Geological disaster risk intelligent management and control method and system based on big data and intelligent algorithm

Abstract

According to the geological disaster risk intelligent management and control method and system based on big data and an intelligent algorithm, safety can be effectively guaranteed, and development is promoted. Multi-source big data of terrains, geology and meteorology are collected and preprocessed, so that the data are ensured to be rich and accurate, and a foundation is laid for subsequent evaluation. A database is constructed by using a GIS, a spatial analysis function of the database is exerted, and landslide susceptibility calculation is supported. A neural network artificial intelligence model is adopted, historical data training is combined, and the risk assessment precision is improved. And the susceptibility interval is divided by using a mean shift clustering algorithm, which is scientific and objective. And a visual susceptibility graph is generated, so that risk information can be quickly mastered. Differentiated control measures are taken for different levels of areas according to the graph, and the method is accurate and efficient. According to the scheme, early warning and disaster prevention can be performed in advance, casualties and property loss are reduced, planning and construction can be reasonably performed, negative effects of disasters on social economy are reduced, and regional sustainable development is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control of geological disaster risks. Specifically, it relates to a method and system for intelligent control of geological disaster risks based on big data and intelligent algorithms. Background Art

[0002] Geological disasters, including landslides, pose a serious threat to the safety of human life and property and infrastructure construction. Accurately assessing and controlling geological disaster risks is crucial for disaster prevention and mitigation. Traditional methods for assessing geological disaster risks often rely on experience and simple statistical analysis, lacking the effective use of a large amount of complex geological data, and the accuracy and reliability of the assessment results are limited.

[0003] With the development of big data and artificial intelligence technologies, new ideas and methods have been provided for geological disaster risk assessment. By using big data, more comprehensive and detailed geological information can be obtained, and intelligent algorithms can deeply mine and analyze these data, so as to more accurately assess the risks of geological disasters. However, the existing methods for assessing geological disaster risks based on big data and intelligent algorithms currently still have problems of being less scientific and reasonable in the division of landslide susceptibility, and cannot well meet the needs of practical applications. Therefore, we have made improvements in this regard and proposed a method and system for intelligent control of geological disaster risks based on big data and intelligent algorithms. Summary of the Invention

[0004] The purpose of the present invention is to address the problems raised in the existing background art. To achieve the above-mentioned invention purpose, the present invention provides the following technical solutions: A method for intelligent control of geological disaster risks based on big data and intelligent algorithms, comprising the following steps: Step 1, data collection and preprocessing: Collect big data related to geological disasters, covering topographic and geomorphic data, geological structure data, meteorological data, and land use data, and perform cleaning, integration, and transformation operations on the collected data to remove noise data and unify the data format;

[0005] Step 2, constructing an artificial intelligence model to calculate landslide susceptibility: Based on the preprocessed data, use geographic information system (GIS) technology to construct a spatial database of geological data; Select an artificial intelligence model algorithm, use the GIS data as input, and use historical landslide data as samples to train the model; Input the GIS data of the area to be evaluated into the trained model, and calculate the landslide susceptibility value of each grid in the area;

[0006] Step 3, using the mean shift clustering algorithm to determine the clustering value: Apply the mean shift clustering algorithm to the susceptibility values of all grids in the GIS to obtain the clustering center point values of the susceptibility values of all grids;

[0007] Step 4. Determine the threshold and divide the susceptibility intervals: Determine the threshold based on the average value of the central points of two adjacent clusters, and divide the susceptibility into different intervals according to the determined threshold;

[0008] Step 5. Generate the susceptibility map: According to the divided susceptibility intervals, assign corresponding susceptibility levels to each grid in GIS, and use the visualization function of GIS to generate the susceptibility map of geological disaster landslides;

[0009] Step 6. Risk control decision-making: According to the generated susceptibility map, take corresponding risk control measures for areas with different susceptibility levels.

[0010] As a preferred technical solution of the present invention, the artificial intelligence model algorithm includes neural network, decision tree, and support vector machine. The mean shift clustering algorithm is a density-based clustering algorithm, which divides sample points into different clusters by continuously iteratively updating the positions of the sample points.

[0011] As a preferred technical solution of the present invention, the division of the susceptibility intervals includes: the susceptibility value range of the entire area is 0-1. When the central point values of the clusters are 0.26, 0.38, 0.6, 0.74, and 0.9, the thresholds are 0.32, 0.49, 0.67, and 0.82. The susceptibility intervals are divided as follows: 0-0.32 is extremely low susceptibility; 0.32-0.49 is low susceptibility; 0.49-0.67 is medium susceptibility; 0.67-0.82 is high susceptibility; 0.82-1 is extremely high susceptibility.

[0012] As a preferred technical solution of the present invention, the risk control measures include:

[0013] For areas with extremely high and high susceptibility, formulate emergency plans;

[0014] For areas with medium susceptibility, conduct regular inspections and monitoring, and take appropriate preventive measures;

[0015] For areas with low and extremely low susceptibility, keep attention.

