is adjusted unidirectionally or bidirectionally based on the road risk triggering factors and their severity. If factors directly threatening traffic, such as suspended rocks, falling rocks obstructing the road, or road surface cracking and subsidence, are identified, the warning level is adjusted upwards (increased by 1 or 2 levels). If only factors affecting visibility but not directly causing landslides, such as heavy fog or dust, are identified, the warning level remains unchanged or is only marked as a warning. Then, using a target detection model, vehicles and pedestrians in the auxiliary image information are detected and counted. The traffic flow of the monitored road segment at the current moment, i.e., the number of vehicles passing through per unit time, and the tourist density, i.e., the ratio of the number of visible pedestrians to the length of the monitored road segment, are extracted. The personnel exposure coefficient is determined based on traffic flow and tourist density. Specifically, the traffic flow is normalized by dividing the traffic flow by the historical maximum traffic flow of the road segment to obtain a normalized traffic flow value (0 to 1). The tourist density is normalized by dividing the tourist density by the maximum pedestrian density corresponding to the designed capacity of the road segment to obtain a density normalized value (0 to 1). The two values ​​are then weighted and summed using preset weights to obtain the personnel exposure coefficient. The weights are set according to the traffic characteristics of tourist roads. For tourist roads where tourist buses and pedestrians are the main traffic objects, the weight of tourist density is higher than that of traffic flow. The server simultaneously acquires the current real-time meteorological data and the rainfall forecast data within a preset time period for the monitored road segment area. The real-time meteorological data comes from meteorological monitoring stations or public meteorological service interfaces near the monitored road segment and includes at least the current rainfall intensity and recent temperature changes. The preset time period is taken as 2 to 6 hours in the future, based on the response time window for landslide disasters. The landslide triggering factor is determined based on the current rainfall intensity, the cumulative rainfall in the rainfall forecast, and the temperature change range. The greater the rainfall intensity, the higher the expected continuous rainfall, or the greater the temperature change (such as the severe freeze-thaw cycle that accelerates rock weathering), the greater the landslide triggering factor. This factor ranges from 0 to 1 and is determined by the server based on a pre-established meteorological parameter classification table. The server also calculates the shortest evacuation time required from the monitored road section to the nearest safe area based on the geographical coordinates of the monitored road section and the distribution of surrounding emergency evacuation routes (derived from evacuation route data in the emergency plan maintained by the highway management department). The nearest safe area is the open area, parking lot, or service area closest to the monitored road section that is not threatened by the current slope collapse. The shortest evacuation time is obtained by dividing the road distance from the road section to the safe area by the estimated driving speed under current road conditions, and the shortest evacuation time is used as a response tolerance parameter.Finally, the initial warning level is corrected using the personnel exposure coefficient, landslide triggering factor, and response tolerance parameter. The correction rules are as follows: when the personnel exposure coefficient exceeds its preset correction threshold (e.g., 0.6), the initial warning level is corrected by 1 level; when the landslide triggering factor exceeds its preset correction threshold (e.g., 0.5), the initial warning level is corrected by 1 level; when the response tolerance parameter, i.e., the shortest evacuation time, is less than the preset safe evacuation time threshold (e.g., 10 minutes, which is set by management personnel based on road length and traffic speed), it indicates insufficient evacuation time, and the initial warning level is corrected by 1 level; when two or more of the above three correction conditions are met simultaneously, the initial warning level is corrected by 2 levels; when all three correction conditions are met simultaneously, the initial warning level is corrected by 3 levels, and the corrected warning level does not exceed the red evacuation level; if the level was already red evacuation level before correction, the highest level remains unchanged, and the corrected level is the final warning level. By employing a correction mechanism based on three dimensions—personnel exposure coefficient, landslide triggering and promotion coefficient, and response tolerance parameter—the warning level not only reflects the severity of the landslide disaster itself but also comprehensively considers actual factors affecting the severity of the disaster consequences, such as population density, external meteorological conditions, and safe evacuation conditions. This significantly improves the scientific rigor and relevance of the warning decision-making process.

