Mudslide emergency supervision system, method and medium based on internet of things large model

Abstract

This invention provides a debris flow emergency monitoring system, method, and medium based on an IoT big data model. The system includes an emergency monitoring management platform. The emergency monitoring management platform is configured to execute a debris flow emergency monitoring method based on an IoT big data model. The method includes: dividing a target area into multiple sub-regions; determining enhanced multimodal data for each sub-region at preset intervals based on the original multimodal data of each sub-region and the positional relationships between the multiple sub-regions; determining an independent risk value for each sub-region based on the enhanced multimodal data; determining a first risk value for each sub-region based on the independent risk value and the positional relationships; and generating a data collection command based on the first risk value, downstream population density, and enhanced multimodal data to control a drone to collect data based on the data collection command. This method can be executed after the computer instructions stored on a computer-readable storage medium are read.

Description

Technical Field

[0001] This specification relates to the field of debris flow monitoring, and in particular to a debris flow emergency monitoring system, method and medium based on a large Internet of Things (IoT) model. Background Technology

[0002] Debris flows are a serious natural disaster characterized by their suddenness, high velocity, large volume, large material capacity, and strong destructive power, causing significant casualties and property damage.

[0003] However, the formation and development of debris flows are influenced by a variety of factors. Monitoring rainfall or displacement alone cannot accurately analyze complex geological changes, resulting in low accuracy and timeliness of monitoring results. This makes it impossible to take effective measures to prevent and control risks in advance, posing a great threat to people's lives and property.

[0004] Therefore, there is a need to provide a debris flow emergency monitoring system, method, and medium based on the Internet of Things (IoT) big data model. By conducting multi-dimensional monitoring and integrating and analyzing the monitoring data, it is possible to effectively monitor and warn of debris flow occurrence trends, and provide accurate and reliable risk prevention and control guidance for relevant departments and personnel. Summary of the Invention

[0005] The invention includes a debris flow emergency monitoring system based on an Internet of Things (IoT) big data model. The system comprises an emergency monitoring and management platform. The platform is configured to: divide a target area into multiple sub-regions; at preset intervals, determine enhanced multimodal data for each sub-region based on the original multimodal data of each sub-region and the positional relationships between the sub-regions; determine an independent risk value for each sub-region based on the enhanced multimodal data; determine a first risk value for each sub-region based on the independent risk value and the positional relationships; and generate a data collection command based on the first risk value, downstream population density, and the enhanced multimodal data, and send it to the internal sensing and control platform of the emergency monitoring system to control a drone to collect data based on the data collection command, wherein the data collection command includes a data collection path, data collection points, and data collection volume.

[0006] The invention includes a debris flow emergency monitoring method based on an Internet of Things (IoT) big data model. The method is executed on an emergency monitoring and management platform and includes: dividing a target area into multiple sub-regions; determining enhanced multimodal data for each sub-region at preset intervals based on the original multimodal data of each sub-region and the positional relationships between the sub-regions; determining an independent risk value for each sub-region based on the enhanced multimodal data; determining a first risk value for each sub-region based on the independent risk value and the positional relationships; and generating a data collection instruction based on the first risk value, downstream population density, and the enhanced multimodal data, and sending it to an internal perception and control platform for emergency monitoring to control a drone to collect data based on the collection instruction, wherein the collection instruction includes a collection path, collection points, and collection volume.

[0007] The invention includes a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes a debris flow emergency monitoring method based on an Internet of Things (IoT) big data model.

[0008] Beneficial effects: By conducting multi-dimensional monitoring and integrating and analyzing the monitoring data, it is possible to effectively monitor and warn of debris flow occurrence trends, providing accurate and reliable risk prevention and control guidance for relevant departments and personnel. Adjusting the parameters during the drone data collection process based on the first risk value can reduce the flight cost and flight time of the drone when collecting data. Attached Figure Description

[0009] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0010] Figure 1 This is a platform structure diagram of a debris flow emergency monitoring system based on an Internet of Things (IoT) big data model, as shown in some embodiments of this specification.

[0011] Figure 2 This is an exemplary flowchart of a debris flow emergency monitoring method based on an IoT big data model, as shown in some embodiments of this specification.

[0012] Figure 3 This is an exemplary flowchart illustrating the generation of a spraying command according to some embodiments of this specification; and

[0013] Figure 4 This is an exemplary schematic diagram of a risk assessment model shown according to some embodiments of this specification. Detailed Implementation

[0014] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0015] Unless the context clearly indicates an exception, words such as "a," "an," "a kind," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0016] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0017] Figure 1 This is a platform structure diagram of a debris flow emergency monitoring system based on an Internet of Things (IoT) big data model, as shown in some embodiments of this specification.

[0018] In some embodiments, such as Figure 1 As shown, the debris flow emergency monitoring system 100 based on the Internet of Things large model includes an emergency monitoring user platform 110, an emergency monitoring service platform 120, an emergency monitoring management platform 130, an emergency monitoring sensor network platform 140, and an emergency monitoring perception and control platform 150 that interact sequentially.

[0019] In some embodiments, the emergency monitoring user platform 110 includes a government monitoring user platform 111 and a citizen user platform 112. The government monitoring user platform 111 is a platform for government monitoring users to monitor the operation of the entire system, and the government monitoring users can be personnel from the safety management department; the citizen user platform 112 is a platform used to obtain emergency monitoring notifications and early warning information.