[0016] An intelligent geological disaster risk control system based on big data and intelligent algorithms, including a data collection and preprocessing module, the data collection and preprocessing module includes a data collection unit and a data preprocessing unit, the data collection and preprocessing module is connected to a landslide susceptibility calculation module, the landslide susceptibility calculation module includes a GIS data management unit, an artificial intelligence model training unit and a landslide susceptibility calculation unit; the landslide susceptibility calculation module is connected to a clustering analysis module, the clustering analysis module includes a mean shift clustering unit, a threshold determination unit, a susceptibility interval division unit, a susceptibility level assignment unit and a graphic display unit; the clustering analysis module is connected to a risk control decision-making module, the risk control decision-making module includes a decision rule formulation unit and a decision execution unit.

[0017] As a preferred technical solution of the present invention, the data collection unit is connected to multiple data sources for collecting big data related to geological disasters, the big data includes topographic and geomorphic data, geological structure data, meteorological data, land use data, and the data sources include geological exploration departments, meteorological monitoring stations, and geographic information mapping agencies; the data preprocessing unit is used to perform cleaning, integration and transformation operations on the collected data, the cleaning is to remove noisy, duplicate and incorrect data, the integration is to integrate data in different formats and sources into a unified data structure, and the transformation includes normalization and standardization operations.

[0018] As a preferred technical solution of the present invention, the GIS data management unit uses geographic information system (GIS) technology to construct a geological data spatial database to store and manage the preprocessed data;

[0019] The artificial intelligence model training unit selects an artificial intelligence model algorithm, uses GIS data as input, and historical landslide data as samples to train the artificial intelligence model algorithm, and optimizes the model performance by adjusting model parameters;

[0020] The landslide susceptibility calculation unit inputs the GIS data of the area to be evaluated into the trained artificial intelligence model to calculate the landslide susceptibility value of each grid in the area.

[0021] As a preferred technical solution of the present invention, the mean shift clustering unit applies the mean shift clustering algorithm to the susceptibility values of all grids in the GIS to obtain the clustering center point values of the susceptibility values of all grids. The mean shift clustering algorithm is a density-based unsupervised clustering algorithm, which iteratively calculates the mean shift vector of the sample points to move the sample points to the area with higher density and converge to the clustering center point;

[0022] The threshold determination unit determines the threshold according to the average value of the adjacent two clustering center point values;

[0023] The susceptibility interval division unit divides the susceptibility into different intervals according to the determined threshold value;

[0024] The susceptibility map generation module, which includes:

[0025] The susceptibility level assignment unit assigns corresponding susceptibility levels to each grid in the GIS according to the divided susceptibility intervals;

[0026] The graphic display unit uses the visualization function of the GIS to display the grids with different susceptibility levels in different colors, symbols or textures, and generates the susceptibility map of geological disaster landslides.

[0027] As a preferred technical solution of the present invention, the decision rule formulation unit formulates risk control decision rules for different susceptibility level regions according to the characteristics of geological disasters and historical experience;

[0028] The decision execution unit automatically generates risk control tasks according to the susceptibility map and decision rules, assigns them to the corresponding departments or personnel for execution, and tracks and feedbacks the task execution situation.

[0029] As a preferred technical solution of the present invention, the system further includes a data storage module for storing various types of data during the operation of the system, including the collected data, the processed data, the model training data, the clustering results, the susceptibility level data, and the risk control task data.

[0030] Compared with the prior art, the beneficial effects of the present invention:

[0031] By collecting multi-source big data of topography, geologic structure, meteorology, and land use, the present invention can reflect the potential influencing factors of geological disasters from multiple dimensions, avoiding the limitations of a single data source, and providing a rich and comprehensive information basis for accurately evaluating geological disaster risks in the follow-up. At the same time, preprocessing operations such as cleaning, integrating, and transforming the collected data effectively remove the noise data and unify the data format, further improving the data quality and ensuring the accuracy of subsequent analysis and modeling.

[0032] The present invention uses geographic information system (GIS) technology to construct a geological data spatial database, realizing the efficient storage and management of data. The powerful spatial analysis function of the GIS can intuitively display the spatial relationships between data, facilitating researchers to conduct in-depth analysis and mining of geological data, and providing strong support for accurately calculating landslide susceptibility.

[0033] The present invention uses artificial intelligence models such as neural networks, decision trees, and support vector machines, and trains with historical landslide data as samples, which can fully learn the laws and characteristics of geological disasters, so as to more accurately calculate the landslide susceptibility values of each grid in the area to be evaluated. Compared with traditional methods, the artificial intelligence model has stronger non-linear fitting ability and adaptability, can handle complex influencing factors of geological disasters, and greatly improves the accuracy of risk assessment.

[0034] The present invention uses the mean shift clustering algorithm to perform clustering analysis on the grid susceptibility values, obtains the clustering center point values, and determines the threshold according to the average value of adjacent center point values to divide the susceptibility intervals. This data-driven division method is more scientific and objective, can fully reflect the actual distribution of geological disaster susceptibility, avoids the interference of subjective factors, and provides a reliable basis for subsequent risk control.

[0035] The present invention uses the visualization function of GIS to generate a geological disaster landslide susceptibility map, presenting complex geological disaster risk information in an intuitive graphical form. This enables decision-makers and relevant personnel to quickly and clearly understand the geological disaster susceptibility levels in different regions, facilitating the formulation of targeted risk control strategies and improving decision-making efficiency and accuracy.

