Disaster risk early warning method and system combined with multi-source data analysis and electronic equipment

Abstract

The invention discloses a disaster risk early warning method and system combined with multi-source data analysis and electronic equipment, and relates to the technical field of data analysis. The method comprises the following steps: carrying out monitoring range screening according to real-time latitude and longitude coordinates to obtain K disaster risk monitoring subareas; executing multi-source information cross-platform crawling to obtain K pieces of disaster monitoring multi-source data; retrieving and outputting the geographic feature data of the K subareas from the open GIS data platform; acquiring disaster risk characteristics of K subareas; carrying out risk dispersion fitting on the disaster risk characteristics of the K subareas, and outputting disaster risk evolution characteristics; and performing spatial consistency analysis according to the disaster risk evolution characteristics and the position deviation of the real-time latitude and longitude coordinates, outputting real-time disaster risk early warning, and transmitting the real-time disaster risk early warning to a risk monitoring user through a mobile terminal. The technical problems of lagged response and insufficient accuracy of disaster risk early warning in the prior art are solved, and the technical effect of improving the timeliness and accuracy of disaster early warning is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to a disaster risk early warning method, system and electronic device that combines multi-source data analysis. Background Art

[0002] Most of the existing disaster risk early warning systems rely on a single data source for monitoring and evaluation, lacking comprehensive analysis of multi-dimensional data. This often leads to a problem of lagging response in disaster risk early warning, and it is unable to provide effective early warning information to relevant personnel in a timely and accurate manner. In addition, traditional systems are usually based on static geographical data and historical disaster data, lacking the ability to respond to real-time dynamic changes, and it is difficult to effectively predict the occurrence and development of disasters in a complex and changeable environment. Summary of the Invention

[0003] This application provides a disaster risk early warning method, system and electronic device that combines multi-source data analysis, and solves the technical problems of lagging response and insufficient accuracy in disaster risk early warning in the prior art.

[0004] In the first aspect of this application, a disaster risk early warning method that combines multi-source data analysis is provided. The method includes:

[0005] Screening the monitoring range according to the real-time latitude and longitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas; using the K sub-area longitude and latitude spans of the K disaster risk monitoring sub-areas as geographical boundaries, performing cross-platform crawling of multi-source information to obtain K multi-source data for disaster monitoring; retrieving and outputting K sub-area geographical feature data from the open GIS data platform according to the K sub-area longitude and latitude spans; fusing and analyzing the K multi-source data for disaster monitoring and the K sub-area geographical feature data to obtain K sub-area disaster risk characteristics; performing risk diffusion fitting on the K sub-area disaster risk characteristics according to the connection relationship of the K disaster risk monitoring sub-areas, and outputting disaster risk evolution characteristics; performing spatial consistency analysis according to the position deviation between the disaster risk evolution characteristics and the real-time latitude and longitude coordinates, outputting real-time disaster risk early warning, and transmitting the real-time disaster risk early warning to the risk monitoring user through a mobile terminal.

[0006] In the second aspect of this application, a disaster risk early warning system that combines multi-source data analysis is provided. The system includes:

[0007] A screening module for screening the monitoring range according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas; an information crawling module for performing cross-platform crawling of multi-source information by using the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries to obtain K multi-source data for disaster monitoring; a retrieval module for retrieving and outputting geographical feature data of the K sub-areas from an open GIS data platform according to the longitude and latitude spans of the K sub-areas; a fusion analysis module for fusing and analyzing the K multi-source data for disaster monitoring and the geographical feature data of the K sub-areas to obtain disaster risk characteristics of the K sub-areas; a fitting module for performing risk diffusion fitting on the disaster risk characteristics of the K sub-areas according to the connection relationship of the K disaster risk monitoring sub-areas and outputting disaster risk evolution characteristics; an analysis module for performing spatial consistency analysis according to the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, outputting a real-time disaster risk warning, and transmitting the real-time disaster risk warning to the risk monitoring user through a mobile terminal.

[0008] In a third aspect of the present application, an electronic device is provided, including: a memory for storing executable instructions; a processor for implementing the disaster risk warning method combining multi-source data analysis provided by the present application when executing the executable instructions stored in the memory.