[0020] In some embodiments, the emergency monitoring user platform 110 can be a terminal device.

[0021] The 120 Emergency Supervision Service Platform is a platform that provides supervision services to users who need supervision services.

[0022] In some embodiments, the emergency monitoring service platform 120 may be configured in a processor and / or server.

[0023] In some embodiments, the emergency monitoring service platform 120 interacts upward with the emergency monitoring user platform 110 and downward with the emergency monitoring management platform 130.

[0024] The Emergency Monitoring and Management Platform 130 is a platform that coordinates and integrates the connections and collaborations between various functional platforms, gathers all the information of the Internet of Things, and provides sensing, management and control functions for the Internet of Things operation system.

[0025] In some embodiments, the emergency monitoring and management platform 130 may be configured in a processor and / or server. The emergency monitoring and management platform 130 may include a database. A database refers to a database used to store monitoring data. For example, the database may be used to store raw multimodal data, interpolation models, etc.

[0026] In some embodiments, the emergency monitoring and management platform 130 is configured to divide the target area into multiple sub-regions; at preset intervals, based on the original multimodal data of each sub-region and the positional relationship between the multiple sub-regions, determine the enhanced multimodal data of each sub-region; based on the enhanced multimodal data, determine the independent risk value of each sub-region; based on the independent risk value and the positional relationship, determine the first risk value of each sub-region; and based on the first risk value, downstream population density, and enhanced multimodal data, generate a collection instruction and send it to the emergency monitoring internal perception and control platform 151 to control the drone to collect data based on the collection instruction.

[0027] The emergency monitoring sensor network platform 140 is a functional platform for managing sensor communication. In some embodiments, the emergency monitoring sensor network platform 140 can realize the functions of sensing communication for perception information and sensing communication for control information.

[0028] In some embodiments, the emergency monitoring sensor network platform 140 may be configured as a communication device and / or gateway, etc.

[0029] In some embodiments, the emergency monitoring sensor network platform 140 interacts upward with the emergency monitoring management platform 130 and downward with the emergency monitoring perception and control platform 150.

[0030] The emergency monitoring and control platform 150 is a functional platform for generating sensing information and executing control information. In some embodiments, the emergency monitoring and control platform 150 includes an internal emergency monitoring and control platform 151 and an external emergency monitoring and control platform 152.

[0031] In some embodiments, the emergency monitoring internal sensing and control platform 151 includes multiple sensors disposed in the target area, the multiple sensors being configured to collect raw multimodal data of the target area, the multiple sensors including at least one of an image sensor, a rain sensor, an anemometer, a conductivity sensor, a spectrometer, etc.

[0032] In some embodiments, the emergency monitoring internal sensing and control platform 151 also includes a drone. The drone is configured to collect data based on a collection command, dispense flocculant based on a first dispensing command, dispense flocculant based on a second dispensing command, etc.

[0033] In some embodiments, the emergency monitoring external sensing control platform 152 includes multiple independent external sensing data interface modules for collecting relevant data of the target area. These external sensing data interface modules include, but are not limited to, a weather forecast sensing data interface module (for obtaining weather forecast information for the target area from meteorological departments), a regional population density sensing data interface module (for obtaining downstream population density data of the target area from relevant departments such as civil affairs), and a geological structure sensing data interface module (for obtaining geological data of the target area from geological departments).

[0034] For more information on the various platforms within the 100-platform debris flow emergency monitoring system based on the Internet of Things (IoT) big data model, please refer to [link / reference]. Figures 2-4 The relevant explanations are in the text.

[0035] In some embodiments of the present invention, the debris flow emergency monitoring system based on the Internet of Things big data model can form an information operation closed loop among various functional platforms, operate in a coordinated and regular manner, and can efficiently and accurately predict the risk of debris flows in different sub-regions.

[0036] It should be noted that the above description of the debris flow emergency monitoring system and platform based on the Internet of Things (IoT) big data model is for ease of description only and should not be construed as limiting this specification to the scope of the embodiments described. It is understood that those skilled in the art, after understanding the principles of this system, may arbitrarily combine the various platforms or construct subsystems that connect with other platforms without departing from these principles.

[0037] Figure 2 This is an exemplary flowchart illustrating an emergency monitoring method for debris flow based on a large-scale Internet of Things (IoT) model, according to some embodiments of this specification. Figure 2 As shown, process 200 includes the following steps. In some embodiments, process 200 may be executed by an emergency monitoring and management platform.

[0038] Step 210: Divide the target area into multiple sub-regions.

[0039] The target area refers to the area that requires emergency monitoring for debris flows. For example, the target area can be a valley, a hillside, etc.

[0040] In some embodiments, the emergency monitoring and management platform can identify target areas based on historical data, specifically areas where mudslides have previously occurred.

[0041] In some embodiments, the emergency monitoring and management platform can also identify areas where mudslides are more likely to occur as target areas based on experience. For example, the target area could be an area that has experienced heavy rainfall.

[0042] The target area can also be determined by any other feasible method.

[0043] In some embodiments, the emergency monitoring and management platform can obtain a topographic map of the target area via the Internet and divide the target area into multiple sub-regions based on geographical location and altitude. For example, when the target area is a hillside, the emergency monitoring and management platform can identify multiple sub-regions based on the different mountain peaks (e.g., east peak, south peak, west peak, north peak), mountainside, and foot of the mountain.