[0036] The present invention takes corresponding risk control measures for areas with different susceptibility levels according to the susceptibility map, realizing precise and differentiated risk control. For areas with extremely high and high susceptibility, emergency plans are formulated, which can respond quickly when disasters occur, reducing casualties and property losses; for areas with medium susceptibility, regular inspections and monitoring are carried out and preventive measures are taken to timely discover and handle potential hidden dangers; for areas with low and extremely low susceptibility, attention is maintained to avoid accidental disasters caused by negligence. This hierarchical management method improves the efficiency of resource utilization and ensures the effectiveness and pertinence of geological disaster risk control work.

[0037] By accurately assessing geological disaster risks and taking effective control measures, the present invention can give early warnings and prevent the occurrence of geological disasters, minimizing casualties and property losses to the greatest extent and ensuring the life and property safety of the people. The scientific and reasonable geological disaster risk control of the present invention helps to reasonably plan land use and infrastructure construction, avoid unreasonable development activities in high-risk areas, thereby reducing the negative impact of geological disasters on social and economic development and promoting the sustainable development of the region. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the experimental data provided by the present invention;

[0039] Figure 2 It is a schematic diagram of the experimental data structure provided by the present invention;

[0040] Figure 3 Schematic diagram of the experimental data structure provided by the present invention;

[0041] Figure 4 Schematic diagram of the experimental data structure provided by the present invention;

[0042] Figure 5 Schematic diagram of the method flow provided by the present invention. Detailed implementation manners

[0043] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Apparently, the described embodiments are some but not all of the embodiments of the present invention.

[0044] Therefore, the following detailed description of the embodiments of the present invention is not intended to limit the scope of the claimed present invention, but merely represents some embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention. It should be noted that, without conflict, the embodiments in the present invention and the features and technical solutions in the embodiments may be combined with each other. It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0045] Embodiment 1: An intelligent geological disaster risk control method based on big data and intelligent algorithms, comprising the following steps: Step 1, data collection and preprocessing: Collect big data related to geological disasters, covering topographic and geomorphic data, geological structure data, meteorological data, and land use data, and perform cleaning, integration, and transformation operations on the collected data to remove noise data and unify the data format;

[0046] Step 2, constructing an artificial intelligence model to calculate landslide susceptibility: Based on the preprocessed data, use geographic information system (GIS) technology to construct a spatial database of geological data; Select an artificial intelligence model algorithm, use the GIS data as input, and use historical landslide data as samples to train the model; Input the GIS data of the area to be evaluated into the trained model to calculate the landslide susceptibility value of each grid in the area;

[0047] Step 3, using the mean shift clustering algorithm to determine the clustering value: Apply the mean shift clustering algorithm to the susceptibility values of all grids in the GIS to obtain the clustering center point values of the susceptibility values of all grids;

[0048] Step 4: Determine the threshold and divide the susceptibility intervals: Determine the threshold based on the average value of the central points of two adjacent clusters, and divide the susceptibility into different intervals according to the determined threshold;

[0049] Step 5: Generate the susceptibility map: According to the divided susceptibility intervals, assign corresponding susceptibility levels to each grid in GIS, and use the visualization function of GIS to generate the susceptibility map of geological disaster landslides;

[0050] Step 6: Risk control decision-making: According to the generated susceptibility map, take corresponding risk control measures for areas with different susceptibility levels.

[0051] Artificial intelligence model algorithms include neural networks, decision trees, and support vector machines. The mean shift clustering algorithm is a density-based clustering algorithm that divides sample points into different clusters by continuously iteratively updating the positions of the sample points.

[0052] The division of susceptibility intervals includes: The susceptibility value range of the entire area is 0 - 1. When the central point values of the clusters are 0.26, 0.38, 0.6, 0.74, and 0.9, the thresholds are 0.32, 0.49, 0.67, and 0.82. The susceptibility intervals are divided as follows: 0 - 0.32 is extremely low susceptibility; 0.32 - 0.49 is low susceptibility; 0.49 - 0.67 is medium susceptibility; 0.67 - 0.82 is high susceptibility; 0.82 - 1 is extremely high susceptibility.

[0053] Risk control measures include:

[0054] For areas with extremely high susceptibility and high susceptibility, formulate emergency plans;

[0055] For areas with medium susceptibility, conduct regular inspections and monitoring, and take appropriate preventive measures;

[0056] For areas with low susceptibility and extremely low susceptibility, keep an eye on them.

[0057] Embodiment 2: A geological disaster risk intelligent control system based on big data and intelligent algorithms, including a data acquisition and preprocessing module. The data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit, and the data acquisition and preprocessing module is connected to the landslide susceptibility calculation module;

[0058] The landslide susceptibility calculation module includes a GIS data management unit, an artificial intelligence model training unit, and a landslide susceptibility calculation unit; the landslide susceptibility calculation module is connected to the clustering analysis module. The clustering analysis module includes a mean shift clustering unit, a threshold determination unit, a susceptibility interval division unit, a susceptibility level assignment unit, and a graphic display unit;

[0059] The clustering analysis module is connected to the risk control decision-making module, which includes a decision rule formulation unit and a decision execution unit.