[0035] Furthermore, based on the final warning level, alarm commands of the corresponding level are sent to two audible and visual alarms deployed at a second preset distance before and after the monitored road section. The second preset distance is greater than the first preset distance; in this embodiment, the second preset distance is 100 meters. The two RS485 audible and visual alarms are respectively installed on the roadside 100 meters before the entrance and 100 meters after the exit of the monitored road section. For details, please refer to [link / reference]. Figure 3 The audible and visual alarms activate corresponding alarm modes based on the warning level: Yellow (Attention) corresponds to flashing yellow lights and a voice prompt to slow down; Orange (Alert) corresponds to flashing orange lights and a voice prompt to proceed with caution; and Red (Evacuation) corresponds to flashing red lights and a voice prompt to stop due to landslide. Simultaneously, the server pushes the location information of the monitored road section, the final warning level, and auxiliary video information to the management personnel's terminals via the 4G network. The terminals display the warning information in a pop-up window and automatically display real-time camera footage, allowing management personnel to remotely confirm the on-site situation and decide whether to initiate emergency response measures such as traffic control and tourist evacuation.

[0036] In one embodiment, a vehicle classification and detection model is used to classify and identify vehicles passing through and lingering in the current monitored road segment for each frame of the auxiliary image information. The vehicles are divided into three categories: tourist buses, dangerous goods transport vehicles, and other vehicles. Tourist buses are identified by their body appearance features (such as body length, window density, and roof rack), while dangerous goods transport vehicles are identified by their dangerous goods identification signs and tank shape features. Vehicles that do not belong to the above two categories are classified into the other vehicle category. Based on the proportion of each type of vehicle on the current monitored road section, the proportions of tourist buses, hazardous materials transport vehicles, and other vehicles are determined. The weighted sum of these three proportions yields the vehicle impact factor. The weights corresponding to the proportions of tourist buses and hazardous materials transport vehicles are higher than those corresponding to the proportions of other vehicles. This is because tourist buses have a large passenger capacity and pose a higher risk of casualties in the event of a landslide, while hazardous materials transport vehicles may cause secondary disasters such as explosions and leaks if damaged. The potential consequences of both are significantly more severe than those of ordinary vehicles. In this embodiment, the weight of the proportion of tourist buses is set to 0.5, the weight of the proportion of hazardous materials transport vehicles is set to 0.35, and the weight of the proportion of other vehicles is set to 0.15.

[0037] The server further obtains the number of currently stranded vehicles in the monitored road segment. This number is calculated by frame-by-frame detection and tracking of vehicles in the auxiliary visual information, and by counting the number of vehicles whose dwell time within the monitored road segment exceeds a preset dwell time threshold (e.g., 2 minutes). Based on the number of stranded vehicles, a congestion index is determined for the monitored road segment. The congestion index equals the number of stranded vehicles divided by the maximum number of vehicles the monitored road segment can accommodate under the current number of lanes and road segment length. The congestion index ranges from 0 to 1, where 0 indicates no congestion and 1 indicates the road segment is completely blocked. Simultaneously, the server obtains real-time traffic status data for each available detour route within a preset range (e.g., 5 kilometers before and after) before and after the monitored road segment through a traffic management data interface. Based on the current traffic flow and designed capacity of each detour route, the remaining capacity of each detour route is calculated. The remaining capacity equals the designed capacity minus the current traffic flow. The server calculates the ratio of the number of stranded vehicles to the total remaining capacity of all available detour routes. This ratio reflects the supply and demand relationship between the number of vehicles currently needing evacuation and the available evacuation capacity. The evacuation difficulty factor is determined by combining the congestion index and this ratio. The higher the congestion index or the larger the ratio of the number of stranded vehicles to the total remaining traffic capacity, the greater the evacuation difficulty factor, indicating that it is more difficult for vehicles on the currently monitored road section to be safely evacuated in a short period of time. Finally, the initial warning level is corrected based on the vehicle impact factor and the evacuation difficulty factor. When the vehicle impact factor exceeds the preset impact threshold (e.g., 0.4, which is set by the management personnel according to the safety management requirements of tourist highways) and the evacuation difficulty factor exceeds the preset evacuation threshold (e.g., 0.5), the initial warning level is corrected upward by at least one level to obtain the final warning level. When only one of them exceeds the corresponding threshold, the initial warning level remains unchanged, but a prompt indicating vehicle composition or evacuation difficulty is added to the warning information pushed to the management personnel's terminal. If the level was already the highest level before correction, it remains unchanged.