[0044] In some embodiments, the emergency monitoring and management platform may execute steps 220 to 250 at preset intervals.

[0045] In some embodiments, the emergency monitoring and management platform can set a preset period based on experience. For example, the preset period can be 5 minutes, 10 minutes, etc.

[0046] Step 220: Based on the original multimodal data of each sub-region and the positional relationship between the multiple sub-regions, determine the enhanced multimodal data of each sub-region.

[0047] Raw multimodal data refers to unprocessed data of various types related to the target region.

[0048] In some embodiments, the raw multimodal data includes multiple types of data corresponding to each sub-region at multiple times. The raw multimodal data is acquired from both the internal emergency monitoring perception and control platform and the external emergency monitoring perception and control platform based on the emergency monitoring sensor network platform, and then aggregated and processed by the emergency monitoring management platform. For example, the raw multimodal data includes geological information, climate information, hydraulic information, etc., corresponding to each sub-region at multiple times.

[0049] Geological information is information related to the composition and structure of strata in a subregion. For example, geological information includes the geological type, topography, and geomorphological information of the subregion.

[0050] Climate information refers to information related to the weather in a sub-region. For example, climate information includes the amount of rainfall, wind direction, and wind speed in the sub-region.

[0051] Hydraulic information refers to information related to the distribution of moisture in a sub-region. For example, hydraulic information includes the groundwater level and soil moisture in the sub-region.

[0052] In some embodiments, the emergency monitoring and management platform can acquire raw multimodal data through multiple sensors installed in the target area. For example, the sensors may include image sensors, rain sensors, anemometers, conductivity sensors, spectrometers, etc.

[0053] In some embodiments, the raw multimodal data also includes plant stress data and stratigraphic data.

[0054] Plant stress data refers to data used to reflect the impact of changes in the natural environment on plants. For example, plant stress data includes physiological signals of target plants in a subregion.

[0055] Target plants are indicator plants that are sensitive to minute soil displacements or changes in soil moisture. Examples of target plants include oxalis, chickweed, and rush.

[0056] Physiological signals are signals released by plants during their growth. For example, physiological signals include leaf spectra.

[0057] In some embodiments, the emergency monitoring and management platform can continuously collect leaf spectra of target plants using hyperspectral cameras located in the target area.

[0058] Stratigraphic data refers to data related to geological formations. For example, stratigraphic data includes vibration and sound data within the formations.

[0059] In some embodiments, the emergency monitoring and management platform can collect ground data through sensing optical cables laid inside the ground.

[0060] In some embodiments, monitoring plant stress data and stratigraphic data can detect minute changes in the soil, improving the accuracy and timeliness of debris flow emergency monitoring.

[0061] Positional relationships refer to the relative relationships between different sub-regions in space. For example, positional relationships include the elevation difference between sub-regions, whether they are directly connected, and their relative orientation.

[0062] In some embodiments, the emergency monitoring and management platform can determine the positional relationships between multiple sub-regions based on a topographic map of the target area.

[0063] In some embodiments, the original multimodal data may be missing data at a certain time or of a certain type, and it is necessary to supplement the missing data.

[0064] Enhanced multimodal data refers to complete multimodal data after supplementing the original multimodal data. Enhanced multimodal data includes multimodal data of all time periods and all types.

[0065] In some embodiments, for each data-missing sub-region, the emergency monitoring and management platform can determine the missing data location in the original multimodal data corresponding to the data-missing sub-region and supplement the missing data location through linear interpolation.

[0066] A missing data sub-region refers to a sub-region where multimodal data is missing. A missing data location refers to the location where the missing multimodal data is situated, including both the time of the data loss and the type of data loss.

[0067] In some embodiments, in response to the data missing location corresponding to the data missing sub-region being data missing time 1 and data missing type a, the emergency monitoring and management platform can supplement the data missing location by linear interpolation based on multiple multimodal data of data missing type a adjacent to data missing time 1 in the original multimodal data.

[0068] In some embodiments, in response to the data missing location corresponding to the data missing sub-region being all times and data missing type b, the emergency monitoring platform can determine multiple adjacent sub-regions of the data missing sub-region based on location relationships, obtain multimodal data of data missing type b of the multiple adjacent sub-regions at all times, and supplement the data missing location through linear interpolation. The quantity and density of the multimodal data determined by linear interpolation can be the mean quantity and mean density of the multimodal data of data missing type b of the multiple adjacent sub-regions, respectively. The density of the multimodal data refers to the quantity of a certain type of multimodal data per unit time.

[0069] In some embodiments, the emergency monitoring and management platform can also determine the missing data location and multiple adjacent sub-regions corresponding to the missing data sub-region based on the original multimodal data and location relationships; and determine the enhanced multimodal data corresponding to the missing data sub-region through an interpolation model based on the original multimodal data corresponding to the missing data sub-region, the missing data location, and the original multimodal data of multiple adjacent sub-regions.

[0070] An interpolation model (also known as a large interpolation model) is a model used to determine augmented multimodal data through linear interpolation. In some embodiments, the interpolation model is a machine learning model. For example, the interpolation model is a neural network (NN) model, a large model, or other user-defined models and any combination thereof.

[0071] The input to the interpolation model includes the original multimodal data corresponding to the missing data sub-regions, the location of the missing data, and the original multimodal data of multiple adjacent sub-regions. The output of the interpolation model includes the enhanced multimodal data corresponding to the missing data sub-regions.