[0060] The data acquisition unit is connected to multiple data sources and is used to collect big data related to geological disasters. The big data includes topographic and geomorphic data, geological structure data, meteorological data, and land use data. The data sources include geological exploration departments, meteorological monitoring stations, and geospatial information mapping agencies. The data preprocessing unit is used to perform cleaning, integration, and transformation operations on the collected data. Cleaning is to remove noisy, duplicate, and incorrect data. Integration is to integrate data in different formats and from different sources into a unified data structure. Transformation includes normalization and standardization operations.

[0061] The GIS data management unit uses geographic information system (GIS) technology to construct a spatial database of geological data and stores and manages the preprocessed data.

[0062] The artificial intelligence model training unit selects an artificial intelligence model algorithm, uses the GIS data as input, and uses historical landslide data as samples to train the artificial intelligence model algorithm, and optimizes the model performance by adjusting the model parameters. The landslide susceptibility calculation unit inputs the GIS data of the area to be evaluated into the trained artificial intelligence model and calculates the landslide susceptibility value of each grid in the area.

[0063] The mean shift clustering unit applies the mean shift clustering algorithm to the susceptibility values of all grids in the GIS to obtain the clustering center point values of the susceptibility values of all grids. The mean shift clustering algorithm is a density-based unsupervised clustering algorithm that iteratively calculates the mean shift vector of sample points to move the sample points to areas with higher density and converge to the clustering center point.

[0064] The threshold determination unit determines the threshold according to the average value of the adjacent two clustering center point values.

[0065] The susceptibility interval division unit divides the susceptibility into different intervals according to the determined threshold.

[0066] The susceptibility map generation module includes:

[0067] The susceptibility level assignment unit assigns corresponding susceptibility levels to each grid in the GIS according to the divided susceptibility intervals.

[0068] The graphic display unit uses the visualization function of the GIS to display grids with different susceptibility levels in different colors, symbols, or textures to generate a landslide susceptibility map of geological disasters.

[0069] The decision rule formulation unit formulates risk control decision rules for areas with different susceptibility levels according to the characteristics of geological disasters and historical experience.

[0070] The decision execution unit automatically generates risk control tasks according to the susceptibility map and decision rules, assigns them to the corresponding departments or personnel for execution, and tracks and feedbacks on the task execution situation.

[0071] The system further includes a data storage module for storing various types of data during the operation of the system, including the collected data, the processed data, the model training data, the clustering results, the susceptibility level data, and the risk control task data.

[0072] Data collection and preprocessing

[0073] Carry out geological disaster risk assessment work in a certain mountainous area. First, obtain the topographic and geomorphic data of this area from the geological exploration department, including elevation, slope, and aspect; obtain the precipitation data of many years from the meteorological department; obtain the land use data from the surveying and mapping department. Integrate these data and store them in the database. Then, clean the data to remove outliers and missing values; perform data transformation to uniformly convert different formats of data into a format suitable for subsequent analysis.

[0074] Build an artificial intelligence model to calculate landslide susceptibility

[0075] Use the Python programming language and related machine learning libraries, including Scikit-learn, and select the neural network model as the artificial intelligence model. Divide the preprocessed GIS data into a training set and a test set, where the training set accounts for 80% and the test set accounts for 20%. Use the historical landslide data as samples to train the neural network model. During the training process, continuously adjust the parameters of the model, including the learning rate and the number of hidden layer nodes, until the model achieves better performance on the test set. Input the GIS data of this mountainous area into the trained neural network model to calculate the landslide susceptibility value of each grid.

[0076] Use the mean shift clustering algorithm to determine the clustering values

[0077] Use the MeanShift class in the Scikit-learn library of Python to implement the mean shift clustering algorithm. Use the susceptibility values of all grids as the input, run the mean shift clustering algorithm, and obtain the clustering center point values of 0.26, 0.38, 0.6, 0.74, and 0.9.

[0078] Determine the thresholds and divide the susceptibility intervals

[0079] Determine the thresholds according to the average of the values of two adjacent clustering center points, and calculate the thresholds to be 0.32, 0.49, 0.67, and 0.82. Divide the susceptibility into the following intervals:

[0080] 0 - 0.32: Very low susceptibility;

[0081] 0.32 - 0.49: Low susceptibility;

[0082] 0.49 - 0.67: Medium susceptibility;

[0083] 0.67 - 0.82: High susceptibility;

[0084] 0.82 - 1: Extremely high susceptibility;

[0085] Generate susceptibility map: Using ArcGIS software, according to the divided susceptibility intervals, assign corresponding susceptibility levels to each grid. Utilize the symbolization function of ArcGIS to set different colors for areas with different susceptibility levels and generate the landslide susceptibility map of geological disasters.

[0086] Risk control decision

[0087] Based on the generated susceptibility map, take corresponding risk control measures for different regions of the mountainous area. For extremely high susceptibility and high susceptibility regions, set up real - time monitoring equipment, arrange special personnel for regular inspections, and formulate detailed emergency plans; for medium susceptibility regions, install simple monitoring equipment and conduct data analysis regularly; for low susceptibility and extremely low susceptibility regions, conduct a comprehensive inspection once a year.

