Debris flow disaster early warning method and emergency response system based on big data analysis

Abstract

The invention discloses a debris flow disaster early warning method and an emergency response system based on big data analysis, and belongs to the technical field of geological monitoring, and the method specifically comprises the steps: collecting and preprocessing watershed terrain, weather, geology, vegetation coverage and historical debris flow event record data, and obtaining image data through combining a remote sensing technology; extracting and integrating the data and image features to form a debris flow feature data set; a debris flow early warning model is constructed by using a finite volume model, a deep learning algorithm and an Internet of Things detection technology in combination with an integrated learning method, and training and verification are performed through historical data; when the model predicts a risk, issuing early warning information and starting a dynamic response mechanism; and finally, the early warning model is optimized and adjusted according to response effect feedback. According to the invention, the accuracy and timeliness of debris flow early warning are improved, and support is provided for disaster prevention and reduction.

Description

Technical Field

[0001] The present invention belongs to the field of geological monitoring technology, and specifically to a debris flow disaster early warning method and emergency response system based on big data analysis. Background Art

[0002] Debris flow is a serious geological disaster that usually occurs in mountainous and hilly areas. It is a flowing body formed by a large amount of solid matter, such as soil and rock fragments, mixed with water. It flows rapidly along slopes or valleys and has great destructive power. Debris flow is often accompanied by natural phenomena such as heavy rain, snowmelt, and earthquakes. Therefore, the development of effective debris flow early warning methods is of great significance for reducing disaster losses. Traditional debris flow early warning methods mainly include traditional empirical methods and remote sensing technology methods. Traditional empirical methods have the problems of strong subjectivity and low accuracy. Remote sensing technology methods use satellite remote sensing images to monitor terrain changes and vegetation coverage to assist in early warning. However, the temporal resolution and spatial resolution of remote sensing images are limited, which makes it difficult to meet the needs of real-time early warning.

[0003] For example, a Chinese patent with the authorization announcement number CN115471980B discloses a debris flow disaster warning method, including: receiving disaster sensing data; obtaining continuous time series video images; extracting multiple continuous image frame groups; identifying foreground points and background points of the target scene; constructing multiple foreground images corresponding to each continuous image frame group; generating a foreground mask; performing morphological processing on the foreground mask to generate a first preprocessed foreground mask; performing connected domain analysis on the first preprocessed foreground mask to generate a second preprocessed foreground mask; obtaining the minimum bounding rectangle of the final screened moving target in the second preprocessed foreground mask and the two-dimensional plane coordinates of the minimum bounding rectangle; generating a continuous sequence motion area detection image; using an image classification algorithm to classify the cropped continuous sequence motion area detection image; when classified into the debris flow subclass, generating a debris flow disaster warning. This technical solution uses digital image processing technology to detect the motion area in the video, which can effectively improve the warning detection accuracy of debris flow disasters.

[0004] The above existing technologies all have the following problems: single data source and high environmental dependence; lack of real-time performance; limited warning range; and lack of dynamic adjustment mechanism. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes a debris flow disaster warning method and emergency response system based on big data analysis, which collects and preprocesses the basin topography, meteorology, geology, vegetation coverage and historical debris flow event record data, and obtains image data in combination with remote sensing technology; extracts and integrates the above data and image features to form a debris flow feature data set; uses finite volume models, deep learning algorithms and Internet of Things detection technology, combined with integrated learning methods to build a debris flow warning model, and trains and verifies it through historical data; when the model predicts a risk, it issues warning information and starts a dynamic response mechanism; finally, the warning model is optimized and adjusted according to the response effect feedback; the present invention improves the accuracy and timeliness of debris flow warnings.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The debris flow disaster early warning method based on big data analysis includes:

[0008] Step S1: Collect basin topographic data, meteorological data, geological data, and vegetation coverage data, and perform preprocessing to generate comprehensive debris flow warning data. At the same time, collect historical debris flow event record data, and use remote sensing technology to obtain remote sensing image data in the basin;

[0009] Step S2: extracting features from debris flow warning comprehensive data and remote sensing image data, and integrating them to form debris flow feature data;

[0010] Step S3: Based on the debris flow characteristic data, the debris flow early warning model is constructed by using an integrated learning method in combination with a finite volume model, a deep learning algorithm and an Internet of Things detection method, and the early warning model is trained and verified using historical debris flow event record data;

[0011] Step S4: When the verified debris flow warning model predicts a debris flow risk, a warning message is issued and a rule-based dynamic response mechanism is established;

[0012] Step S5: Optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism.