[0009] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0010] First, the monitoring range is screened according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas. Then, the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas are used as geographical boundaries to perform cross-platform crawling of multi-source information to obtain K multi-source data for disaster monitoring. At the same time, geographical feature data of the K sub-areas are retrieved and output from an open GIS data platform according to the longitude and latitude spans of the K sub-areas. Next, the K multi-source data for disaster monitoring and the geographical feature data of the K sub-areas are fused and analyzed to obtain disaster risk characteristics of the K sub-areas. Then, according to the connection relationship of the K disaster risk monitoring sub-areas, risk diffusion fitting is performed on the disaster risk characteristics of the K sub-areas to output disaster risk evolution characteristics. Finally, spatial consistency analysis is performed according to the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, a real-time disaster risk warning is output, and the real-time disaster risk warning is transmitted to the risk monitoring user through a mobile terminal. The technical problems of lagging response and insufficient accuracy in disaster risk warning in the prior art are solved, and the technical effects of improving the timeliness and accuracy of disaster warning are achieved. Description of the Drawings

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts.

[0012] Figure 1 Schematic flow chart of the disaster risk early warning method combining multi-source data analysis provided by the embodiments of the present application;

[0013] Figure 2 Schematic structural diagram of the disaster risk early warning system combining multi-source data analysis provided by the embodiments of the present application;

[0014] Figure 3 Schematic structural diagram of an exemplary electronic device of the present application.

[0015] Explanation of reference numerals: screening module 11, information crawling module 12, retrieval module 13, fusion analysis module 14, fitting module 15, analysis module 16, processor 21, memory 22, input device 23, output device 24. Detailed implementation manners

[0016] The present application provides a disaster risk early warning method, system and electronic device combining multi-source data analysis, and solves the technical problems of lagging response and insufficient accuracy in disaster risk early warning in the prior art.

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0018] It should be noted that the terms "include" and "have" are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or modules not clearly listed or inherent to these processes, methods, products or devices.

[0019] Embodiment 1, as Figure 1 shown, the present application provides a disaster risk early warning method combining multi-source data analysis, wherein the method includes:

[0020] Screen the monitoring range according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas.

[0021] After obtaining the user's permission, the real-time latitude and longitude coordinates of the risk monitoring user are obtained through a mobile terminal or other devices. These coordinates are obtained through GPS or other positioning technologies and reflect the user's current geographical location. After obtaining the user's real-time location, the system determines the monitoring range according to preset rules. Usually, a circular area or an area of other shapes is set with the latitude and longitude coordinates as the center as the monitoring range. The size of this monitoring range can be dynamically adjusted according to different disaster types, regional characteristics, and user needs. Then, the system divides the monitoring range into several sub-areas, and each sub-area is called a disaster risk monitoring sub-area. The boundary of each sub-area is set by the latitude and longitude coordinate span to ensure that the range of each sub-area is suitable for disaster risk assessment and early warning. Specifically, the number of sub-areas (i.e., K sub-areas) can be flexibly adjusted according to the size of the monitoring range, the influence range of the disaster, and the risk assessment criteria.

[0022] Furthermore, the monitoring range is screened based on the real-time latitude and longitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas. The methods include:

[0023] Using the geographical location tracking technology of the mobile terminal to obtain the real-time latitude and longitude coordinates of the risk monitoring user; using the real-time latitude and longitude coordinates as the center of the circle and the reachable range of the response materials as the radius to frame the disaster risk monitoring range in the GIS twin model; presetting the monitoring range segmentation conditions, where the monitoring range segmentation conditions include the annular segmentation distance interval and the sector segmentation angle parameter; using the monitoring range segmentation conditions to perform multi-layer angular division on the disaster risk monitoring range to obtain the K disaster risk monitoring sub-areas.

[0024] First, through mobile terminal geographical location tracking technologies such as GPS positioning technology, base station positioning, or Wi-Fi positioning technology, the real-time longitude and latitude coordinates of the risk monitoring user are obtained, and the real-time longitude and latitude coordinates reflect the user's current location. Then, with the real-time longitude and latitude coordinates as the center and the set reachable range of response supplies (i.e., the distance within which disaster emergency supplies can effectively arrive) as the radius, the disaster risk monitoring range is delineated in the GIS (Geographic Information System) twin model. Among them, the GIS twin model can simulate the actual geographical environment during a disaster and provide digital support for monitoring and analysis by combining geographical spatial data and virtual models. Then, the system presets the monitoring range segmentation conditions, including the annular segmentation distance interval and the sector segmentation angle parameter. For example, the annular segmentation distance interval can be set to one interval every 5 kilometers, and the sector segmentation angle parameter can be set to one segment every 30 degrees to ensure that the monitoring area covers the main geographical directions; the annular segmentation divides the monitoring range into multiple concentric circular areas, and the sector segmentation divides the monitoring area according to the angle to adapt to the expansion characteristics of different disasters. Within the determined monitoring range, according to the preset annular segmentation distance interval, the monitoring range is divided into several levels. For example, from the center point outwards, it is successively divided into multiple concentric circular areas, and each circular area represents a monitoring sub-area; at the same time, according to the preset sector segmentation angle parameter, the area within each annular area is further divided into multiple sector sub-areas. Through multi-layer angle division, the monitoring range is finally divided into K disaster risk monitoring sub-areas, and these sub-areas represent the units of disaster risk monitoring. Each sub-area corresponds to a certain geographical area and can be independently evaluated for risk and warned.