[0072] In some embodiments, the emergency monitoring and management platform can train an interpolation model based on multiple first training samples with first labels. The platform can input the first training samples into the initial interpolation model, construct a loss function based on the output of the initial interpolation model and the first labels, iteratively update the parameters of the initial interpolation model based on the loss function, and terminate the iteration when the iteration termination condition is met, thus obtaining the trained interpolation model. The iterative update method includes, but is not limited to, gradient descent, and the iteration termination condition can be the convergence of the loss function or the reaching of a threshold number of iterations.

[0073] The first training sample includes multimodal data of samples corresponding to the missing data sub-regions, the missing data locations of samples, and the original multimodal data of multiple adjacent sub-regions of samples. The first label includes the enhanced multimodal data corresponding to the missing data sub-regions of samples.

[0074] In some embodiments, the emergency monitoring and management platform can identify sub-regions in the original multimodal data where no data is missing based on historical data, define these sub-regions as sample data missing sub-regions, and designate the original multimodal data as the first label. The emergency monitoring and management platform can randomly remove one or more locations of multimodal data from the original multimodal data corresponding to the sample data missing sub-regions to obtain sample multimodal data, where the one or more locations are the sample data missing locations. The emergency monitoring and management platform can also identify multiple adjacent sub-regions of the sample data missing sub-regions as sample adjacent sub-regions, obtaining original multimodal data for multiple sample adjacent sub-regions.

[0075] Some embodiments in this specification use interpolation models to determine enhanced multimodal data, which can more accurately and quickly supplement missing data in the original multimodal data, making the independent risk values ​​determined subsequently based on the enhanced multimodal data more accurate.

[0076] In some embodiments, the emergency monitoring and management platform can sequentially supplement the missing data locations corresponding to multiple missing data sub-regions using the above method to obtain enhanced multimodal data for each sub-region.

[0077] Step 230: Based on the enhanced multimodal data, determine the independent risk value for each sub-region.

[0078] A risk value is an indicator used to measure the probability of a debris flow occurring in a sub-region. In some embodiments, the risk value can be represented by a numerical value, with a higher value indicating a greater probability of a debris flow.

[0079] Independent risk value refers to the risk value determined based on enhanced multimodal data of a single sub-region.

[0080] In some embodiments, the emergency monitoring and management platform can determine the independent risk value of a sub-region based on enhanced multimodal data of a single sub-region through cluster analysis. For example, the emergency monitoring and management platform can obtain historical enhanced multimodal data of multiple sub-regions based on historical data, construct a first cluster vector corresponding to each sub-region based on the historical enhanced multimodal data, and use whether a debris flow occurred within a preset period after the corresponding time in the historical enhanced multimodal data as the label corresponding to the first cluster vector, and construct a first cluster vector library based on multiple first cluster vectors and their corresponding labels.

[0081] The emergency monitoring and management platform can construct a first target vector based on enhanced multimodal data of the current sub-region. Based on the first target vector and multiple first clustering vectors, it performs clustering to obtain multiple clusters, and identifies the cluster containing the first target vector as the first target cluster. The platform calculates the proportion of first clustering vectors labeled "debris flow occurred" in the first target cluster to all first clustering vectors in the target cluster, and determines this proportion as the independent risk value of the current sub-region. Clustering methods include, but are not limited to, K-Means clustering algorithm and DBSCAN clustering algorithm.

[0082] Step 240: Determine the first risk value for each sub-region based on the independent risk value and location relationship.

[0083] The first risk value refers to the risk value determined based on enhanced multimodal data from multiple sub-regions.

[0084] In some embodiments, for each sub-region, the emergency monitoring and management platform can determine the directly connected neighboring sub-regions based on location relationships, and determine the average of the independent risk values ​​of the sub-region and its neighboring sub-regions as the first risk value of the sub-region.

[0085] In some embodiments, the emergency monitoring and management platform can construct a debris flow risk map based on multiple independent risk values ​​and location relationships corresponding to multiple sub-regions; based on the debris flow risk map, multiple first risk values ​​corresponding to multiple sub-regions are determined through a risk assessment model. For more details, see [link to relevant documentation]. Figure 4 And related explanations.

[0086] Step 250: Based on the first risk value, downstream population density, and enhanced multimodal data, a data collection instruction is generated and sent to the emergency monitoring internal perception and control platform to control the drone to collect data based on the data collection instruction.

[0087] Downstream population density reflects the population living conditions below a target area. For example, downstream population density includes downstream population density and downstream building distribution density.

[0088] In some embodiments, the emergency monitoring and management platform can use drones to collect images of the area below the target region and determine the downstream population density through image recognition. Image recognition methods include convolutional neural networks, target detection algorithms, etc. In some embodiments, the emergency monitoring and management platform can also obtain the downstream population density of the target region from the civil affairs department through an emergency monitoring external sensing and control platform.

[0089] Data acquisition commands are instructions used to control a drone to collect data. For example, data acquisition commands include the acquisition path, acquisition points, and acquisition volume.

[0090] The data acquisition path refers to the movement path of the drone during the data acquisition process. Data acquisition points refer to the locations where the drone collects data. Data acquisition volume refers to the amount of data collected at each data acquisition point.

[0091] In some embodiments, the emergency monitoring and management platform can determine the collection path, collection points, and collection volume by querying a first preset table based on a first risk value, downstream population density, and enhanced multimodal data.