[0013] Specifically, the specific steps of step S3 include:

[0014] S3.1: Obtain debris flow characteristic data ,in, represents the nth debris flow characteristic data, n represents the number of debris flow characteristic data, and the debris flow characteristic data includes debris flow early warning comprehensive characteristic data and remote sensing image characteristic data;

[0015] S3.2: Based on , using the finite volume model to simulate and predict the movement of debris flows and calculate the probability of debris flows occurring and strength ;

[0016] S3.3: Introduce deep learning algorithms to classify remote sensing image feature data into debris flow prone areas and non-prone areas , and identify the characteristics of debris flow-prone areas;

[0017] S3.4: Use the Internet of Things detection method to monitor the changes in debris flow characteristic data in real time by deploying sensor networks to obtain new debris flow characteristic data ,in, represents the mth debris flow characteristic data, and m represents the number of debris flow characteristic data;

[0018] S3.5: Use ensemble learning methods to integrate and optimize the finite volume model, deep learning algorithm, and IoT detection method to generate a debris flow warning model;

[0019] S3.6: Use historical debris flow event record data to verify the debris flow warning model, and deploy the verified warning model into the debris flow warning system.

[0020] Specifically, the specific steps of S3.2 include:

[0021] S3.21: Acquisition ,according to Set the initial conditions and boundary conditions of the simulation, combine the physical characteristics and movement laws of the debris flow, and build a finite volume model. The initial conditions include the initial position of the debris flow. , initial velocity v, initial volume V and initial concentration C, the boundary conditions include terrain height, obstacle position, flow velocity and pressure at the boundary, where, , , They respectively represent the three-dimensional coordinate positions of the starting points of the debris flow;

[0022] S3.22: Input into the finite volume model, run the finite volume model, simulate the movement process of debris flow, and monitor the parameters in the simulation process;

[0023] S3.23: Post-process the simulation results to extract the probability and intensity information of debris flow. The formula is:

[0024] ;

[0025] in, represents the intercept term, represents the finite volume model parameters, Indicates factors that affect the occurrence of debris flow, including rainfall, terrain slope, soil type, and vegetation coverage. represents the density of debris flow, represents the volume of debris flow, k represents the influence coefficient of obstacles on the absorption of debris flow energy, It indicates the speed of debris flow without obstacles and slope influence. represents the friction coefficient between the debris flow and the ground, represents the terrain slope, represents the exponential function, represents the cosine function, Represents the sine function.

[0026] Specifically, the specific steps of S3.5 include:

[0027] S3.51: Use the finite volume model, deep learning algorithm and IoT detection method in steps S3.2 to S3.4 as the base model for ensemble learning;

[0028] S3.52: Obtain historical debris flow event record data and , and trained the finite volume model, deep learning algorithm and IoT detection method respectively to obtain their respective preliminary prediction models;

[0029] S3.53: Using the three trained preliminary prediction models, Make predictions and obtain the prediction results of each base model;

[0030] S3.54: The prediction results of each base model are integrated by using the ensemble learning method based on weighted average to obtain a preliminary debris flow warning model;

[0031] S3.55: Evaluate the performance of the preliminary debris flow warning model and optimize the preliminary debris flow warning model based on the evaluation results.

[0032] Specifically, the specific steps of step S4 include:

[0033] S4.1: After receiving the real-time monitored debris flow characteristic data, the debris flow warning model performs risk assessment calculations;

[0034] S4.2: Set a risk threshold. If the risk assessment result exceeds the risk threshold, the early warning mechanism will be triggered;

[0035] S4.3: Determine the warning level and scope based on the risk assessment results and the output of the debris flow warning model, and transmit the warning level and scope information to the emergency command organization under the rule-based dynamic response plan framework;

[0036] S4.4: The emergency command organization shall formulate a dynamic response plan based on the warning level and scope information;

[0037] S4.5: When debris flow risk occurs, activate the dynamic response plan and execute response actions according to the predetermined measures.

[0038] Specifically, the rule-based dynamic response mechanism in step S4 automatically generates a corresponding emergency response strategy according to the predicted debris flow risk level, including formulating an emergency plan, clarifying the warning level and response measures, and tracking the response execution status in real time through the Internet of Things technology.

[0039] The debris flow disaster emergency response system based on big data analysis includes: data processing module, feature extraction module, early warning model construction module, dynamic response module, and model optimization module;

[0040] The data processing module is used to collect debris flow warning comprehensive data and remote sensing image data, and perform preliminary processing and integration;

[0041] The feature extraction module is used to extract features from the debris flow warning comprehensive data and remote sensing image data, and integrate them to form debris flow feature data;

[0042] The early warning model building module is used to build a debris flow early warning model based on debris flow characteristic data, combined with a finite volume model, a deep learning algorithm and an Internet of Things detection method, and use historical debris flow event record data for training and verification;

[0043] The dynamic response module is used to issue warning information and establish a rule-based dynamic response mechanism when the early warning model predicts a debris flow risk;

[0044] The model optimization module is used to optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism.