[0025] Taking the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries, multi-source information cross-platform crawling is performed to obtain K multi-source data for disaster monitoring.

[0026] The specific positions of each sub-area are located according to the longitude and latitude spans of the sub-areas of each disaster risk monitoring sub-area (i.e., the geographical boundaries of the sub-areas). For example, if the longitude range of a sub-area is [30°E, 32°E] and the latitude range is [40°N, 42°N], then the longitude and latitude span of this sub-area is this longitude and latitude range. After the geographical boundaries of each sub-area are determined, the system can further obtain multi-source data for disaster monitoring, including meteorological data, geological data, environmental data, etc., from multiple data platforms (such as news platforms, social media platforms, weather forecast platforms, etc.) according to these boundaries.

[0027] Furthermore, taking the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries, multi-source information cross-platform crawling is performed to obtain K multi-source data for disaster monitoring. The method includes:

[0028] Crawl news report information using the first filtering condition to obtain first disaster feature information, where the first filtering condition includes a filtering keyword feature and a filtering time feature; crawl social media information using the second filtering condition to obtain second disaster feature information, where the second filtering condition includes a filtering topic feature and the filtering time feature; use the K sub-region longitude and latitude spans of the K disaster risk monitoring sub-regions as geographical boundaries, perform weather data crawling to obtain K sub-region weather data; perform geographical feature keyword recognition on the first disaster feature information and the second disaster feature information, and traverse the K sub-region longitude and latitude spans using the recognition results to decompose the first disaster feature information and the second disaster feature information into K integrated disaster information; store the K sub-region weather data and the K integrated disaster information in association with the K disaster risk monitoring sub-regions, and output the K multi-source disaster monitoring data.

[0029] Crawl news report information using the first filtering condition to obtain first disaster feature information (such as the type of disaster, the affected area, the occurrence time, etc.). The first filtering condition includes a filtering keyword feature and a filtering time feature. Among them, the filtering keyword feature includes keywords related to disasters (such as "earthquake", "flood", etc.), and the filtering time feature is used to screen for disaster events that occurred recently (such as reports within the past 24 hours). Secondly, crawl social media information using the second filtering condition to obtain second disaster feature information. The second filtering condition includes a filtering topic feature and a filtering time feature. Among them, the filtering topic feature mainly identifies relevant information through popular topics on social media (such as #typhoon, #earthquake, etc.), and the filtering time feature ensures that the content captured is the latest published content.

[0030] By taking the longitude and latitude spans of K disaster risk monitoring sub - regions as geographical boundaries, weather data crawling is performed to obtain sub - region weather data related to each sub - region. These weather data include meteorological parameters such as temperature, precipitation, wind speed, etc., which can provide real - time environmental data for disaster risk assessment. Specifically, data crawling can be carried out through open weather data interfaces (such as OpenWeatherMap, meteorological bureaus, etc.). The system will obtain accurate sub - region weather data from the weather platform according to the geographical boundaries (longitude and latitude ranges) of each sub - region. Then, natural language processing (NLP) algorithms, such as named entity recognition (NER), are used to identify geographical feature keywords in the first disaster feature information (from news reports) and the second disaster feature information (from social media), and through the recognition results, the longitude and latitude spans of K sub - regions are traversed to match the disaster information with the corresponding sub - regions, forming K integrated disaster information. Each integrated disaster information integrates disaster features from multi - source data (news, social media, meteorological data). Finally, the first disaster feature information, the second disaster feature information, and the weather data of each sub - region are associated and stored, and K multi - source data for disaster monitoring are output. These data integrate disaster information from news, social media, and meteorological data sources, and can provide comprehensive and accurate data support for disaster risk assessment and real - time early warning.

[0031] Retrieve and output K sub - region geographical feature data from the open GIS data platform according to the longitude and latitude spans of the K sub - regions.

