Multi-source data fusion urban inland inundation and secondary disaster dynamic early warning method and system

Abstract

The invention belongs to the technical field of disaster early warning systems, and particularly relates to a dynamic early warning method and system for urban inland inundation and secondary disasters based on multi-source data fusion, and the method comprises a data collection module which is used for collecting the related data of urban inland inundation and secondary disasters from multiple sources in real time, the data includes but is not limited to meteorological data, rainfall data, ground water level data, urban drainage system state data, historical waterlogging event data and geographic information data; the data preprocessing module is used for carrying out cleaning, denoising, missing value filling and normalization processing on the data; the data fusion module is used for fusing data from different data sources and analyzing the data through a multi-source data fusion algorithm to generate a comprehensive urban inland inundation risk assessment model; through multi-source data fusion, accurate prediction and precision of urban inland inundation and secondary disaster occurrence are improved, so that the system can identify disaster risks earlier and more accurately.

Description

Technical Field

[0001] The present invention belongs to the technical field of disaster warning systems, and particularly relates to a dynamic warning method and system for urban waterlogging and secondary disasters with multi-source data fusion. Background Art

[0002] With the acceleration of the urbanization process, urban waterlogging and secondary disasters have become one of the important challenges faced by major cities around the world. Due to climate change, insufficient urban drainage system design, the increase in extreme weather events, and land cover changes brought about by the urbanization process, urban waterlogging occurs frequently, causing serious economic losses, traffic interruptions, and public safety problems. In particular, secondary disasters such as traffic accidents, power outages, and water pollution often exacerbate the impact of disasters.

[0003] To better address these challenges, it is particularly important to develop an efficient dynamic warning system for urban waterlogging and secondary disasters. These systems usually predict the occurrence of disasters and issue warnings by collecting, processing, and analyzing data from various sources in real time. However, existing urban waterlogging and disaster warning systems still have several deficiencies and challenges in data acquisition, fusion, prediction, and emergency response, as follows:

[0004] 1. Most existing urban waterlogging and disaster warning systems mainly rely on a single data source, making it difficult to effectively integrate multiple data sources. In particular, urban waterlogging warning requires comprehensive consideration of multiple factors such as meteorology, geography, traffic, drainage systems, and social feedback. However, these data are often stored in different systems or platforms separately, resulting in ineffective data sharing and the formation of information islands. Therefore, the system is limited in terms of prediction accuracy and timeliness.

[0005] Insufficient data fusion and analysis capabilities. Although machine learning and artificial intelligence technologies have been widely applied in recent years, current dynamic warning systems for urban waterlogging and secondary disasters still face challenges in multi-source data fusion. Existing fusion algorithms usually focus on simple weighted averaging or rule-based methods, failing to fully consider the complex interrelationships and dependencies between data. In addition, the interaction between historical data and real-time data is not fully utilized during the fusion process, which also results in insufficient prediction capabilities of the model. Especially in complex disaster scenarios, it is often difficult to accurately assess the probability of disaster occurrence and the scope of impact.

[0006] 2. The accuracy and robustness of disaster prediction models are insufficient. Although methods such as machine learning and neural networks have been introduced into existing disaster prediction systems for disaster prediction, existing prediction models are usually limited to the processing of a single data source and lack comprehensive evaluation after fusing multi-source data. Therefore, the performance of these models is often unstable in extreme climates or sudden disasters, with low prediction accuracy and robustness, affecting the reliability and timeliness of warnings.

[0007] Therefore, in view of the above problems, a dynamic early warning method and system for urban waterlogging and secondary disasters based on multi-source data fusion are proposed. Summary of the Invention

[0008] (1) Technical Problems to be Solved

[0009] In view of the deficiencies of the prior art, the present invention provides a dynamic early warning method and system for urban waterlogging and secondary disasters based on multi-source data fusion, which is used to solve the problems raised in the background art.