[0045] Specifically, the early warning model construction module includes: an Internet of Things detection unit, a model construction unit, a model training unit, and a model verification unit;

[0046] The Internet of Things detection unit is used to monitor debris flow characteristic data in real time through the Internet of Things technology;

[0047] The model building unit is used to build a debris flow early warning model using a finite volume model, a deep learning algorithm and an Internet of Things detection method;

[0048] The model training unit is used to train the debris flow warning model using historical debris flow event record data;

[0049] The model verification unit is used to verify the accuracy and reliability of the debris flow warning model.

[0050] Specifically, the dynamic response module includes: an early warning information release unit, a response mechanism construction unit, and a response effect monitoring unit;

[0051] The warning information issuing unit is used to issue the warning information to the working department in a timely manner;

[0052] The response mechanism building unit is used to establish a rule-based dynamic response mechanism and action plan according to the level and urgency of the warning information;

[0053] The response effect monitoring unit is used to monitor and evaluate the execution effect of the dynamic response mechanism in real time.

[0054] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a debris flow disaster early warning method based on big data analysis when executing the computer program.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] 1. The present invention proposes a debris flow disaster emergency response system based on big data analysis, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.

[0057] 2. The present invention proposes a debris flow disaster early warning method based on big data analysis. By collecting and integrating multi-source data, including basin topography, meteorology, geology, vegetation coverage and historical debris flow event records, and combining remote sensing technology to obtain image data, comprehensive monitoring and early warning of debris flow risks are achieved. This comprehensive data processing method significantly improves the accuracy and reliability of the early warning model.

[0058] 3. The present invention proposes a debris flow disaster early warning method based on big data analysis. The process also adopts an advanced integrated learning method to construct a debris flow early warning model, and optimizes it in combination with a finite volume model, a deep learning algorithm and an Internet of Things detection method. When the model predicts a debris flow risk, it can quickly release early warning information and start a rule-based dynamic response mechanism. At the same time, according to the feedback on the execution effect of the response mechanism, the early warning model is continuously optimized and adjusted to form a closed-loop monitoring, early warning and response system, which not only improves the timeliness of debris flow early warnings, but also effectively reduces disaster losses and protects people’s lives and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of a debris flow disaster early warning method based on big data analysis of the present invention;

[0060] Figure 2 This is a principle flow chart of the debris flow disaster early warning method based on big data analysis of the present invention;

[0061] Figure 3 A flow chart of debris flow warning model generation for a debris flow disaster warning method based on big data analysis of the present invention;

[0062] Figure 4 This is an architecture diagram of the debris flow disaster emergency response system based on big data analysis of the present invention. DETAILED DESCRIPTION

[0063] Example 1

[0064] See also Figure 1-Figure 3 The present invention provides an embodiment: a debris flow disaster early warning method based on big data analysis, comprising the following steps:

[0065] Step S1: Collect basin topographic data, meteorological data, geological data, and vegetation coverage data, and perform preprocessing to generate comprehensive debris flow warning data. At the same time, collect historical debris flow event record data, and use remote sensing technology to obtain remote sensing image data in the basin;

[0066] Among them, basin topographic data: high-precision topographic data is obtained through geographic information system, including key parameters such as elevation, slope, and aspect;

[0067] Meteorological data: Obtain real-time meteorological data from meteorological departments or meteorological observation stations, especially information on short-term heavy rainfall, which is one of the key factors triggering debris flows;

[0068] Geological data: Collect geological information such as rock types, rock layer structure, soil properties, etc. in the basin, which helps to understand the geological background of debris flow;

[0069] Vegetation coverage: remote sensing technology is used to obtain vegetation coverage data. Vegetation coverage has a certain impact on the occurrence of debris flow;

[0070] Historical debris flow event records: Organize and analyze historical debris flow event data, including the time, location, scale, influencing factors, etc., to provide training data and verification basis for early warning models;

[0071] Remote sensing image data: Remote sensing image data of the basin is obtained using remote sensing satellites or drones to extract information such as terrain features and vegetation distribution.

[0072] Step S2: extracting features from debris flow warning comprehensive data and remote sensing image data, and integrating them to form debris flow feature data;

[0073] Furthermore, the specific steps of step S2 include:

[0074] (1) Obtain comprehensive debris flow warning data and remote sensing image data;

[0075] (2) Feature extraction of comprehensive data for debris flow warning:

[0076] 1) According to the requirements of the debris flow warning model, the characteristics of rainfall, rainfall intensity, rainfall duration, etc. in the meteorological data are extracted using statistical and computational methods. The statistical and computational methods are the prior art content in this field and are not the inventive solutions of this application, so they will not be described in detail here;

[0077] 2) Extract geological structure, lithology, fault distribution and other characteristics from geological data through geological survey;

[0078] 3) Extracting features such as slope, elevation, and watershed division from watershed terrain data based on geographic information systems;

[0079] (3) Feature extraction of remote sensing image data:

[0080] The principal component analysis method is used to extract the spectral features, texture features, and shape features in the image. Based on the performance characteristics of debris flow in remote sensing images, such as color, brightness, texture changes, etc., image features related to debris flow are extracted.