[0038] It should be noted that the vehicle impact factor is a dimensionless coefficient reflecting the severity of the potential disaster consequences for the vehicle composition on the currently monitored road segment. The congestion index reflects the degree of congestion on the road segment. The evacuation difficulty factor is a comprehensive evaluation index reflecting the difficulty of evacuating vehicles on the currently monitored road segment. By distinguishing vehicle types and assigning differentiated weights to calculate the vehicle impact factor, the early warning correction can reflect the differences in the consequences of landslides for different vehicle types. The characteristics of tourist buses with large passenger volumes and dangerous goods vehicles that may cause secondary disasters are fully considered. By introducing the congestion index and the remaining capacity of detour routes to construct the evacuation difficulty factor, the early warning level is expanded from simply reflecting the severity of geological disasters to also considering the actual conditions for safe and timely evacuation of vehicles, making the early warning decision more in line with the actual scenario of dense traffic flow and limited detour conditions on tourist highways.

[0039] In one embodiment, when it is determined that there are signs of impending landslide on the current monitored road section, the server identifies the monitoring point that triggered the landslide as the target monitoring point and further cross-validates the landslide precursor using rainfall data and soil moisture content data to confirm the reliability of the precursor. Specifically, the server first obtains rainfall data for the area where the monitored road section is located over a preset period of time. This preset period is determined based on the slope geological type and the time scale of rainfall infiltration. For loess slopes, the data is taken from the past 24 to 72 hours, and for rock slopes, it is taken from the past 12 to 48 hours. The rainfall data comes from hourly rainfall records provided by meteorological monitoring stations or public meteorological service interfaces near the monitored road section. At the same time, the server obtains soil moisture content data collected by soil moisture sensors embedded at different depths inside the slope where the monitored road section is located. The soil moisture sensors are deployed at different depths below the slope surface (e.g., 0.5 meters, 1.0 meters, and 1.5 meters) along the potential sliding surface depth direction of the slope to obtain time-series soil moisture content data at different depths. The server calculates the infiltration expansion rate of the slope over a preset period of time based on the temporal changes of rainfall data and soil moisture content data. The infiltration expansion rate is calculated as follows: the difference between the start times of moisture content rise of two soil moisture sensors at adjacent depths is taken as the time difference. The depth difference between the two sensors is divided by this time difference to obtain the infiltration depth value of rainfall water from shallow to deep layers. This value is the infiltration expansion rate. The larger the infiltration expansion rate, the faster the rainfall water infiltrates into the interior of the slope, the more rapidly the slope moisture content rises, and the higher the risk of the slope's soil and rock strength decreasing and pore water pressure increasing due to the increased moisture content.

[0040] Based on all sampling sequences from the tilt and strain sensors in the target cellular cluster-type landslide monitoring equipment deployed at the target monitoring point over a preset time period, the server uses a curve fitting method (such as least squares polynomial fitting or spline interpolation) to fit the sampling time as the horizontal axis and the combined value of the cumulative change in tilt angle and the cumulative change in strain as the vertical axis to form a cumulative deformation curve for the target monitoring point. This curve reflects the cumulative development process of slope deformation at the target monitoring point over time. The server calculates the second derivative of the cumulative deformation curve to obtain the time series of deformation acceleration. In this time series, it searches for inflection points where the deformation acceleration suddenly changes from a stable or slow change to a rapid increase. These inflection points are identified as abrupt change points in deformation acceleration, and their corresponding occurrence times are recorded. The criteria for identifying abrupt change points in deformation acceleration are that the change in deformation acceleration before and after this moment exceeds a preset multiple (e.g., 3 times) of the average acceleration value of the stable segment before the abrupt change. This preset multiple is determined based on the slope geological type and the statistical characteristics of historical deformation data.

[0041] The server compares the occurrence time of the deformation acceleration abrupt change with the time series of the seepage expansion rate to determine whether the occurrence time falls within the time window when the seepage expansion rate exceeds a preset rate threshold. The preset rate threshold is determined based on the permeability characteristics of the slope's soil and rock mass, and can be taken as a preset proportion of the permeability coefficient of the slope's soil and rock mass under saturation (e.g., 60% to 80%). When the seepage expansion rate reaches or approaches the saturated permeability capacity of the soil and rock mass, it indicates that the slope is undergoing a rapid seepage process. If the occurrence time of the deformation acceleration abrupt change does indeed fall within the time window when the seepage expansion rate exceeds the preset rate threshold, it indicates that there is a causal relationship between the acceleration of slope deformation and the rapid increase in slope moisture content caused by rainfall infiltration. Based on this, it is confirmed that there is a precursor to a landslide induced by rainfall infiltration in the current monitored section. This precursor has passed the time series correlation verification of rainfall-moisture content-deformation, and the subsequent risk value calculation and final warning level determination process can continue.