[0010] (2) Technical Solutions

[0011] To achieve the above object, the present invention provides the following technical solutions: A dynamic early warning method and system for urban waterlogging and secondary disasters based on multi-source data fusion include:

[0012] A data acquisition module, which is used to collect data related to urban waterlogging and secondary disasters from multiple sources in real time. The data includes but is not limited to meteorological data, rainfall data, ground water level data, urban drainage system status data, historical waterlogging event data, and geographical information data;

[0013] A data preprocessing module, which is used to clean, denoise, fill in missing values, and normalize the data;

[0014] A data fusion module, which is used to fuse data from different data sources and analyze the data through a multi-source data fusion algorithm to generate a comprehensive urban waterlogging risk assessment model;

[0015] A disaster prediction module, which is used to predict the probability and severity of urban waterlogging and its resulting secondary disasters in real time according to the risk assessment model;

[0016] An early warning release module, which is used to generate early warning information according to the disaster prediction module and release the early warning information to relevant departments and the public through various methods;

[0017] An emergency response module, which is used to provide response measures according to the early warning information level, including evacuation route planning, traffic control, resource scheduling, and post-disaster recovery.

[0018] Preferably, the data preprocessing module includes the following functions:

[0019] Data cleaning. Since multi-source data contains duplicate records, especially when data is collected simultaneously by multiple monitoring devices or data sources, the data preprocessing module should be able to detect and delete duplicate data;

[0020] Data denoising. By removing noise in the data, the quality of the data is improved to ensure the accuracy of subsequent analysis, modeling, or decision-making;

[0021] Missing value filling. In the dataset, there will be missing values due to data collection failures, transmission problems, or equipment malfunctions. The missing values are filled using the mean, median, or mode.

[0022] Normalization processing. The data is transformed to a unified scale or range to eliminate the dimensional differences between different features.

[0023] Preferably, the levels of multi-source data fusion include:

[0024] Data-level fusion. The raw data is directly integrated.

[0025] Feature-level fusion. The features of each data source are extracted using a convolutional neural network; then the features of each data source are fused.

[0026] Decision-level fusion. The independent decision results of each data source are fused.

[0027] Multi-source data fusion algorithms include:

[0028] Weighted average method. Weights are assigned according to the reliability of each data source, and the data is weighted and summed.

[0029] Kalman filter. Based on the state space model, the fusion result is optimized through recursive estimation.

[0030] D-S evidence theory. Uncertainties are processed through basic probability assignment and evidence combination rules.

[0031] Deep learning method. Neural networks are used to automatically learn the complex relationships between data.

[0032] Bayesian network. The dependence relationships between variables are represented through conditional probability distributions for probability inference.

[0033] The data fusion module adopts machine learning algorithms. Weights are dynamically assigned through the entropy weight method and the analytic hierarchy process to ensure that high-value data sources obtain higher weights and automatically adjust the weights of different data sources to improve the accuracy and flexibility of disaster prediction.

[0034] Preferably, the disaster prediction module adopts a neural network or a support vector machine prediction model to predict the occurrence probabilities of urban waterlogging and secondary disasters based on historical data and real-time data.

[0035] Preferably, the early warning release module automatically selects a suitable early warning level through an intelligent decision-making system and adjusts the early warning content and release method according to the disaster type.

[0036] An intelligent decision-making system is a computing system that integrates advanced technologies such as artificial intelligence, big data, machine learning, and knowledge engineering to provide scientific and intelligent decision-making support for decision-makers;

[0037] System architecture of the intelligent decision-making system:

[0038] Data layer:

[0039] Data collection and integration: Integrate multi-source heterogeneous data and achieve data cleaning and standardization through ETL technology;

[0040] Data storage and management: Use a distributed database or real-time stream processing framework to store structured and unstructured data;

[0041] Model layer:

[0042] Machine learning models: Build prediction and optimization models based on supervised learning, unsupervised learning, or reinforcement learning algorithms;

[0043] Knowledge graph: Integrate domain knowledge through entity-relationship modeling to support semantic reasoning and association analysis;

[0044] Simulation and optimization engine: Use Monte Carlo simulation and linear programming algorithms for multi-objective optimization and risk assessment;

[0045] Decision-making layer:

[0046] Decision rule engine: Achieve automated decision-making based on rule-based reasoning or case-based reasoning;

[0047] Multi-criteria decision analysis: Support the analytic hierarchy process to assist in dealing with complex decision-making problems.