[0088] As can be seen from the above embodiments, the intelligent geological disaster risk control method based on big data and intelligent algorithms of the present invention can effectively evaluate geological disaster risks and provide a scientific decision - making basis for risk control.

[0089] Based on Python and related geographic information processing libraries, used to implement the above - mentioned intelligent geological disaster risk control method based on big data and intelligent algorithms.

[0090] Experimental environment preparation

[0091] Install Python libraries:

[0092] ```bash

[0093] pip install geopandas rasterio scikit - learn numpy pandas matplotlib

[0094] ```

[0095] Code implementation

[0096] ```

[0097] python

[0098] import geopandas as gpd

[0099] import rasterio

[0100] import numpy as np

[0101] from sklearn.neural_network import MLPClassifier

[0102] from sklearn.cluster import MeanShift

[0103] Step 1: Data collection and preprocessing

[0104] def data_collection_and_preprocessing:

[0105] Simulate data collection

[0106] Replace with actual data loading code according to the actual situation

[0107] terrain_data = np.random.rand(100, 100) Terrain data

[0108] geological_data = np.random.rand(100, 100) Geological structure data

[0109] meteorological_data = np.random.rand(100, 100) Meteorological data

[0110] land_use_data = np.random.rand(100, 100) Land use data

[0111] Data cleaning, integration, transformation

[0112] Remove NaN values and unify data format

[0113] data = np.stack([terrain_data, geological_data, meteorological_data, land_use_data], axis=-1)

[0114] data = np.nan_to_num(data)

[0115] return data

[0116] Step 2: Build an artificial intelligence model to calculate landslide susceptibility

[0117] def calculate_landslide_susceptibility(data):

[0118] Build a spatial database of geological data

[0119] rows, cols, _ = data.shape

[0120] X = data.reshape(-1, 4) Feature data

[0121] Simulate historical landslide data

[0122] y = np.random.randint(0, 2, size=(rows * cols)) 0 indicates no landslide, 1 indicates landslide

[0123] Select an artificial intelligence model and use a neural network

[0124] model = MLPClassifier

[0125] model.fit(X, y)

[0126] Calculate the landslide susceptibility value for each grid

[0127] susceptibility_values = model.predict_proba(X)[:, 1]

[0128] susceptibility_values = susceptibility_values.reshape(rows, cols)

[0129] return susceptibility_values

[0130] Step 3: Determine the clustering values using the Mean Shift clustering algorithm

[0131] def determine_clustering_values(susceptibility_values):

[0132] values = susceptibility_values.flatten().reshape(-1, 1)

[0133] ms = MeanShift

[0134] ms.fit(values)

[0135] cluster_centers = ms.cluster_centers_.flatten

[0136] cluster_centers = np.sort(cluster_centers)

[0137] return cluster_centers

[0138] Step 4: Determine the thresholds and divide the susceptibility intervals

[0139] def determine_thresholds_and_intervals(cluster_centers):

[0140] thresholds = []

[0141] for i in range(len(cluster_centers) - 1):

[0142] threshold = (cluster_centers[i] + cluster_centers[i + 1]) / 2

[0143] thresholds.append(threshold)

[0144] intervals = [0] + thresholds + [1]

[0145] return thresholds, intervals

[0146] Step 5: Generate the susceptibility map

[0147] def generate_susceptibility_map(susceptibility_values, intervals):

[0148] rows, cols = susceptibility_values.shape

[0149] susceptibility_map = np.zeros((rows, cols), dtype = int)

[0150] for i in range(len(intervals) - 1):

[0151] mask = (susceptibility_values >= intervals[i]) & (susceptibility_values < intervals[i + 1])

[0152] susceptibility_map[mask] = i

[0153] return susceptibility_map

[0154] Step 6: Risk management decision

[0155] def risk_management_decision(susceptibility_map):

[0156] risk_levels = []

[0157] for level in np.unique(susceptibility_map):

[0158] if level == 0:

[0159] risk_levels.append("Very low susceptibility: Keep an eye on")

[0160] elif level == 1:

[0161] risk_levels.append("Low susceptibility: Keep an eye on")

[0162] elif level == 2:

[0163] risk_levels.append("Medium susceptibility: Conduct regular inspections and monitoring, and take appropriate preventive measures")

[0164] elif level == 3:

[0165] risk_levels.append("High susceptibility: Develop an emergency plan")

[0166] elif level == 4:

[0167] risk_levels.append("Very high susceptibility: Develop an emergency plan")

[0168] return risk_levels

[0169] Main function

[0170] def main:

[0171] data=data_collection_and_preprocessing

[0172] susceptibility_values=calculate_landslide_susceptibility(data)

[0173] cluster_centers=determine_clustering_values(susceptibility_values)

[0174] thresholds,intervals=determine_thresholds_and_intervals(cluster r_centers)

[0175] susceptibility_map=generate_susceptibility_map(susceptibility_values,intervals)

[0176] risk_levels=risk_management_decision(susceptibility_map)

[0177] print("Cluster center value:",cluster_centers)

[0178] print("Threshold:",thresholds)

[0179] print("Probability interval:",intervals)

[0180] print("Risk control measures:",risk_levels)

[0181] if__name__=="__main__":

[0182] main.