[0042] In another embodiment, a feasible way to determine the final warning level is as follows: After determining the initial warning level, based on pre-cached tourist highway congestion records, obtain multiple historical time periods (which may contain identical time periods) that have experienced traffic congestion on tourist highways during the same period. These tourist highway congestion records include, but are not limited to, historical time periods and road segments where traffic congestion occurred within a historical period, with the historical period being the past 5 years. For example, if the current period is a holiday or a specific month, then the historical time periods on tourist highways during the same period can be understood as the time periods during the same holiday or month within the same historical period that have experienced traffic congestion.

[0043] The frequency of a single historical time period repeating across multiple historical time periods is counted. If the frequency exceeds a preset frequency threshold, the historical time period is identified as a period of concern for vehicle congestion, i.e., a historical time period where congestion is likely to occur on tourist highways, and at least one such period exists. Then, based on the aforementioned tourist highway congestion records, multiple historical road segments that experienced vehicle congestion within a single period of concern are obtained (these segments may be identical; the road segments are pre-defined segments for tourist highways). The number of times a single historical segment repeats across these multiple historical segments is counted. If the number of repetitions exceeds a preset frequency threshold, the historical segment is identified as a road segment of concern for vehicle congestion, i.e., a road segment prone to congestion during the period of concern.

[0044] Furthermore, a first occurrence index for vehicle congestion within a single attention period is determined. This first occurrence index is the ratio of the frequency of recurrence within a single attention period to the sum of the frequencies of recurrence across all attention periods, representing the probability of vehicle congestion occurring within that single attention period. Then, a second occurrence index is determined for each attention road segment corresponding to the single attention period. This second occurrence index is the ratio of the number of times a particular attention road segment recurs to the sum of the numbers of times all attention road segments recur, representing the probability of vehicle congestion occurring within that attention road segment. See [link to details] for more information. Figure 4 , Figure 4 This is a diagram showing the relationship between time periods of interest and corresponding road segments of interest. There are time periods of interest E and F. Time period E corresponds to road segments of interest E1 and E2; time period F corresponds to road segments of interest F1 and F2.

[0045] Furthermore, using a preset mapping table, based on the risk value, the estimated warning duration for a landslide on the monitored road section is determined relative to the current time. The higher the risk value, the greater the likelihood of a landslide, and the longer the time after the current time, the more vigilant one needs to be regarding the possibility of a landslide. The mapping table includes intervals for different risk values ​​and corresponding warning duration ranges. In this embodiment, a duration is randomly selected from the warning duration range matched to the risk value as the estimated warning duration. Specifically, the settings are: when the risk value is in the range of 0 to 0.2, the estimated duration is less than 1 hour; when the risk value is in the range of 0.2 to 0.4, the estimated duration is 1 to 6 hours; when the risk value is in the range of 0.4 to 0.6, the estimated duration is 6 to 24 hours; when the risk value is in the range of 0.6 to 0.8, the estimated duration is 24 to 72 hours; and when the risk value is greater than 0.8, the estimated duration is greater than 72 hours. The estimated warning duration is added to the current time to determine the landslide warning period.

[0046] If a monitored road segment is included in the monitored road segment within a single attention period, then that attention road segment is designated as the target time period, i.e., the time period during which congestion is likely to occur on the monitored road segment. Simultaneously, the attention road segment containing the monitored road segment is also designated as the target road segment. If the landslide warning period includes a target time period, then the first occurrence index of the target time period is multiplied by the second occurrence index of the target road segment to obtain a risk index for traffic congestion on the monitored road segment during the target time period, representing the probability of traffic congestion. For each target time period included within the landslide warning period, a corresponding risk index is determined according to this logic. Then, the various risk indices are summed to obtain the overall risk index for traffic congestion on the monitored road segment during the landslide warning period, representing the overall probability of congestion on the monitored road segment during the entire landslide warning period. Finally, using the overall risk index, the corresponding warning level is matched from a preset index-level matching table as the final warning level. The index-level matching table includes warning levels corresponding to different index intervals. The warning level corresponding to the index interval in which the overall risk index falls is determined as the final warning level; the higher the overall risk index, the higher the corresponding warning level.