[0048] Preferably, the emergency response module includes a real-time feedback mechanism, which automatically adjusts the emergency response strategy and resource scheduling plan according to the real-time data feedback after the disaster occurs.

[0049] Preferably, the disaster prediction module further includes a risk assessment and simulation module, which is used to simulate the consequences under different disaster scenarios to help decision-makers formulate emergency measures.

[0050] Preferably, the system further includes a disaster assessment and post-disaster recovery module, which is used to evaluate the disaster impact and propose a recovery plan according to the loss data after urban waterlogging and secondary disasters occur.

[0051] Preferably, the data collection module further includes a sensor network, which is used to monitor the ground water level, rainfall intensity, and other key indicators in real time.

[0052] A dynamic early warning method for urban waterlogging and secondary disasters with multi-source data fusion, the steps include:

[0053] S1: Obtain information such as rainfall, temperature, and humidity through weather stations, satellites, and radars, obtain real-time data on river, lake, and groundwater levels, obtain the working status of the urban drainage system through sensors and monitoring devices, obtain traffic flow, waterlogging locations, and road conditions through the urban traffic management platform, obtain photos, videos, or text information on urban waterlogging uploaded by citizens from social platforms, and obtain urban geographical information through satellites, drones, or ground surveys;

[0054] S2: Clean, standardize, and format the data to ensure the effective integration of data from different sources on the same platform;

[0055] S3: Generate a global and reliable urban waterlogging and secondary disaster risk model through different data fusion methods;

[0056] S4: Based on the fused data, conduct risk assessment of urban waterlogging and secondary disasters, and predict the possibility and scope of disasters;

[0057] S5: Generate disaster warning information based on the risk assessment results and release it to relevant departments and the public;

[0058] S6: Coordinate emergency responses after disasters occur, and conduct subsequent data analysis and optimization according to actual situations;

[0059] S7: Continuously optimize the data fusion and prediction models through post-disaster analysis and feedback of real-time data, and improve the accuracy and robustness of the system.

[0060] Beneficial Effects

[0061] Compared with the prior art, the present invention provides a dynamic early warning method and system for urban waterlogging and secondary disasters based on multi-source data fusion, having the following beneficial effects:

[0062] 1. In this invention, through multi-source data fusion, the system can integrate information from different data sources, eliminate errors and biases existing in a single data source, thereby improving the accurate prediction and accuracy of urban waterlogging and secondary disasters, enabling the system to identify disaster risks earlier and more accurately.

[0063] 2. In this invention, by collecting multi-source data in real time, it can immediately reflect the urban waterlogging situation. Based on the rapid processing and analysis of real-time data, it can provide dynamic early warnings before disasters occur, enabling the government and the public to obtain early warning information in a timely manner, thereby effectively taking emergency measures and reducing disaster losses.

[0064] 3. The invention integrates multiple data sources such as meteorological data, geographical information, and the status of the drainage system, providing an integrated early warning system that not only covers urban waterlogging itself but also involves secondary disasters, forming a comprehensive disaster prediction and early warning platform. This comprehensiveness greatly improves the efficiency and effectiveness of disaster management. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings described herein are used to provide a further understanding of the present invention, form a part of this application, and the schematic embodiments and descriptions thereof are used to explain the present invention without unduly limiting the present invention. In the drawings:

[0066] Figure 1 is the system flow chart of the present invention;

[0067] Figure 2 is the data fusion flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0069] Specific embodiments are given below.