[0183] Experimental example: Experimental purpose

[0184] Verify the effectiveness of the intelligent geological disaster risk management and control method based on big data and intelligent algorithms in actual geological disaster risk assessment and control, and provide a scientific basis for geological disaster prevention and control in relevant areas.

[0185] Experimental area selection

[0186] A mountainous area was selected as the experimental area. The geological conditions in this mountainous area are complex. Landslide geological disasters have occurred many times in history. The topography, geological structure, meteorological conditions and land use are representative.

[0187] Experimental Preparation

[0188] 1. Personnel preparation

[0189] Form a professional experimental team, including geologists, meteorologists, geographic information system (GIS) technicians and data analysts.

[0190] Train team members to familiarize them with experimental procedures and related technical methods.

[0191] 2. Equipment Preparation

[0192] Equipped with high-precision topographic survey equipment, including total stations and GPS receivers, to obtain topographic and geomorphic data.

[0193] Prepare geological exploration equipment, including geological radar and drilling equipment, to collect geological structure data.

[0194] Install meteorological monitoring instruments, including weather stations and rain gauges, to obtain meteorological data.

[0195] Use remote sensing satellite images and drone mapping technology to obtain land use data.

[0196] Prepare high-performance computers and GIS software, including ArcGIS and QGIS, for data processing and analysis.

[0197] Experimental procedures

[0198] Step 1: Data collection and preprocessing

[0199] 1. Data Collection

[0200] Topographic data: Use total stations and GPS receivers to measure the topography of the experimental area and obtain topographic information such as height, slope, and aspect. At the same time, use drones to conduct low-altitude photogrammetry to generate a high-precision digital elevation model (DEM).

[0201] Geological structure data: The geological structure of the experimental area is detected through geological radar and drilling equipment to obtain geological structure data such as stratum lithology, fault distribution, and rock mechanical properties.

[0202] Meteorological data: Install a weather station and rain gauge to monitor the temperature, precipitation, and wind speed meteorological elements in the experimental area in real time. At the same time, collect historical meteorological data, including precipitation records and temperature changes over the years.

[0203] Land use data: Remote sensing satellite images and drone mapping technology are used to obtain the land use types of the experimental area, including cultivated land, forest land, and construction land. Combined with field surveys, the land use data are verified and revised.

[0204] 2. Data Preprocessing

[0205] Data cleaning: Check the collected data and remove duplicate, erroneous and missing data. Examples include checking topographic survey data and removing abnormal elevation values; quality control of meteorological data and removing obviously unreasonable meteorological records.

[0206] Data integration: Integrate data from different sources and formats and store them in a unified database. For example, integrate topographic data, geological structure data, meteorological data, and land use data into a GIS database for subsequent analysis and processing.

[0207] Data transformation: Standardize and normalize the data to unify the data format and scale. For example, normalize the terrain slope data and meteorological precipitation data to make their values ​​range between 0 and 1.

[0208] Step 2: Build an AI model to calculate landslide susceptibility

[0209] 1. Constructing a spatial database of geological data

[0210] Using GIS technology, the pre-processed data is organized and stored according to spatial location to build a geological data spatial database. For example, topographic data, geological structure data, meteorological data and land use data are stored as different layers, and spatial indexes are established for fast query and analysis.

[0211] 2. Choose the AI ​​model algorithm

[0212] The neural network model is selected as the artificial intelligence algorithm for this experiment. Neural networks have strong nonlinear mapping capabilities and can better process complex geological disaster data.

[0213] 3. Model training

[0214] Collect historical landslide data in the experimental area, including the location, time, and scale of the landslide. Use these historical landslide data as samples and match them with relevant data in the geological data spatial database to construct a training data set.

[0215] The neural network model is trained using the training data set, and the parameters of the model are adjusted so that the model can accurately predict landslide susceptibility.

[0216] 4. Calculate the landslide susceptibility value

[0217] The GIS data of the area to be evaluated is input into the trained neural network model to calculate the landslide susceptibility value of each grid in the area. The calculation results are stored in the GIS database as the basis for subsequent analysis.

[0218] Step 3: Determine cluster values ​​using mean shift clustering algorithm

[0219] 1. Data extraction

[0220] The landslide susceptibility values ​​of all rasters were extracted from the GIS database and used as input data for the mean shift clustering algorithm.

[0221] 2. Cluster analysis

[0222] The mean shift clustering tool in the GIS software was used to perform cluster analysis on the susceptibility values ​​of all rasters. Appropriate clustering parameters, including bandwidth, were set, and the sample points were divided into different clusters by continuously iteratively updating the positions of the sample points.

[0223] 3. Determine the cluster center value

[0224] The center point value of each cluster is extracted as the representative value of the cluster. These cluster center point values ​​are recorded for subsequent threshold determination and susceptibility interval division.

[0225] Step 4: Determine thresholds and divide susceptibility intervals

[0226] 1. Calculate the threshold

[0227] The threshold is determined based on the average of the values ​​of two adjacent cluster centers. For example, if the cluster center values ​​are 0.26, 0.38, 0.6, 0.74, and 0.9, the thresholds are (0.26+0.38) / 2=0.32, (0.38+0.6) / 2=0.49, (0.6+0.74) / 2=0.67, and (0.74+0.9) / 2=0.82, respectively.