[0047] In another embodiment, the first occurrence index of a single attention period is multiplied by the second occurrence index of each corresponding attention road segment to obtain multiple products corresponding to the single attention period. Each product represents the probability of congestion occurring in the corresponding attention road segment within a single attention period. The products corresponding to the same attention road segment are summed across all attention periods to obtain a congestion index, representing the overall probability of congestion occurring in the attention road segment. Based on pre-cached tourist highway landslide records, the number of landslides occurring within the attention road segment is counted. This number is divided by the total number of landslides in the historical period to obtain the landslide index for that attention road segment, representing the risk of a landslide. The tourist highway landslide records include the locations of landslides that occurred on tourist highways within the historical period. The congestion index and landslide index of the same attention road segment are multiplied to obtain an urgency index for landslide monitoring of that attention road segment. N attention road segments are selected from all attention road segments in descending order of urgency index, and these are determined as the monitoring road segments requiring focused attention. In this embodiment, N is the quotient of the total number of attention road segments divided by 2.

[0048] The implementation principle of the landslide early warning method based on cellular landslide equipment in this application embodiment is as follows: multiple cellular cluster landslide equipment deployed at various monitoring points on the slope surface synchronously acquire the tilt angle, strain, and vibration deformation monitoring data of each monitoring point, forming a distributed sensing matrix covering the hidden danger area, eliminating single-point monitoring blind spots, and making comprehensive judgments based on multi-point spatial correlation and temporal progression, effectively identifying precursor signals before the slope enters accelerated instability, thereby improving the timeliness of landslide early warning.

[0049] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0050] Please see Figure 5 This is a schematic diagram of the structure of a landslide early warning device based on a cellular landslide equipment provided in an embodiment of this application. This landslide early warning device, applied to a cellular landslide equipment, can be implemented as all or part of a device through software, hardware, or a combination of both. The device includes a data acquisition module 11, a precursor judgment module 12, an image acquisition module 13, and a landslide early warning module 14.

[0051] The data acquisition module 11 is used to acquire deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway through multiple preset target cellular cluster collapse devices. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least an angle sensor, a strain sensor and a vibration sensor. The precursor detection module 12 is used to determine whether there are precursors to a landslide in the current monitored road section based on the deformation monitoring data of each monitoring point. The image acquisition module 13 is used to determine the risk value of a landslide in the monitored road section if there are signs of a landslide in the current monitored road section, and to acquire auxiliary image information of the monitored road section through cameras deployed at a first preset distance before and after the monitored road section. The landslide early warning module 14 is used to identify road risk triggering factors based on auxiliary screen information, determine the final early warning level based on the road risk triggering factors and risk values, and issue early warning reminders through audible and visual alarms deployed at a second preset distance before and after the monitored road section according to the final early warning level. It also sends the location information of the monitored road section, the final early warning level, and auxiliary screen information to the terminal of the management personnel. The second preset distance is greater than the first preset distance.

[0052] It should be noted that the landslide early warning device based on a cellular landslide device provided in the above embodiments, when executing the landslide early warning method based on a cellular landslide device, is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the landslide early warning device based on a cellular landslide device and the landslide early warning method embodiment based on a cellular landslide device provided in the above embodiments belong to the same concept, and the implementation process is detailed in the method embodiment, which will not be repeated here.

[0053] This application also discloses a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements a landslide early warning method based on a cellular landslide device as described in the above embodiments.

[0054] The computer program can be stored in a computer-readable medium. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or certain middleware. The computer-readable medium includes any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the computer-readable medium includes, but is not limited to, the above-mentioned components.

[0055] The above-described method for early warning of landslides based on cellular landslide equipment is stored in the computer-readable storage medium and loaded and executed on the processor to facilitate the storage and application of the above method.

[0056] like Figure 6 As shown in the embodiments of this application, an electronic device is also disclosed. The electronic device adopts electronic devices such as servers, and the electronic device includes, but is not limited to, a processor 201 and a memory 202. For example, the electronic device may also include a bus 203 and a network interface 204.

[0057] The processor 201 can be a central processing unit (CPU). Of course, depending on the actual use, it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc., and this application does not limit it in this regard.

[0058] The memory 202 can be an internal storage unit of an electronic device, such as a hard disk or memory of the electronic device, or an external storage device of the electronic device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD) or flash memory card (FC) equipped on the electronic device. Furthermore, the memory 202 can also be a combination of an internal storage unit and an external storage device of the electronic device. The memory is used to store computer programs and other programs and data required by the electronic device. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0059] In this electronic device, a landslide early warning method based on a cellular landslide device according to the above embodiment is stored in the memory 202 of the electronic device and loaded and executed on the processor 201 of the electronic device for convenient use.