[0081] (4) Feature integration and data fusion:

[0082] The extracted debris flow warning comprehensive data features and remote sensing image data features are integrated through data splicing to form debris flow characteristic data.

[0083] Step S3: Based on the debris flow characteristic data, the debris flow early warning model is constructed by using an integrated learning method in combination with a finite volume model, a deep learning algorithm and an Internet of Things detection method, and the early warning model is trained and verified using historical debris flow event record data;

[0084] Step S4: When the verified debris flow warning model predicts a debris flow risk, a warning message is issued and a rule-based dynamic response mechanism is established;

[0085] Step S5: Optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism.

[0086] Furthermore, the specific steps of step S5 include:

[0087] (1) Collect relevant data from the execution of the dynamic response mechanism, including response time, effectiveness of response measures, and extent of disaster losses. At the same time, collect performance data of the debris flow early warning model during the prediction process, such as prediction accuracy, missed alarm rate, and false alarm rate;

[0088] (2) Conduct in-depth analysis of the collected feedback data to identify problems and deficiencies in the implementation of the response mechanism and analyze the sources of error in the debris flow warning model during the prediction process, such as data quality, model parameter settings, and algorithm selection;

[0089] (3) Based on the analysis results, determine the direction in which the debris flow warning model needs to be optimized, such as improving the prediction accuracy, reducing missed reports and false alarms, and at the same time, consider whether the rules and strategies of the dynamic response mechanism need to be adjusted to improve the response effect;

[0090] (4) For the early warning model, take measures such as optimizing algorithms, adjusting model parameters, and increasing training data to improve the model's prediction performance. For the dynamic response mechanism, adjust the response rules and strategies to ensure that the response measures are more effective and timely;

[0091] (5) Use the new data set to verify the optimized debris flow warning model and evaluate whether its prediction performance has been improved. Verify the performance of the optimized dynamic response mechanism through simulation or actual disaster events.

[0092] (6) Based on the verification results, the debris flow warning model and dynamic response mechanism are continuously adjusted and optimized to form a closed loop of continuous improvement.

[0093] The specific steps of step S3 include:

[0094] S3.1: Obtain debris flow characteristic data ,in, represents the nth debris flow characteristic data, n represents the number of debris flow characteristic data, and the debris flow characteristic data includes debris flow early warning comprehensive characteristic data and remote sensing image characteristic data;

[0095] S3.2: Based on , using the finite volume model to simulate and predict the movement of debris flows and calculate the probability of debris flows occurring and strength ;

[0096] S3.3: Introduce deep learning algorithms to classify remote sensing image feature data into debris flow prone areas and non-prone areas , and identify the characteristics of debris flow-prone areas;

[0097] Furthermore, the specific steps of S3.3 include:

[0098] (1) Data acquisition and preprocessing: Acquire remote sensing image data and preprocess it to meet model input requirements;

[0099] (2) Model selection and construction: Select the TensorFlow framework according to the task requirements and build a deep learning model. The use of the TensorFlow framework to build a deep learning model is a prior art in this field and is not an inventive solution of the present application, and is not elaborated here.

[0100] (3) Feature learning and extraction: using a deep learning model to automatically learn and extract features from remote sensing images. The training and classification process of the deep learning model is the prior art in this field and is not an inventive solution of the present application, and will not be described in detail here.

[0101] (4) Model training and optimization: Use the training data set to train the deep learning model and optimize it by adjusting the parameters;

[0102] (5) Classification and result analysis: Use the trained deep learning model for classification and analyze the characteristics of debris flow-prone areas.

[0103] S3.4: Use the Internet of Things detection method to monitor the changes in debris flow characteristic data in real time by deploying sensor networks to obtain new debris flow characteristic data ,in, represents the mth debris flow characteristic data, and m represents the number of debris flow characteristic data;

[0104] Furthermore, the specific steps of S3.4 include:

[0105] (1) Construct a monitoring network consisting of multiple sensors in debris flow-prone areas;

[0106] (2) Sensors collect real-time debris flow characteristic data, such as rainfall, mud level, and surface displacement;

[0107] (3) The collected data is transmitted to the data center via the wireless communication network.

[0108] S3.5: Use ensemble learning methods to integrate and optimize the finite volume model, deep learning algorithm, and IoT detection method to generate a debris flow warning model;

[0109] S3.6: Use historical debris flow event record data to verify the debris flow warning model, and deploy the verified warning model into the debris flow warning system.