[0070] Embodiment

[0071] As Figure 1 and Figure 2 shown, the data acquisition module is the basis of the system and is responsible for collecting real-time data related to urban waterlogging and secondary disasters from multiple sources; the specific implementation is as follows:

[0072] Sensor network deployment: Deploy various types of sensors in key areas of the city, including but not limited to meteorological sensors, hydrological sensors, urban drainage system status sensors, and ground water level sensors;

[0073] Meteorological sensors are used to collect meteorological data such as temperature, humidity, wind speed, and precipitation;

[0074] Hydrological sensors are used to monitor real-time data of rivers, lakes, and underground water levels;

[0075] Urban drainage system status sensors monitor the water flow speed and water level of drainage pipes;

[0076] Ground water level sensors monitor the water level changes in low-lying areas or easy-to-flood points in real time;

[0077] Diverse data sources: In addition to sensor networks, a wide range of meteorological data is obtained through means such as weather stations, satellite remote sensing, and radar detection; traffic flow, waterlogging locations, road conditions, etc. are obtained using the urban traffic management platform; at the same time, information such as photos, videos, or text of urban waterlogging uploaded by citizens is collected from social platforms as an important source of social feedback data;

[0078] Data transmission and storage: The collected data is transmitted to the system center using a high-speed and stable data transmission protocol, and a distributed database or real-time stream processing framework is used for data storage and management to ensure the real-time nature and accessibility of the data.

[0079] The data preprocessing module cleans, denoises, fills in missing values, and normalizes the collected raw data to improve data quality and provide a reliable data basis for subsequent data fusion and disaster prediction; the specific implementation methods are as follows:

[0080] Data cleaning: A rule-based cleaning algorithm is used to automatically identify and delete duplicate records, outliers, and obviously incorrect data points;

[0081] Data denoising: Mean filtering technology is used to remove the noise components in the data and improve the accuracy and reliability of the data;

[0082] In the process of data denoising, mean filtering is a simple and effective method. Suppose there is a set of time series data x1, x2,... x n , the data after mean filtering can be calculated by the following formula:

[0083]

[0084] where m is the half-width of the filtering window;

[0085] Missing value filling: For the missing values in the dataset, according to the distribution characteristics and correlations of the data, mean interpolation, median interpolation, or a prediction model based on machine learning is used for filling;

[0086] For missing values, the mean of the previous and subsequent data can be used for filling. Suppose the i-th data point is missing, and its filling value is Its formula is:

[0087]

[0088] Normalization processing: The data is converted to a unified scale or range to eliminate the dimensional differences between different features and improve the efficiency and accuracy of subsequent data fusion and model training;

[0089] The data fusion module is the core part of the system, responsible for fusing data from different data sources and generating a comprehensive urban waterlogging risk assessment model through multi-source data fusion algorithms. The specific implementation methods are as follows:

[0090] Data layer fusion: After data preprocessing, directly perform preliminary integration on the original data to form a multi-dimensional data set;

[0091] Feature layer fusion: Use a convolutional neural network deep learning model to extract the feature information of each data source and fuse these features to form a more representative feature vector;

[0092] Decision layer fusion: Based on the independent decision-making results of each data source, use methods such as weighted average method, Bayesian network or D-S evidence theory for fusion to generate the final urban waterlogging risk assessment model;

[0093] Dynamic weight allocation: Use machine learning algorithms such as entropy weight method and analytic hierarchy process to dynamically allocate weights according to the quality and reliability of the data, ensuring that high-value data sources obtain higher weights in the fusion process, and improving the accuracy and flexibility of disaster prediction;

[0094] The disaster prediction module predicts the probability and severity of urban waterlogging and its resulting secondary disasters in real time according to the risk assessment model. The specific implementation methods are as follows:

[0095] Prediction model selection: Use neural network or support vector machine machine learning models as prediction models, and train and optimize according to historical data and real-time data;

[0096] For a simple fully connected neural network layer, assuming the input is x, the weight is w, the bias is b, and the activation function is σ, then the output y can be expressed as:

[0097] y = σ(W X + b);

[0098] In practical applications, the neural network may contain multiple hidden layers, and the selection of activation functions and loss functions will also affect the specific form of the formula;

[0099] In a linear regression model, assuming the input feature is x, the weight is w, and the bias is b, then the predicted value can be expressed as:

[0100]

[0101] For more complex neural network models, the prediction formula will involve iterative calculations of multiple layers of networks and the application of activation functions;

[0102] Scenario simulation and risk assessment: Use hydrodynamic models or other physical models to simulate the consequences under different disaster scenarios, and combine with risk assessment models to quantitatively evaluate the possibility and impact scope of disasters;

[0103] Real-time update and early warning: Continuously update the risk assessment model according to real-time data, and generate early warning information when the disaster risk reaches the early warning threshold;

[0104] Suppose the risk assessment result is R, and the early warning level L can be judged according to the preset threshold. If R > R high , then L = high-level early warning; if R low <R ≤ R high , then L = medium-level early warning; if R ≤ R low , then L = low-level early warning;

[0105] Here, R high and R low are the preset early warning thresholds.