[0060] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A landslide early warning method based on a cellular landslide device, characterized in that, The method includes: By using multiple preset target cellular cluster collapse devices, deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway is obtained. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least tilt sensors, strain sensors and vibration sensors. Based on the deformation monitoring data of each monitoring point, determine whether there are any signs of an impending landslide in the currently monitored road section; If there are signs of an impending landslide in the monitored road section, the risk value of a landslide in the monitored road section is determined, and auxiliary image information of the monitored road section is obtained by cameras deployed at a first preset distance before and after the monitored road section. Based on the auxiliary screen information, road risk triggering factors are identified. Based on the road risk triggering factors and the risk value, the final warning level is determined. Based on the final warning level, warning reminders are issued by sound and light alarms deployed at a second preset distance before and after the monitored road section. The location information of the monitored road section, the final warning level, and the auxiliary screen information are sent to the terminal of the management personnel. The second preset distance is greater than the first preset distance.

2. The landslide early warning method based on cellular landslide equipment according to claim 1, characterized in that, Determining the risk value of a landslide in the monitored road section specifically includes: From the deformation monitoring data of each monitoring point in the monitored road section, key monitoring points whose deformation exceeds a preset deformation threshold are selected, and the deformation rate of each key monitoring point is calculated based on the temporal change of the deformation monitoring data of each key monitoring point. Based on the distribution location of each key monitoring point on the slope surface, a minimum convex hull region covering all key monitoring points is constructed, and the area and shape dispersion of the minimum convex hull region are calculated. Using the deformation rate of each key monitoring point as a weight, the area of ​​the minimum convex hull region and the shape dispersion are weighted and fused to obtain a deformation development index that characterizes the development trend of deformation space. Obtain the slope geological type, slope parameters, and historical landslide records of the monitored road section, and determine the basic vulnerability coefficient of the monitored road section based on the matching relationship between the slope geological type, slope parameters, and historical landslide records. Based on the deformation development index and the basic vulnerability coefficient, the risk value of landslides in the monitored road section is obtained. The more concentrated the key monitoring points, the higher the deformation rate, and the larger the basic vulnerability coefficient, the greater the risk value.

3. The landslide early warning method based on cellular landslide equipment according to claim 1, characterized in that, The determination of the final warning level based on the road risk triggering factors and the risk value specifically includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; The traffic flow and tourist density of the monitored road section at the current moment are extracted from the auxiliary screen information, and the personnel exposure coefficient is determined based on the traffic flow and tourist density. The system acquires real-time meteorological data of the area where the monitored road section is located and rainfall forecast data for a preset period of time in the future. Based on the rainfall intensity and temperature change range in the real-time meteorological data and rainfall forecast data, the system determines the landslide triggering and promoting coefficient. Based on the location of the monitored road section and the distribution of surrounding emergency evacuation routes, calculate the shortest evacuation time required to reach the nearest safe area from the current location, and use the shortest evacuation time as a response tolerance parameter; The initial warning level is corrected by the personnel exposure coefficient, the landslide triggering factor, and the response tolerance parameter to obtain the final warning level. The higher the personnel exposure coefficient or the landslide triggering factor, the greater the upward correction of the initial warning level. The smaller the response tolerance parameter, the greater the upward correction of the initial warning level.

4. The landslide early warning method based on cellular landslide equipment according to claim 1, characterized in that, The determination of the final warning level based on the road risk triggering factors and the risk value further includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; Based on the auxiliary screen information, the proportion of tourist buses, hazardous materials transport vehicles and other vehicles in the current monitored road segment is determined, and the proportion of tourist buses, hazardous materials transport vehicles and other vehicles is weighted and summed to obtain the vehicle impact factor. The weights corresponding to the proportion of tourist buses and hazardous materials transport vehicles are higher than the weights corresponding to other vehicles. The system obtains the current number of vehicles stranded in the monitored road section and the real-time traffic status data of each available detour route within a preset range before and after the monitored road section. Based on the number of stranded vehicles, the congestion index of the monitored road segment is determined, and the remaining capacity of each available detour route is determined based on all the real-time traffic status data. The ratio of the number of stranded vehicles to the total remaining capacity is calculated, and the evacuation difficulty factor is determined based on the congestion index and the ratio. The initial warning level is revised based on the vehicle impact factor and the evacuation difficulty factor to obtain the final warning level.