[0110] The specific steps of S3.2 include:

[0111] S3.21: Acquisition ,according to Set the initial conditions and boundary conditions of the simulation, combine the physical characteristics and movement laws of the debris flow, and build a finite volume model. The initial conditions include the initial position of the debris flow. , initial velocity v, initial volume V and initial concentration C, the boundary conditions include terrain height, obstacle position, flow velocity and pressure at the boundary, where, , , They respectively represent the three-dimensional coordinate positions of the starting points of the debris flow;

[0112] S3.22: Input into the finite volume model, run the finite volume model, simulate the movement process of debris flow, and monitor the parameters in the simulation process;

[0113] Furthermore, the specific steps of constructing the finite volume model and solving it include:

[0114] (1) Determine the simulation area based on the actual location and scope of the debris flow, and divide the simulation area into grids. The size and shape of the grid should be determined based on the flow characteristics of the debris flow and the calculation accuracy;

[0115] (2) Setting initial conditions and boundary conditions. Initial conditions include the initial position, velocity, volume and concentration of the debris flow. Boundary conditions include the undulation of the terrain, the location of obstacles, and the velocity and pressure at the boundary. The setting of these conditions and parameters should be based on actual observation data or expert experience.

[0116] (3) Based on the physical characteristics and movement laws of debris flows, equations describing the movement of debris flows are established. These equations include the mass conservation equation, momentum conservation equation, and energy conservation equation. When constructing the equations, it is necessary to consider the structural two-phase flow characteristics inside the debris flow, that is, the non-settling slurry composed of water and fine-grained sediment is regarded as the "liquid phase", and the coarse sediment or stones larger than a certain critical particle size are regarded as the "solid phase". Among them, the mass conservation equation, momentum conservation equation, and energy conservation equation are the existing technical contents in this field, and are not the inventive solution of this application, and are not elaborated here;

[0117] (4) Converting the established debris flow motion equation into a differential equation for numerical solution. During the solution process, the flow variables in the grid, such as velocity, pressure, and temperature, are continuously updated, and the interaction between adjacent grids is considered. The numerical solution of the equation is a prior art in the art and is not an inventive solution of the present application, and is not described in detail herein.

[0118] (5) Post-process the solution results, such as visualization and data statistical analysis.

[0119] S3.23: Post-process the simulation results to extract the probability and intensity information of debris flow. The formula is:

[0120] ;

[0121] in, represents the intercept term, represents the finite volume model parameters, Indicates factors that affect the occurrence of debris flow, including rainfall, terrain slope, soil type, and vegetation coverage. represents the density of debris flow, represents the volume of debris flow, k represents the influence coefficient of obstacles on the absorption of debris flow energy, It indicates the speed of debris flow without obstacles and slope influence. represents the friction coefficient between the debris flow and the ground, represents the terrain slope, represents the exponential function, represents the cosine function, Represents the sine function.

[0122] The specific steps of S3.5 include:

[0123] S3.51: Use the finite volume model, deep learning algorithm and IoT detection method in steps S3.2 to S3.4 as the base model for ensemble learning;

[0124] S3.52: Obtain historical debris flow event record data and , and respectively train a finite volume model, a deep learning algorithm, and an Internet of Things detection method to obtain respective preliminary prediction models, wherein using known data to train the model is a prior art content in this field, is not an inventive solution of the present application, and is not elaborated here;

[0125] S3.53: Using the three trained preliminary prediction models, Perform prediction to obtain prediction results of each base model, wherein, when the input data is known, the prediction process of the preliminary prediction model is the prior art content in this field, which is not the inventive solution of the present application and will not be described in detail here;

[0126] S3.54: adopting an ensemble learning method based on weighted average to integrate the prediction results of each base model to obtain a preliminary debris flow warning model, wherein the weighted average method is a prior art content in this field and is not an inventive solution of the present application, and is not described in detail here;

[0127] S3.55: Evaluate the performance of the preliminary debris flow warning model and optimize the preliminary debris flow warning model based on the evaluation results.

[0128] Specifically, S3.51, as a preparatory stage, clarified the base model and data source of ensemble learning, providing a basis for subsequent steps; S3.52 used the extracted features to train the base model, providing preliminary prediction results for ensemble learning; S3.53 obtained the respective prediction results through the base model prediction, providing input for ensemble learning; S3.54 and S3.55 obtained the final debris flow warning model through ensemble learning and optimization, and carried out performance evaluation and optimization.