[0032] An open GIS (Geographic Information System) data platform refers to those online platforms that publicly provide various geospatial data, such as OpenStreetMap, geographical information platforms of countries or regions, etc. These platforms usually provide a wide range of geographical information data, including terrain, landform, road network, building distribution, population density, land use type, etc. According to the longitude and latitude spans of each sub - region, a request is sent to the GIS data platform to retrieve the geographical feature data of the corresponding area, thereby outputting K sub - region geographical feature data. Each sub - region geographical feature data includes various geographical information such as terrain, land use type, transportation network, building distribution, etc.

[0033] Fusion - analyze the K multi - source data for disaster monitoring and the K sub - region geographical feature data to obtain the disaster risk characteristics of the K sub - regions.

[0034] The system preprocesses the acquired K disaster monitoring multi-source data (such as meteorology, geology, environment, etc.) with the geographical feature data of K sub-areas to ensure that the data are integrated under the same standard. Then, the disaster monitoring data of each sub-area is combined with the geographical feature data using weighted average method, machine learning model or multidimensional data fusion method to analyze the disaster risk characteristics of the area. Finally, the system outputs the disaster risk characteristics of K sub-areas, including disaster type, intensity, probability of occurrence, etc., to provide data support for subsequent disaster assessment and emergency response.

[0035] Furthermore, the K disaster monitoring multi-source data and the K sub-region geographic feature data are integrated and analyzed to obtain the disaster risk characteristics of the K sub-regions. The method includes:

[0036] A data validity window is preset; based on the data validity window and timestamp, the K disaster monitoring multi-source data are divided into K groups of stage monitoring multi-source data; disaster feature extraction is performed on the K groups of stage monitoring multi-source data to obtain K groups of stage disaster features, wherein the stage disaster features include stage disaster types and stage disaster urgency; based on the K sub-area geographic feature data, the K groups of stage disaster features are upgraded and compensated to obtain the K sub-area disaster risk features, wherein the K sub-area disaster risk features correspond to K groups of compensated disaster features.

[0037] The data validity window refers to the time range within which the acquired disaster monitoring data is considered valid. This window is set based on the timeliness and accuracy of the data. For example, the system can set a time window of 24 hours, 48 ​​hours or longer, indicating that the data within this time period can be used for disaster characteristics analysis.

[0038] In the process of disaster monitoring, each piece of data has a timestamp, which indicates the time when the data was collected. Based on the set data validity window and the timestamp of each piece of data, the system divides K disaster monitoring multi-source data into K groups of stage monitoring multi-source data. Each group of stage monitoring multi-source data represents a time period within the validity window. The system cuts the original data into multiple stage data sets based on the timestamp. For example, for a 24-hour validity window, all disaster monitoring data will be divided into multiple 24-hour stages based on the timestamp. The data in each stage has similar time characteristics, which is convenient for subsequent analysis.

[0039] By extracting disaster features from K groups of multi-source data for stage monitoring, the stage disaster type and stage disaster urgency can be extracted from each group of stage data. Specifically, the stage disaster type may be typhoon, earthquake, flood, etc. The stage disaster urgency (i.e., the severity and urgency of the disaster) is determined based on the potential impact of the disaster on the region or the proximity of the disaster. Optionally, the urgency of each stage disaster is evaluated based on the results of sentiment analysis and disaster information extraction. For example, the urgency of a typhoon may depend on its distance and intensity to the region, while the urgency of an earthquake is determined based on the distance from the epicenter to the sub-region and the magnitude of the epicenter.

[0040] After completing the disaster feature extraction, the system upgrades and compensates for the K groups of stage disaster features based on the geographical feature data of each sub-area. Upgrade compensation refers to weighted adjustment of the urgency of disaster features based on the geographical features of each sub-area (such as terrain, infrastructure, population density, etc.). For example, if a sub-area is located in a low-lying area, it may be greatly affected by floods, so its disaster urgency will be increased accordingly; if the transportation network in a sub-area is underdeveloped, the time for emergency response may be delayed, so the disaster urgency in the area will also increase accordingly. Finally, the system will output disaster risk characteristics for each sub-area based on these compensated disaster characteristics. The disaster risk characteristics of each sub-area include not only the type and urgency of the disaster, but also the risk value adjusted according to the geographical characteristics of the area.