[0228] 2. Divide the susceptibility interval

[0229] The susceptibility is divided into different intervals based on the determined threshold: 0-0.32 is very low susceptibility; 0.32-0.49 is low susceptibility; 0.49-0.67 is medium susceptibility; 0.67-0.82 is high susceptibility; 0.82-1 is extremely high susceptibility.

[0230] Step 5: Generate Susceptibility Map

[0231] 1. Assignment operation

[0232] According to the divided susceptibility interval, each grid is assigned a corresponding susceptibility level in GIS. For example, grids with susceptibility values ​​between 0 and 0.32 are assigned a very low susceptibility level, represented by blue; grids with susceptibility values ​​between 0.32 and 0.49 are assigned a low susceptibility level, represented by green; and so on.

[0233] 2. Visualization

[0234] By using the visualization function of GIS, the raster data with susceptibility levels are rendered to generate a geological disaster landslide susceptibility map. Legends, scales, and compass elements are added to the map to make the susceptibility map more intuitive and clear.

[0235] Step 6: Risk management decisions

[0236] 1. Regional division

[0237] According to the generated susceptibility map, the experimental area is divided into regions with different susceptibility levels, namely, very low susceptibility region, low susceptibility region, medium susceptibility region, high susceptibility region and very high susceptibility region.

[0238] 2. Develop risk management measures

[0239] Extremely high and high susceptibility areas: Organize relevant experts and emergency management departments to formulate emergency plans, clarify emergency response processes, rescue measures and division of responsibilities. Strengthen monitoring and early warning in these areas, establish a real-time monitoring system, and immediately activate emergency plans once abnormal conditions are found.

[0240] - Medium-prone areas: Arrange professional personnel to conduct regular inspections and monitoring to check the changes in geological disaster risk points. Take appropriate preventive measures, including strengthening the mountains and building drainage facilities, to reduce the risk of geological disasters.

[0241] Low-susceptibility and extremely low-susceptibility areas: Establish a daily attention mechanism to regularly collect and analyze geological and meteorological data in these areas to promptly identify potential geological disaster risks.

[0242] Experimental data

[0243] Step 1: Data collection and preprocessing

[0244]

[0245] Data preprocessing

[0246]

[0247] Step 4: Determine thresholds and divide susceptibility intervals

[0248]

[0249] Interval Division

[0250] The division of the susceptibility intervals includes: 0 - 0.105 is extremely low susceptibility 0.105 - 0.36 is low susceptibility 0.36 - 0.625 is medium susceptibility 0.625 - 0.74 is high susceptibility 0.74 - 1 is extremely high susceptibility

[0251] Experimental results evaluation

[0252] 1. Accuracy Assessment

[0253] Collect new geological disaster data in the experimental area and compare them with the generated geological disaster landslide susceptibility map to evaluate the accuracy of the susceptibility map. Calculate the overlap between the high susceptibility and extremely high susceptibility areas in the susceptibility map and the actual geological disaster areas as an evaluation indicator of accuracy.

[0254] 2. Effectiveness Evaluation

[0255] Observe the effectiveness of risk control measures and evaluate whether the risk control measures taken in different susceptibility levels have effectively reduced the probability of geological disasters and the extent of losses. For example, count the number of geological disasters and losses in high and extremely high susceptibility areas after the implementation of the emergency plan, and compare them with those before implementation.

[0256] Experimental Summary

[0257] According to the experimental results, the advantages and disadvantages of the intelligent management and control methods of geological disaster risks based on big data and intelligent algorithms are summarized, and improvement suggestions are put forward. At the same time, the experimental results are applied to the actual prevention and control of geological disasters to provide technical support for protecting people's lives and property and the stability of the ecological environment.

[0258] The above embodiments are only used to illustrate the present invention and are not intended to limit the technical solutions described in the present invention. Although the present invention has been described in detail with reference to the above embodiments, the present invention is not limited to the above specific implementation methods. Therefore, any modification or replacement of the present invention; and all technical solutions and improvements thereof that do not depart from the spirit and scope of the invention are included in the scope of the claims of the present invention.

Claims

1. A method for intelligent management and control of geological disaster risks based on big data and intelligent algorithms, characterized in that: The following steps are involved: Step 1: Data collection and preprocessing: Collect big data related to geological disasters, including topographic data, geological structure data, meteorological data, and land use data, and clean, integrate, and transform the collected data to remove noise data and unify the data format; Step 2: Construct an artificial intelligence model to calculate landslide susceptibility: Based on the preprocessed data, a spatial database of geological data is constructed using geographic information system (GIS) technology; Select an artificial intelligence model algorithm, use GIS data as input, and use historical landslide data as samples to train the model; input the GIS data of the area to be evaluated into the trained model, and calculate the landslide susceptibility value of each grid in the area; Step 3: Determine cluster values ​​using the mean shift clustering algorithm: Apply the mean shift clustering algorithm to the susceptibility values ​​of all rasters in the GIS to obtain the cluster center point values ​​of all rasters’ susceptibility values; Step 4, determine the threshold and divide the susceptibility interval: determine the threshold according to the average value of the two adjacent cluster center points, and divide the susceptibility into different intervals according to the determined threshold; Step 5: Generate a susceptibility map: According to the divided susceptibility intervals, assign a corresponding susceptibility level to each grid in GIS, and use the visualization function of GIS to generate a geological disaster landslide susceptibility map; Step 6: Risk management decision: Based on the generated susceptibility map, take corresponding risk management measures for areas with different susceptibility levels.