[0092] The first preset table includes the first risk value, downstream population density, the correspondence between enhanced multimodal data and collection path, collection points and collection volume. The first preset table can be preset by technical personnel based on experience.

[0093] In some embodiments, the drone can move along a collection path and collect data according to the collection volume once it reaches the collection point.

[0094] In some embodiments, the drone can collect other data related to the occurrence of debris flows. For example, the drone can collect vegetation cover data, and the emergency monitoring and management platform can determine a second risk value for each sub-area based on the vegetation cover data. Based on the second risk value, a dispersal command is generated and sent to the internal perception and control platform of the emergency monitoring system to control the drone to dispersal flocculant according to the dispersal command. See more details. Figure 3 And related explanations.

[0095] Some embodiments in this specification, through multi-dimensional monitoring and fusion analysis of monitoring data, can effectively monitor and warn of debris flow occurrence trends, providing accurate and reliable risk prevention and control guidance for relevant departments and personnel. Adjusting parameters during the drone data collection process based on a first risk value can reduce the flight cost and flight time of the drone when collecting data.

[0096] Figure 3 This is an exemplary flowchart illustrating the generation of a spraying command according to some embodiments of this specification. Figure 3 As shown, process 300 includes the following steps. In some embodiments, process 300 may be executed by an emergency monitoring and management platform.

[0097] Step 310: Based on the vegetation cover data, determine the vegetation cover characteristics of each sub-region.

[0098] Vegetation cover data refers to image data that reflects the vegetation cover situation.

[0099] In some embodiments, vegetation cover data can be collected by drones and uploaded to an emergency monitoring and management platform.

[0100] Vegetation cover characteristics refer to data reflecting vegetation distribution and health status. Vegetation distribution includes vegetation type and its corresponding area. Health status includes the presence of withered or unhealthy plants and the proportion of withered or unhealthy plants.

[0101] In some embodiments, the emergency monitoring and management platform can determine vegetation cover characteristics and vegetation cover rate based on vegetation cover data and through image recognition.

[0102] Step 320: Based on vegetation cover characteristics and geological information, determine the vegetation impact characteristics of each sub-region.

[0103] For more information on geological information, please refer to the relevant instructions in step 220.

[0104] Vegetation impact characteristics refer to data used to reflect the influence of vegetation on debris flows. For example, vegetation impact characteristics include impact values; a positive impact value indicates that vegetation exacerbates debris flows, while a negative impact value indicates that vegetation blocks debris flows. The larger the absolute value of the impact value, the stronger the exacerbation or blocking effect.

[0105] In some embodiments, the emergency monitoring and management platform can determine the vegetation impact characteristics of each sub-region through cluster analysis based on vegetation cover characteristics and geological information. For example, the platform can obtain historical vegetation cover characteristics and historical geological information of multiple historical sub-regions before debris flows occurred based on historical data, construct a second cluster vector corresponding to each historical sub-region based on the historical vegetation cover characteristics and historical geological information, and determine the flow rate and velocity of debris flows that occurred after the second cluster vector as the labels corresponding to the second cluster vector. A second cluster vector library is constructed based on multiple second cluster vectors and their corresponding labels.

[0106] The emergency monitoring and management platform can construct a second target vector based on the vegetation cover characteristics and geological information of the current sub-region, perform clustering based on the second target vector and multiple second clustering vectors to obtain multiple clusters, and determine the cluster containing the second target vector as the second target cluster.

[0107] In some embodiments, the emergency monitoring and management platform calculates the difference between the debris flow rate and the reference flow rate, and the difference between the debris flow velocity and the reference velocity for each second cluster vector in the second target cluster. The weighted sum of all differences is then used to determine the impact value corresponding to the second target vector. The weights for the weighted sum can be set based on experience.

[0108] Reference flow rate and reference velocity refer to the flow rate and velocity of debris flows in the absence of vegetation influence.

[0109] In some embodiments, the emergency monitoring and management platform can determine the reference flow rate and reference velocity corresponding to the second cluster vector by querying a second preset table based on the geological information corresponding to the second cluster vector.

[0110] The second preset table includes the correspondence between geological information and reference flow rate and reference velocity. The second preset table can be set by technicians based on experience.

[0111] Step 330: Based on the vegetation impact characteristics and the first risk value, determine the second risk value for each sub-region.

[0112] The second risk value is the risk value determined by taking into account the characteristics of vegetation impact.

[0113] For more information on risk values ​​and the first risk value, please refer to the relevant instructions in step 230.

[0114] In some embodiments, the emergency monitoring and management platform normalizes the impact value and the first risk value and then performs a weighted summation, determining the weighted sum as the second risk value. The normalization method includes Min-Max normalization, among others.

[0115] Step 340: Based on the second risk value, generate a dispensing instruction and send it to the emergency monitoring internal perception and control platform to control the drone to dispense flocculant based on the dispensing instruction.

[0116] Flocculants can cause colloidal particles in the soil to form flocculent precipitates, thereby reducing soil fluidity and thus reducing the risk of debris flows.

[0117] The dispensing command (also known as the first dispensing command) is a command used to control the drone to dispense flocculant. For example, the first dispensing command includes the dispensing path, dispensing point (also known as the first dispensing point), and dispensing amount (also known as the first dispensing amount).

[0118] In some embodiments, the spraying path, the first spraying point, and the first spraying amount can be preset values ​​for the system. For example, the spraying path can be a circular path that encircles the sub-region, and the first spraying point can include multiple points evenly distributed along the spraying path.