5. The landslide early warning method based on a cellular landslide device according to claim 1, characterized in that, The method further includes: When it is determined that there are signs of a landslide in the current monitored road section, the monitoring point with signs of a landslide is identified as the target monitoring point, and the rainfall data of the area where the monitored road section is located in the past preset time period and the soil moisture content data collected by the soil moisture sensor embedded in the slope are obtained. Based on the temporal changes of the rainfall data and the soil moisture content data, the infiltration expansion rate of the slope within the preset past time period is calculated. In the target cellular cluster version of the collapse device deployed based on the target monitoring point, the sampling sequence of the tilt sensor and the strain sensor within the preset past time period is fitted to form the deformation accumulation curve of the target monitoring point, and the deformation acceleration abrupt point and its corresponding occurrence time are extracted from the deformation accumulation curve. Determine whether the occurrence time falls within the time window when the infiltration expansion rate exceeds a preset rate threshold; if it falls within the time window, then confirm that there are signs of a landslide in the currently monitored road section.

6. The landslide early warning method based on cellular landslide equipment according to claim 1, characterized in that, The determination of the final warning level based on the road risk triggering factors and the risk value specifically includes: Based on the road risk triggering factors and the risk values, determine the initial warning level corresponding to the monitored road segment; Obtain multiple historical time periods during which traffic congestion occurred on the tourist highway during the same period, and select at least one time period of interest from the multiple historical time periods where traffic congestion occurred; The system obtains multiple historical road sections of the tourist highway that have experienced traffic congestion during the specified period of interest, and determines at least one traffic congestion-prone road section based on these multiple historical road sections. Based on the multiple historical time periods, a first occurrence index of vehicle congestion within the period of interest is determined, and based on the multiple historical road segments, a second occurrence index of vehicle congestion within each of the road segments of interest is determined. Based on the risk value, the estimated warning duration for a landslide to occur on the monitored road section relative to the current time is determined, and the landslide warning period is determined based on the estimated warning duration and the current time. The higher the risk value, the longer the estimated warning duration. The initial warning level is corrected based on the landslide warning period, the first occurrence index, and the second occurrence index to obtain the final warning level.

7. The landslide early warning method based on cellular landslide equipment according to claim 6, characterized in that, The step of revising the initial warning level based on the landslide warning period, the first occurrence index, and the second occurrence index to obtain the final warning level specifically includes: If the monitored road segment is included in the road segment of the period of concern, then the period of concern is determined as the target period, and the road segment of concern that includes the monitored road segment is determined as the target segment. When the target period is included in the landslide warning period, the risk index of traffic congestion in the monitored road segment during the target period is determined based on the first occurrence index of traffic congestion during the target period and the second occurrence index of traffic congestion during the target road segment. Based on the aforementioned risk indices, the overall risk index of traffic congestion on the monitored road section during the landslide warning period is determined, and the initial warning level is corrected according to the overall risk index to obtain the final warning level.

8. A landslide early warning device based on a cellular landslide equipment, used to implement the landslide early warning method based on a cellular landslide equipment as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module (11) is used to acquire deformation monitoring data of each monitoring point on the slope surface of the monitored section of the tourist highway through a plurality of preset target cellular cluster collapse devices. The target cellular cluster collapse devices are cellular cluster collapse devices deployed at the corresponding monitoring points. The target cellular cluster collapse devices include at least an inclination sensor, a strain sensor and a vibration sensor. The precursor judgment module (12) is used to determine whether there are precursors to landslides in the current monitored road section based on the deformation monitoring data of each monitoring point. The image acquisition module (13) is used to determine the risk value of a landslide in the monitored road section if there are signs of a landslide in the monitored road section, and to acquire auxiliary image information of the monitored road section through cameras deployed at a first preset distance before and after the monitored road section. The landslide early warning module (14) is used to identify road risk triggering factors based on the auxiliary screen information, determine the final early warning level based on the road risk triggering factors and the risk value, and issue early warning reminders through audible and visual alarms deployed at a second preset distance before and after the monitored road section based on the final early warning level, and send the location information of the monitored road section, the final early warning level and the auxiliary screen information to the terminal of the management personnel. The second preset distance is greater than the first preset distance.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is loaded and executed by the processor, it implements the method of any one of claims 1-7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor loads and executes the computer program, it implements the method of any one of claims 1-7.

Images