[0129] The specific steps of step S4 include:

[0130] S4.1: After receiving the real-time monitored debris flow characteristic data, the debris flow warning model performs risk assessment calculations;

[0131] Furthermore, the specific steps of S4.1 include:

[0132] (1) Data reception and preprocessing:

[0133] The debris flow warning model first receives debris flow characteristic data from the real-time monitoring system and pre-processes the received data, including data cleaning, format conversion, missing value processing, etc., to ensure the quality and consistency of the data;

[0134] (2) Feature selection:

[0135] Select features relevant to debris flow risk assessment from the preprocessed data, such as rainfall, terrain slope, vegetation coverage, and geological conditions;

[0136] (3) Risk assessment model construction:

[0137] Based on the extracted debris flow characteristic data, the risk assessment model is trained and validated using historical debris flow event data to adjust the model parameters and improve its prediction performance;

[0138] (4) Risk assessment calculation:

[0139] The real-time monitoring debris flow characteristic data is input into the constructed risk assessment model. The risk assessment model calculates the risk level of debris flow occurrence based on the input debris flow characteristic data. ,in, represents the risk assessment function, represents the sth historical debris flow event data, represents the lth risk assessment model parameter;

[0140] (5) Result output and interpretation:

[0141] Output the results of risk assessment calculations, and interpret and analyze the results to provide a scientific basis for subsequent warning information release and emergency response.

[0142] S4.2: Set a risk threshold. If the risk assessment result exceeds the risk threshold, the early warning mechanism will be triggered;

[0143] S4.3: Determine the warning level and scope based on the risk assessment results and the output of the debris flow warning model, and transmit the warning level and scope information to the emergency command organization under the rule-based dynamic response plan framework;

[0144] Furthermore, the specific steps of S4.3 include:

[0145] (1) Interpret the risk assessment results output by the debris flow warning model and understand their meaning and the corresponding debris flow threat level;

[0146] (2) Based on the risk assessment results and the output of the debris flow warning model, determine the debris flow warning level, such as level 1, level 2, or level 3;

[0147] (3) Determine the geographical boundaries of the warning based on the areas and scope that may be affected by the debris flow;

[0148] (4) Organize warning level and scope information into a standardized format to ensure accuracy and readability of the information;

[0149] (5) Through a secure and reliable dedicated communication network transmission channel, the warning level and scope information is transmitted to the emergency command organization under the rule-based dynamic response plan framework to ensure the timeliness and completeness of information transmission so that the emergency command organization can respond quickly.

[0150] S4.4: The emergency command organization shall formulate a dynamic response plan based on the warning level and scope information;

[0151] Among them, after receiving the warning information, the emergency command organization will activate the corresponding emergency plan according to the warning level and scope, including organizing relevant departments and personnel to take necessary emergency measures, such as evacuating residents, blocking roads, and preparing rescue supplies.

[0152] S4.5: When debris flow risk occurs, activate the dynamic response plan and execute response actions according to the predetermined measures.

[0153] In step S4, the rule-based dynamic response mechanism automatically generates the corresponding emergency response strategy according to the predicted debris flow risk level, including formulating emergency plans, clarifying warning levels and response measures, and tracking the response execution in real time through Internet of Things technology.

[0154] Example 2

[0155] See also Figure 4 Another embodiment provided by the present invention is a debris flow disaster emergency response system based on big data analysis, comprising:

[0156] Data processing module, feature extraction module, early warning model building module, dynamic response module, model optimization module;

[0157] Data processing module, used to collect debris flow warning comprehensive data and remote sensing image data, and perform preliminary processing and integration;

[0158] The feature extraction module is used to extract features from the debris flow warning comprehensive data and remote sensing image data, and integrate them to form debris flow feature data to provide input for the construction of the warning model;

[0159] The early warning model building module is used to build a debris flow early warning model based on debris flow characteristic data, combined with finite volume models, deep learning algorithms and Internet of Things detection methods, and use historical debris flow event record data for training and verification;

[0160] Dynamic response module, which is used to issue early warning information in a timely manner and establish a rule-based dynamic response mechanism when the early warning model predicts the risk of debris flow;

[0161] The model optimization module is used to optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism to improve its prediction accuracy and practicality.

[0162] The data processing module includes: terrain data acquisition unit, meteorological data acquisition unit, geological data acquisition unit, vegetation coverage acquisition unit, historical record unit, remote sensing image acquisition unit, and data preprocessing unit;

[0163] The terrain data collection unit is used to collect terrain data within the watershed, such as elevation, slope, slope direction, etc.

[0164] Meteorological data collection unit, used to collect meteorological data such as rainfall, wind speed, temperature, etc.;

[0165] Geological data acquisition unit, used to obtain geological information such as soil type and geological structure;

[0166] Vegetation coverage collection unit, used to measure and record vegetation coverage in the watershed;

[0167] The historical record unit is used to collect and organize the data of debris flow events that occurred in history;

[0168] A remote sensing image acquisition unit is used to acquire remote sensing images within the watershed using remote sensing technology for subsequent feature extraction and analysis;

[0169] The data preprocessing unit is used to perform preprocessing operations such as cleaning, formatting, and normalization on various types of collected data to generate comprehensive debris flow warning data.