[0119] In some embodiments, the emergency monitoring and management platform can determine the spraying path and the first spraying point based on the topographic information and vegetation cover data of each sub-region.

[0120] Geomorphological information refers to information related to the geomorphology of a sub-region. For example, geomorphological information includes soil looseness at various locations within the sub-region.

[0121] In some embodiments, the emergency monitoring and management platform can obtain geomorphological information from raw multimodal data.

[0122] In some embodiments, the emergency monitoring and management platform can determine multiple locations with soil looseness greater than a preset looseness threshold or vegetation coverage less than a preset coverage threshold based on topographic information and vegetation cover data, and use these locations as first spreading points. Based on the first spreading points, a spreading path is generated using shortest path planning algorithms, etc. The shortest path planning algorithms include, but are not limited to, Dijkstra's algorithm, Floyd's algorithm, etc.

[0123] Some embodiments in this specification determine the application path and first application point based on topographic information and vegetation cover data, enabling precise application of flocculants to areas prone to debris flows, thereby effectively improving the application effect and reducing the risk of debris flows.

[0124] In some embodiments, in response to a second risk value in a sub-region being greater than a first preset risk threshold, the emergency monitoring and management platform generates a first dispensing instruction and sends it to the emergency monitoring internal perception and control platform to control the drone to move in the sub-region based on the dispensing path. When it reaches the first dispensing point, it dispenses flocculant according to the first dispensing amount.

[0125] In some embodiments of this specification, considering the impact of vegetation on the velocity and flow rate of debris flows, the first risk value is adjusted based on the characteristics of vegetation impact to obtain a more accurate second risk value. When the risk of debris flow is high, flocculants can be sprayed by drones to reduce the fluidity of sediment, thereby reducing the probability of debris flow formation and protecting the lives and property of downstream residents to the greatest extent possible.

[0126] In some embodiments, in response to multiple first risk values ​​not meeting preset conditions, the emergency monitoring and management platform generates a second dispensing instruction and sends it to the emergency monitoring internal perception and control platform when generating the collection instruction, so as to control the drone to dispense flocculant based on the second dispensing instruction during the data collection process.

[0127] The preset conditions are used to determine whether to generate a second scattering instruction. In some embodiments, the preset conditions may be that multiple first risk values ​​are all less than a second preset risk threshold. The first and second preset risk thresholds may be set based on experience.

[0128] In some embodiments, different sub-regions correspond to different preset conditions, and the preset conditions may include multiple preset sub-conditions, with one preset sub-condition corresponding to one sub-region. A preset sub-condition may be that the first risk value of the sub-region is less than the corresponding second preset risk threshold.

[0129] In some embodiments, the emergency monitoring and management platform can also determine preset sub-conditions based on multiple first risk values ​​of multiple adjacent sub-regions. For example, the emergency monitoring and management platform can determine a second preset risk threshold based on the average of multiple first risk values ​​of multiple adjacent sub-regions; the larger the average, the smaller the second preset risk threshold.

[0130] A debris flow in a certain sub-region can affect adjacent areas where no debris flow has occurred. Some embodiments of this specification adjust the second preset risk threshold in the corresponding preset sub-conditions of a sub-region based on the first risk value of adjacent sub-regions. This improves the sensitivity of regulation, allowing for timely intervention and control of the sub-region, which helps to delay or prevent debris flows and protect the lives and property of residents.

[0131] The second dispensing command is used to control the drone to dispense flocculant during data collection. For example, the second dispensing command may include a second dispensing point and a second dispensing amount.

[0132] In some embodiments, the second spraying points can be evenly distributed along the drone's collection path. The second spraying amount can be set based on actual needs.

[0133] In some embodiments, during the data collection process, the drone sprays flocculant according to the second spraying amount each time it reaches the second spraying point.

[0134] Some embodiments in this specification control the drone to simultaneously spray flocculant during data collection, which can reduce the drone's flight cost and flight time, effectively improve the timeliness of control, and reduce the risk of debris flow.

[0135] It should be noted that the above descriptions of processes 200 and 300 are for illustrative purposes only and do not limit the scope of this specification. Those skilled in the art can make various modifications and changes to processes 200 and 300 under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0136] Figure 4 This is an exemplary schematic diagram of a risk assessment model shown according to some embodiments of this specification.

[0137] In some embodiments, such as Figure 4 As shown, the emergency monitoring and management platform can construct a debris flow risk map 430 based on multiple independent risk values ​​410 and location relationships 420 corresponding to multiple sub-regions; based on the debris flow risk map 430, multiple first risk values ​​450 corresponding to multiple sub-regions are determined through a risk assessment model 440.

[0138] For more information on independent risk value, positional relationship, and first risk value, please refer to [link to relevant information]. Figure 2 And related explanations.

[0139] A debris flow risk map is a graph used to reflect the risk of debris flows occurring in various sub-regions. The debris flow risk map is a directed graph, consisting of multiple nodes and multiple edges connecting these nodes.

[0140] Each node in the debris flow risk map corresponds to a sub-region. The node characteristic is the independent risk value of the sub-region.

[0141] In debris flow risk maps, edges are used to represent the positional relationships between nodes. For example, if the sub-regions corresponding to two nodes are adjacent sub-regions, then the two nodes are connected by an edge, with the direction of the edge pointing from the upstream node to the downstream node, that is, from the node corresponding to the sub-region with higher elevation to the node corresponding to the sub-region with lower elevation.