2. According to claim 1, a method for intelligent management and control of geological disaster risks based on big data and intelligent algorithms is characterized in that: The artificial intelligence model algorithm includes a neural network, a decision tree, and a support vector machine. The mean shift clustering algorithm is a density-based clustering algorithm that divides sample points into different clusters by continuously iteratively updating the positions of sample points.

3. According to claim 2, a method for intelligent management and control of geological disaster risks based on big data and intelligent algorithms is characterized in that: The susceptibility interval division includes: the susceptibility value range of the entire area is 0-1. When the cluster center point values ​​are 0.26, 0.38, 0.6, 0.74, and 0.9, the thresholds are 0.32, 0.49, 0.67, and 0.82, and the susceptibility interval is divided into: 0-0.32 is extremely low susceptibility; 0.32-0.49 is low susceptibility; 0.49-0.67 is medium susceptibility; 0.67-0.82 is high susceptibility; 0.82-1 is extremely high susceptibility.

4. The method for intelligent management and control of geological disaster risks based on big data and intelligent algorithms according to claim 3 is characterized in that: The risk control measures include: extremely high susceptibility and high susceptibility areas, medium susceptibility areas, and low susceptibility and extremely low susceptibility areas.

5. An intelligent management and control system for geological disaster risks based on big data and intelligent algorithms, characterized in that: It includes a data acquisition and preprocessing module, which includes a data acquisition unit and a data preprocessing unit. The data acquisition and preprocessing module is connected to the landslide susceptibility calculation module by data. The landslide susceptibility calculation module includes a GIS data management unit, an artificial intelligence model training unit and a landslide susceptibility calculation unit. The landslide susceptibility calculation module is connected to the cluster analysis module by data. The cluster analysis module includes a mean shift clustering unit, a threshold determination unit, a susceptibility interval division unit, a susceptibility level assignment unit and a graphic display unit. The cluster analysis module is data-connected to the risk management and control decision module. The risk management and control decision module includes a decision rule formulation unit and a decision execution unit.

6. According to claim 5, a geological disaster risk intelligent management and control system based on big data and intelligent algorithms is characterized in that: The data acquisition unit is connected to multiple data sources and is used to collect big data related to geological disasters. The big data includes topographic data, geological structure data, meteorological data, and land use data. The data sources include geological exploration departments, meteorological monitoring stations, and geographic information mapping agencies; the data preprocessing unit is used to clean, integrate and transform the collected data. The cleaning is to remove noise, duplication and erroneous data, the integration is to integrate data of different formats and sources into a unified data structure, and the transformation includes normalization and standardization operations.

7. According to claim 6, a geological disaster risk intelligent management and control system based on big data and intelligent algorithms is characterized in that: GIS data management unit, which uses geographic information system (GIS) technology to build a geological data spatial database and store and manage pre-processed data; The artificial intelligence model training unit selects an artificial intelligence model algorithm, takes GIS data as input, and uses historical landslide data as samples to train the artificial intelligence model algorithm, and optimizes model performance by adjusting model parameters; the landslide susceptibility calculation unit inputs the GIS data of the area to be evaluated into the trained artificial intelligence model, and calculates the landslide susceptibility value of each grid in the area.

8. According to claim 7, a geological disaster risk intelligent management and control system based on big data and intelligent algorithms is characterized in that: The mean shift clustering unit applies a mean shift clustering algorithm to the susceptibility values ​​of all grids in the GIS to obtain the cluster center point value of the susceptibility values ​​of all grids. The mean shift clustering algorithm is an unsupervised clustering algorithm based on density. The mean shift vector of the sample point is iteratively calculated to move the sample point to a higher density area and converge to the cluster center point. A threshold determination unit, which determines the threshold according to the average value of two adjacent cluster center points; A susceptibility interval division unit divides the susceptibility into different intervals according to a determined threshold; The susceptibility map generation module includes: a susceptibility level assignment unit, which assigns a corresponding susceptibility level to each grid in the GIS according to the divided susceptibility intervals; The graphic display unit uses the visualization function of GIS to display grids of different susceptibility levels in different colors, symbols or textures to generate a geological hazard landslide susceptibility map.

9. The intelligent management and control system for geological disaster risks based on big data and intelligent algorithms according to claim 8 is characterized in that: The decision-making rule formulation unit formulates risk management and control decision-making rules for areas of different susceptibility levels according to the characteristics of geological disasters and historical experience; The decision execution unit automatically generates risk management and control tasks based on the susceptibility diagram and decision rules, assigns them to corresponding departments or personnel for execution, and tracks and provides feedback on the execution of the tasks.

10. The intelligent management and control system for geological disaster risks based on big data and intelligent algorithms according to claim 9 is characterized in that: The system also includes a data storage module for storing various types of data during the operation of the system, including collected data, processed data, model training data, clustering results, susceptibility level data, and risk management and control task data.

Images