[0170] The feature extraction module includes: a feature extraction unit and a feature integration unit;

[0171] The feature extraction unit is used to extract terrain features, meteorological features, geological features and vegetation coverage features from the debris flow warning comprehensive data, and at the same time, to extract terrain features, vegetation distribution and other features from remote sensing images;

[0172] The feature integration unit is used to integrate the features extracted by the feature extraction unit to form a debris flow feature data set.

[0173] The early warning model construction module includes: Internet of Things detection unit, model construction unit, model training unit, and model verification unit;

[0174] The IoT detection unit is used to monitor debris flow characteristic data in real time through IoT technology and provide real-time input for the debris flow early warning model;

[0175] A model building unit, used to build a debris flow warning model using a finite volume model, a deep learning algorithm, and an Internet of Things detection method;

[0176] A model training unit, used to train a debris flow warning model using historical debris flow event record data;

[0177] The model verification unit is used to verify the accuracy and reliability of the debris flow warning model.

[0178] The dynamic response module includes: early warning information release unit, response mechanism construction unit, and response effect monitoring unit;

[0179] Early warning information release unit, used to release early warning information to work departments in a timely manner;

[0180] The response mechanism building unit is used to establish a rule-based dynamic response mechanism and action plan according to the level and urgency of the warning information;

[0181] The response effect monitoring unit is used to monitor and evaluate the execution effect of the dynamic response mechanism in real time.

[0182] The model optimization module includes: performance evaluation unit, model optimization unit, and model update unit;

[0183] Performance evaluation unit, used to evaluate the performance of the early warning model and identify existing problems and deficiencies;

[0184] Model optimization unit, used to optimize and adjust the early warning model according to the evaluation results, such as adjusting model parameters and improving algorithms;

[0185] The model updating unit is used to regularly update the early warning model to reflect the latest changes in geological, meteorological, vegetation and other conditions.

[0186] Example 3

[0187] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a debris flow disaster warning method based on big data analysis are implemented. For details, please refer to the above method embodiment, which will not be repeated here.

[0188] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in the field can also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, which are all within the protection of the present invention.

Claims

1. A debris flow disaster early warning method based on big data analysis, characterized in that: include: Step S1: Collect basin topographic data, meteorological data, geological data, and vegetation coverage data, and perform preprocessing to generate comprehensive debris flow warning data. At the same time, collect historical debris flow event record data, and use remote sensing technology to obtain remote sensing image data in the basin; Step S2: extracting features from debris flow warning comprehensive data and remote sensing image data, and integrating them to form debris flow feature data; Step S3: Based on the debris flow characteristic data, the debris flow early warning model is constructed by using an integrated learning method in combination with a finite volume model, a deep learning algorithm and an Internet of Things detection method, and the early warning model is trained and verified using historical debris flow event record data; Step S4: When the verified debris flow warning model predicts a debris flow risk, a warning message is issued and a rule-based dynamic response mechanism is established; Step S5: Optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism.

2. The debris flow disaster early warning method based on big data analysis according to claim 1, characterized in that: The specific steps of step S3 include: S3.1: Obtain debris flow characteristic data ,in, represents the nth debris flow characteristic data, n represents the number of debris flow characteristic data, and the debris flow characteristic data includes debris flow early warning comprehensive characteristic data and remote sensing image characteristic data; S3.2: Based on , using the finite volume model to simulate and predict the movement of debris flows and calculate the probability of debris flows occurring and strength ; S3.3: Introduce deep learning algorithms to classify remote sensing image feature data into debris flow prone areas and non-prone areas , and identify the characteristics of debris flow-prone areas; S3.4: Use the Internet of Things detection method to monitor the changes in debris flow characteristic data in real time by deploying sensor networks to obtain new debris flow characteristic data ,in, represents the mth debris flow characteristic data, and m represents the number of debris flow characteristic data; S3.5: Use ensemble learning methods to integrate and optimize the finite volume model, deep learning algorithm, and IoT detection method to generate a debris flow warning model; S3.6: Use historical debris flow event record data to verify the debris flow warning model, and deploy the verified warning model into the debris flow warning system.