[0142] In some embodiments, the edge also has corresponding edge features. The edge features include the elevation difference between the sub-regions corresponding to the two nodes connected by the edge.

[0143] In some embodiments, the emergency monitoring and management platform can determine the elevation difference between two adjacent sub-regions based on their location relationships. For more information on location relationships, please refer to the relevant description in step 220.

[0144] A risk assessment model (also known as a large risk assessment model) is a model used to evaluate the first risk value corresponding to each node in a debris flow risk map. In some embodiments, the risk assessment model is a machine learning model, such as a graph neural network (GNN) model, a large model, or other user-defined models, or any combination thereof.

[0145] The input to the risk assessment model includes a debris flow risk map. The output of the risk assessment model includes the first risk value corresponding to each node in the debris flow risk map.

[0146] In some embodiments, the emergency monitoring and management platform can train a risk assessment model based on multiple second training samples with second labels. The training method for the risk assessment model is similar to that for the difference model; see the relevant description in step 220 for more details.

[0147] The second training sample includes multiple historical debris flow risk maps corresponding to different historical target areas, and can be obtained based on historical data. The second label includes the first risk value corresponding to each node in the historical debris flow risk map.

[0148] In some embodiments, the same historical debris flow risk map corresponds to multiple different historical moments. For each node in the historical debris flow risk map, the emergency monitoring and management platform can determine, based on historical data, whether a debris flow will occur at that node within a preset period after each historical moment, and determine the ratio of the number of historical moments in which debris flows occurred to the total number of historical moments as the first risk value corresponding to that node.

[0149] In some embodiments, the debris flow risk map also includes multiple edge weights corresponding to multiple edges. The emergency monitoring and management platform can determine multiple debris flow characteristics corresponding to multiple sub-regions based on geomorphological information, geological information and surface runoff information; and determine multiple edge weights based on multiple debris flow characteristics and location relationships.

[0150] For more information on geological information, please see [link to geological information]. Figure 2 And related explanations. For more information on topography, please see [link / reference]. Figure 3 And related explanations.

[0151] Surface runoff information refers to information related to water flow on the Earth's surface. For example, surface runoff information includes the location and direction of rivers, streams, and other watercourses on the surface.

[0152] In some embodiments, the emergency monitoring and management platform can use drones to collect images of the target area and determine surface runoff information through image recognition.

[0153] Debris flow characteristics refer to the features associated with the flow of debris flows during their occurrence. For example, debris flow flow characteristics include the direction of flow (e.g., southeast, southwest, etc.).

[0154] In some embodiments, the emergency monitoring and management platform can determine multiple debris flow characteristics corresponding to multiple sub-regions through cluster analysis based on geomorphological information, geological information, and surface runoff information. For example, the emergency monitoring and management platform can obtain geomorphological information, geological information, and surface runoff information of multiple historical sub-regions based on historical data, construct a third cluster vector, determine the historical flow direction when debris flows occurred in the historical sub-regions corresponding to the third cluster vector as the label corresponding to the third cluster vector, and construct a third cluster vector library based on multiple third cluster vectors and their corresponding labels.

[0155] The emergency monitoring and management platform can construct a third target vector based on the geomorphological, geological, and surface runoff information of the current sub-region. It then performs clustering based on this third target vector and multiple third clustering vectors, resulting in multiple clusters. The cluster containing the third target vector is then identified as the third target cluster. The platform can then take the union of the labels corresponding to all third clustering vectors within the third target cluster and use this union as the debris flow characteristic for the current sub-region.

[0156] Edge weights reflect the probability that a debris flow will flow from an upstream node to a downstream node when the upstream node of an edge experiences a debris flow. The larger the edge weight, the greater the probability that a debris flow will flow from the upstream node (e.g., sub-region 1) to a downstream node (e.g., the adjacent sub-region 2 of sub-region 1) when the upstream node of that edge experiences a debris flow.

[0157] In some embodiments, in response to the debris flow direction in the debris flow feature corresponding to the upstream node (e.g., sub-region 1) including a direction from the upstream node to the downstream node (adjacent sub-region 2), the edge weight of the edge from the upstream node (sub-region 1) to the downstream node (adjacent sub-region 2) is a high value. In response to the debris flow direction in the debris flow feature corresponding to the upstream node (e.g., sub-region 1) not including a direction from the upstream node to the downstream node (adjacent sub-region 3), the edge weight of the edge from the upstream node (sub-region 1) to the downstream node (adjacent sub-region 3) is a low value. The edge weight can be set based on requirements.

[0158] Some embodiments in this specification, by determining the edge weights in the debris flow risk map, can take into account the impact of debris flows occurring in upstream nodes on downstream nodes, making the first risk value determined by the risk assessment model more accurate.

[0159] Some embodiments in this specification demonstrate that by constructing a debris flow risk map and using a risk assessment model, the first risk value corresponding to all sub-regions can be determined quickly and accurately, resulting in higher data processing efficiency and accuracy.

[0160] This specification provides an embodiment of a debris flow emergency monitoring device based on an Internet of Things (IoT) big data model. The device includes at least one processor and at least one memory. The at least one memory is used to store computer instructions. The at least one processor is used to execute at least a portion of the computer instructions to implement the debris flow emergency monitoring method based on the IoT big data model as described above.