[0106] The early warning release module releases the early warning information to relevant departments and the public through various methods according to the early warning information generated by the disaster prediction module. The specific implementation methods are as follows:

[0107] Intelligent decision-making system: Integrate artificial intelligence, big data, and machine learning technologies, and automatically select suitable release methods and content according to the disaster type and early warning level;

[0108] Multi-channel release: Release the early warning information through multiple channels such as text messages, phone calls, mobile application push, television and radio, and electronic display screens to ensure the wide spread and timely reception of the information;

[0109] The emergency response module provides response measures according to the early warning information level, including evacuation route planning, traffic control, resource scheduling, and post-disaster recovery. The specific implementation methods are as follows:

[0110] Formulation of emergency response plans: Formulate corresponding emergency response plans for different levels of early warning information, and clarify the responsibilities and tasks of each department;

[0111] Real-time feedback and adjustment: After the disaster occurs, automatically adjust the emergency response strategy and resource scheduling plan according to real-time data feedback and social feedback to ensure the pertinence and effectiveness of the emergency response;

[0112] Post-disaster assessment and recovery: After the disaster occurs, conduct loss assessment on the affected area, and put forward recovery plans and resource scheduling suggestions according to the assessment results;

[0113] System continuous optimization and iteration is the key to improving the accuracy and timeliness of early warnings. The specific implementation is as follows:

[0114] Effect evaluation: After each disaster, evaluate the effectiveness of early warning and emergency response, identify deficiencies and make improvements;

[0115] Algorithm optimization: Continuously optimize the data fusion algorithm and disaster prediction model based on the data feedback and prediction accuracy during the occurrence of disasters, and improve the accuracy and timeliness of the next early warning;

[0116] Continuous learning: Continuously train and optimize the model and algorithm by collecting new disaster data and user feedback to make the system more intelligent and precise.

[0117] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. A multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system, characterized in that: The system comprises: A data collection module is used to collect urban waterlogging and secondary disaster related data from multiple sources in real time, including but not limited to meteorological data, rainfall data, ground water level data, urban drainage system status data, historical waterlogging event data, and geographic information data; A data preprocessing module, used for cleaning, denoising, filling missing values ​​and normalizing the data; Data fusion module, used to fuse data from different data sources and analyze the data through multi-source data fusion algorithm to generate a comprehensive urban waterlogging risk assessment model; A disaster prediction module, which predicts the probability and severity of urban waterlogging and the secondary disasters caused by it in real time according to the risk assessment model; An early warning release module generates early warning information according to the disaster prediction module and releases the early warning information to relevant departments and the public in various ways; The emergency response module is used to provide response measures according to the warning information level, including evacuation route planning, traffic control, resource scheduling, and post-disaster recovery.

2. According to the multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system of claim 1, it is characterized by: The data preprocessing module includes the following functions: Data cleaning, multi-source data contains duplicate records, especially when multiple monitoring devices or data sources collect data at the same time, the data preprocessing module should be able to detect and delete duplicate data; Data denoising: improving data quality by removing noise from the data to ensure the accuracy of subsequent analysis, modeling or decision-making; Missing value filling: In a data set, there may be missing values ​​due to data collection failure, transmission problems, or equipment failure. Use the mean, median, or mode to fill in missing values. Normalization is the process of converting data to a uniform scale or range to eliminate dimensional differences between different features.