3. The debris flow disaster early warning method based on big data analysis according to claim 2, characterized in that: The specific steps of S3.2 include: S3.21: Acquisition ,according to Set the initial conditions and boundary conditions of the simulation, combine the physical characteristics and movement laws of the debris flow, and build a finite volume model. The initial conditions include the initial position of the debris flow. , initial velocity v, initial volume V and initial concentration C, the boundary conditions include terrain height, obstacle position, flow velocity and pressure at the boundary, where, , , They respectively represent the three-dimensional coordinate positions of the starting points of the debris flow; S3.22: Input into the finite volume model, run the finite volume model, simulate the movement process of debris flow, and monitor the parameters in the simulation process; S3.23: Post-process the simulation results to extract the probability and intensity information of debris flow. The formula is: ； in, represents the intercept term, represents the finite volume model parameters, Indicates factors that affect the occurrence of debris flow, including rainfall, terrain slope, soil type, and vegetation coverage. represents the density of debris flow, represents the volume of debris flow, k represents the influence coefficient of obstacles on the absorption of debris flow energy, It indicates the speed of debris flow without obstacles and slope influence. represents the friction coefficient between the debris flow and the ground, represents the terrain slope, represents the exponential function, represents the cosine function, Represents the sine function.

4. The debris flow disaster early warning method based on big data analysis according to claim 3, characterized in that: The specific steps of S3.5 include: S3.51: Use the finite volume model, deep learning algorithm and IoT detection method in steps S3.2 to S3.4 as the base model for ensemble learning; S3.52: Obtain historical debris flow event record data and , and trained the finite volume model, deep learning algorithm and IoT detection method respectively to obtain their respective preliminary prediction models; S3.53: Using the three trained preliminary prediction models, Make predictions and obtain the prediction results of each base model; S3.54: The prediction results of each base model are integrated by using the ensemble learning method based on weighted average to obtain a preliminary debris flow warning model; S3.55: Evaluate the performance of the preliminary debris flow warning model and optimize the preliminary debris flow warning model based on the evaluation results.

5. The debris flow disaster early warning method based on big data analysis according to claim 4, characterized in that: The specific steps of step S4 include: S4.1: After receiving the real-time monitored debris flow characteristic data, the debris flow warning model performs risk assessment calculations; S4.2: Set a risk threshold. If the risk assessment result exceeds the risk threshold, the early warning mechanism will be triggered; S4.3: Determine the warning level and scope based on the risk assessment results and the output of the debris flow warning model, and transmit the warning level and scope information to the emergency command organization under the rule-based dynamic response plan framework; S4.4: The emergency command organization shall formulate a dynamic response plan based on the warning level and scope information; S4.5: When debris flow risk occurs, activate the dynamic response plan and execute response actions according to the predetermined measures.

6. The debris flow disaster early warning method based on big data analysis according to claim 5, characterized in that: The rule-based dynamic response mechanism in step S4 automatically generates a corresponding emergency response strategy according to the predicted debris flow risk level, including formulating an emergency plan, clarifying the warning level and response measures, and tracking the response execution status in real time through the Internet of Things technology.

7. A debris flow disaster emergency response system based on big data analysis, which is used to implement the debris flow disaster early warning method based on big data analysis as described in any one of claims 1 to 6, characterized in that: include: Data processing module, feature extraction module, early warning model building module, dynamic response module, model optimization module; The data processing module is used to collect debris flow warning comprehensive data and remote sensing image data, and perform preliminary processing and integration; The feature extraction module is used to extract features from the debris flow warning comprehensive data and remote sensing image data, and integrate them to form debris flow feature data; The early warning model building module is used to build a debris flow early warning model based on debris flow characteristic data, combined with a finite volume model, a deep learning algorithm and an Internet of Things detection method, and use historical debris flow event record data for training and verification; The dynamic response module is used to issue warning information and establish a rule-based dynamic response mechanism when the early warning model predicts a debris flow risk; The model optimization module is used to optimize and adjust the debris flow warning model according to the execution effect feedback of the dynamic response mechanism.

8. The debris flow disaster emergency response system based on big data analysis according to claim 7, characterized in that: The early warning model construction module includes: an Internet of Things detection unit, a model construction unit, a model training unit, and a model verification unit; The Internet of Things detection unit is used to monitor debris flow characteristic data in real time through the Internet of Things technology; The model building unit is used to build a debris flow early warning model using a finite volume model, a deep learning algorithm and an Internet of Things detection method; The model training unit is used to train the debris flow warning model using historical debris flow event record data; The model verification unit is used to verify the accuracy and reliability of the debris flow warning model.

9. The debris flow disaster emergency response system based on big data analysis according to claim 8, characterized in that: The dynamic response module includes: an early warning information release unit, a response mechanism construction unit, and a response effect monitoring unit; The warning information issuing unit is used to issue the warning information to the working department in a timely manner; The response mechanism building unit is used to establish a rule-based dynamic response mechanism and action plan according to the level and urgency of the warning information; The response effect monitoring unit is used to monitor and evaluate the execution effect of the dynamic response mechanism in real time.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the debris flow disaster early warning method based on big data analysis described in any one of claims 1 to 6 are implemented.

Images