[0161] Some embodiments of this specification provide a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the debris flow emergency monitoring method based on the Internet of Things (IoT) big data model as described above.

[0162] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0163] It should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be consistent with the teachings of this specification, rather than as examples or limitations. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A debris flow emergency monitoring system based on an Internet of Things (IoT) big data model, characterized in that, The system includes an emergency monitoring and management platform; The emergency monitoring and management platform is configured as follows: Divide the target area into multiple sub-regions; Every preset period, based on the original multimodal data of each subregion in the plurality of subregions and the positional relationship between the plurality of subregions, the enhanced multimodal data of each subregion is determined; the original multimodal data includes plant stress data and stratigraphic data, wherein the plant stress data refers to data used to reflect the impact of changes in the natural environment on plants; Based on the enhanced multimodal data, an independent risk value is determined for each sub-region; Determining a first risk value for each sub-region based on the independent risk value and the location relationship includes: for each sub-region, determining the adjacent sub-regions directly connected to the sub-region based on the location relationship, and determining the average of the independent risk values ​​of the sub-region and its adjacent sub-regions as the first risk value of the sub-region; Based on vegetation cover data, the vegetation cover characteristics of each sub-region are determined; Based on the vegetation cover characteristics and geological information, the vegetation impact characteristics of each sub-region are determined; Based on the vegetation impact characteristics and the first risk value, a second risk value is determined for each sub-region; Based on the second risk value, a dispensing instruction is generated and sent to the emergency monitoring internal perception and control platform to control the drone to dispense flocculant according to the dispensing instruction. The dispensing instruction includes a dispensing path, dispensing points, and dispensing volume. The dispensing path and dispensing points are determined based on the topographic information of each sub-region and the vegetation cover data. Based on the first risk value, downstream population density, and the enhanced multimodal data, a collection instruction is generated and sent to the emergency monitoring internal perception and control platform to control the drone to collect data based on the collection instruction. The collection instruction includes the collection path, collection point, and collection volume.

2. The system as described in claim 1, characterized in that, The emergency monitoring and management platform is also configured as follows: Based on the multiple independent risk values ​​corresponding to the multiple sub-regions and the location relationships, a debris flow risk map is constructed. Based on the debris flow risk map, multiple first risk values ​​corresponding to the multiple sub-regions are determined by a risk assessment model, wherein the risk assessment model is a machine learning model.

3. The system as described in claim 2, characterized in that, The debris flow risk map also includes multiple edge weights corresponding to multiple edges, and the emergency monitoring and management platform is further configured as follows: Based on geomorphological information, geological information, and surface runoff information, the flow characteristics of multiple debris flows corresponding to the multiple sub-regions are determined. Based on the multiple debris flow flow characteristics and the positional relationships, the multiple edge weights are determined.

4. A debris flow emergency monitoring method based on an Internet of Things (IoT) big data model, characterized in that, The method is executed based on an emergency monitoring and management platform and includes: Divide the target area into multiple sub-areas; Every preset period, based on the original multimodal data of each subregion in the plurality of subregions and the positional relationship between the plurality of subregions, the enhanced multimodal data of each subregion is determined; the original multimodal data includes plant stress data and stratigraphic data, wherein the plant stress data refers to data used to reflect the impact of changes in the natural environment on plants; Based on the enhanced multimodal data, an independent risk value is determined for each sub-region; Determining a first risk value for each sub-region based on the independent risk value and the location relationship includes: for each sub-region, determining the adjacent sub-regions directly connected to the sub-region based on the location relationship, and determining the average of the independent risk values ​​of the sub-region and its adjacent sub-regions as the first risk value of the sub-region; Based on vegetation cover data, the vegetation cover characteristics of each sub-region are determined; Based on the vegetation cover characteristics and geological information, the vegetation impact characteristics of each sub-region are determined; Based on the vegetation impact characteristics and the first risk value, a second risk value is determined for each sub-region; Based on the second risk value, a dispensing instruction is generated and sent to the emergency monitoring internal perception and control platform to control the drone to dispense flocculant according to the dispensing instruction. The dispensing instruction includes a dispensing path, dispensing points, and dispensing volume. The dispensing path and dispensing points are determined based on the topographic information of each sub-region and the vegetation cover data. Based on the first risk value, downstream population density, and the enhanced multimodal data, a collection instruction is generated and sent to the emergency monitoring internal perception and control platform to control the drone to collect data based on the collection instruction. The collection instruction includes the collection path, collection point, and collection volume.

5. The method as described in claim 4, characterized in that, The determination of the first risk value for each sub-region based on the independent risk value and the location relationship includes: Based on the multiple independent risk values ​​corresponding to the multiple sub-regions and the location relationships, a debris flow risk map is constructed. Based on the debris flow risk map, multiple first risk values ​​corresponding to the multiple sub-regions are determined by a risk assessment model, wherein the risk assessment model is a machine learning model.

6. The method as described in claim 5, characterized in that, The debris flow risk map also includes multiple edge weights corresponding to multiple edges, and the method further includes: Based on geomorphological information, geological information, and surface runoff information, the flow characteristics of multiple debris flows corresponding to the multiple sub-regions are determined. Based on the multiple debris flow flow characteristics and the positional relationships, the multiple edge weights are determined.

7. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions. When the computer reads the computer instructions from the storage medium, the computer executes the method as described in any one of claims 4 to 6.

Images