3. According to the multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system of claim 1, it is characterized by: The levels of multi-source data fusion include: Data layer fusion, directly integrating the original data; Feature layer fusion, using convolutional neural network to extract the features of each data source; then fuse the features of each data source; Decision-making layer fusion, the independent decision results of each data source are integrated; Multi-source data fusion algorithms include: The weighted average method assigns weights according to the reliability of each data source and performs a weighted summation of the data; Kalman filtering, based on the state space model, optimizes the fusion result through recursive estimation; DS evidence theory, which handles uncertainty through basic probability assignment and evidence combination rules; Deep learning methods use neural networks to automatically learn complex relationships between data; Bayesian network, which represents the dependency between variables through conditional probability distribution and performs probabilistic reasoning; The data fusion module adopts machine learning algorithms to dynamically allocate weights through the entropy weight method and hierarchical analysis method to ensure that high-value data sources obtain higher weights and automatically adjust the weights of different data sources to improve the accuracy and flexibility of disaster prediction.

4. According to claim 1, a multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system is characterized in that: The disaster prediction module adopts a neural network or support vector machine prediction model to predict the probability of urban waterlogging and secondary disasters based on historical data and real-time data.

5. According to the multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system of claim 1, it is characterized in that: The warning release module automatically selects the appropriate warning level through the intelligent decision-making system, and adjusts the warning content and release method according to the disaster type; Intelligent decision-making system is a computing system that integrates artificial intelligence, big data, machine learning, knowledge engineering and other advanced technologies to provide scientific and intelligent decision-making support for decision makers; System architecture of intelligent decision-making system: Data Layer: Data collection and integration: Integrate multi-source heterogeneous data and use ETL technology to clean and standardize data; Data storage and management: Use distributed databases or real-time stream processing frameworks to store structured and unstructured data; Model layer: Machine learning models: Build prediction and optimization models based on supervised learning, unsupervised learning, or reinforcement learning algorithms; Knowledge graph: Integrate domain knowledge through entity-relationship modeling to support semantic reasoning and association analysis; Simulation and optimization engine: Monte Carlo simulation and linear programming algorithms are used for multi-objective optimization and risk assessment; Decision-making level: Decision rule engine: realizes automated decision-making based on rule reasoning or case reasoning; Multi-criteria decision analysis: supports hierarchical analysis method to assist in dealing with complex decision-making problems.

6. The multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system according to claim 1 is characterized in that: The emergency response module includes a real-time feedback mechanism, which automatically adjusts the emergency response strategy and resource scheduling plan according to the real-time data feedback after the disaster occurs.

7. The multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system according to claim 1 is characterized in that: The disaster prediction module also includes a risk assessment and simulation module for simulating the consequences of different disaster scenarios to help decision makers formulate emergency measures.

8. The multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system according to claim 1 is characterized in that: The system also includes a disaster assessment and post-disaster recovery module, which is used to assess the impact of disasters and propose recovery plans based on the loss data after urban waterlogging and secondary disasters occur.

9. The multi-source data fusion urban waterlogging and secondary disaster dynamic early warning system according to claim 1 is characterized in that: The data acquisition module also includes a sensor network for real-time monitoring of ground water level, rainfall intensity and other key indicators.

10. A multi-source data fusion urban waterlogging and secondary disaster dynamic early warning method, characterized in that the steps include: S1: Obtain rainfall, temperature, and humidity information through weather stations, satellites, and radars; obtain real-time data on rivers, lakes, and groundwater levels; obtain the working status of urban drainage systems through sensors and monitoring equipment; obtain traffic flow, waterlogging locations, and road conditions through urban traffic management platforms; obtain photos, videos, or text information about urban flooding uploaded by citizens from social platforms; and obtain urban geographic information through satellites, drones, or ground measurements; S2: Clean, standardize and format data to ensure that data from different sources can be effectively integrated on the same platform; S3: Generate a global and reliable urban waterlogging risk assessment model through different data fusion methods; S4: Based on the integrated data, conduct risk assessment of urban waterlogging and secondary disasters, and predict the possibility and impact range of disasters; S5: Generate disaster warning information based on risk assessment results and release it to relevant departments and the public; S6: After a disaster occurs, coordinate emergency response and conduct subsequent data analysis and optimization based on actual conditions; S7: Through post-disaster analysis and real-time data feedback, continuously optimize data fusion and prediction models to improve the accuracy and robustness of the system.

Images