[0041] Furthermore, the K groups of stage disaster characteristics are upgraded and compensated according to the geographical feature data of the K sub-areas to obtain the disaster risk characteristics of the K sub-areas, and the method includes:

[0042] Interactively obtain multiple baseline disaster urgency sets, multiple sample geographic feature data sets, and multiple upgraded disaster urgency sets for multiple sample disaster types; use the multiple baseline disaster urgency sets, multiple sample geographic feature data sets, and multiple upgraded disaster urgency sets as training data to construct multiple disaster upgraded analysis models; associate and store the multiple sample disaster types and multiple disaster upgraded analysis models to obtain a disaster upgraded model library; and schedule a real-time model to perform data analysis from the disaster upgraded model library based on the K groups of stage disaster characteristics and the K sub-area geographic feature data to output the K groups of compensatory disaster characteristics.

[0043] Multiple baseline disaster urgency sets, multiple sample geographical feature data sets, and multiple upgraded disaster urgency sets are obtained through interaction. The baseline disaster urgency set includes the disaster urgency calculated according to disaster types (such as typhoons, earthquakes, floods, etc.), representing the impact level of disasters at different times and intensities; the sample geographical feature data set contains sub-region geographical data related to disasters, such as terrain, infrastructure, traffic conditions, population density, etc. These features will affect the spread and impact of disasters. For example, floods may have a greater impact in low-lying areas, so the disaster urgency in low-altitude areas needs to be higher; the upgraded disaster urgency set is the data obtained by weighting and adjusting the baseline disaster urgency according to geographical features, reflecting the disaster risk differences caused by geographical differences in each region. Then, the multiple baseline disaster urgency sets, multiple sample geographical feature data sets, and multiple upgraded disaster urgency sets are used as training data to construct multiple disaster upgrading analysis models; by using machine learning algorithms (such as regression analysis, decision trees, support vector machines, etc.), these data sets are trained to generate models that can adjust disaster urgency based on geographical features. These models can identify the impacts of different types of disasters on different geographical environments and adjust the disaster urgency accordingly. Subsequently, the system stores the obtained multiple disaster upgrading analysis models in the disaster upgrading model library and stores the models in an associated manner according to disaster types and geographical features to ensure that different types of disasters correspond to different upgrading analysis models. Among them, the disaster upgrading model library contains the upgrading analysis models of all disaster types, and the system can call the appropriate model for data analysis in real time as needed. Finally, the system schedules real-time models from the disaster upgrading model library to perform data analysis based on the K sets of stage disaster features and the geographical feature data of K sub-regions, and outputs K sets of compensated disaster features. These compensated disaster features include information such as the disaster urgency, impact range, and warning level adjusted according to geographical features.

[0044] Furthermore, according to the K sets of stage disaster features and the geographical feature data of K sub-regions, scheduling real-time models from the disaster upgrading model library to perform data analysis and outputting the K sets of compensated disaster features, the method includes:

[0045] Aggregate the K groups of stage disaster features to obtain multiple stage disaster feature sets corresponding to the multiple sample disaster types, wherein each stage disaster feature in the multiple stage disaster feature sets has a region-stage time series identifier; cluster the K sub-area geographic feature data again according to the region-stage time series identifier to obtain multiple sub-area geographic feature data sets, wherein the multiple sub-area geographic feature data sets are mapped to multiple stage disaster feature sets; synchronize the mapping of the multiple sub-area geographic feature data sets and the multiple stage disaster feature sets to the multiple disaster escalation analysis models to perform disaster escalation compensation prediction to obtain multiple compensation disaster feature sets; perform data restoration on the multiple compensation disaster feature sets according to the region-stage time series identifier, and output the K groups of compensation disaster features.

[0046] By aggregating K groups of stage disaster features, multiple stage disaster feature sets corresponding to multiple sample disaster types are obtained. Each stage disaster feature set corresponds to the stage features of different disaster types, and each stage disaster feature set contains a region-stage time series identifier, indicating the area and time of the disaster. For example, the system aggregates the stage disaster features of various disaster types such as typhoons, earthquakes, and floods through disaster monitoring data, and each feature set marks the specific time period and area where the disaster occurs. Then, the system re-clustered the K sub-area geographic feature data according to the region-stage time series identifier to obtain multiple sub-area geographic feature data sets; this process clusters the geographic features of the sub-area (such as terrain, population density, transportation network, etc.) according to the time and regional characteristics of the stage disaster, so as to better combine them with the disaster characteristics. For example, the system will classify the low-lying areas with flood disaster feature sets, and associate the areas with obstructed traffic with earthquake disaster features. Subsequently, the system maps multiple sub-area geographic feature data sets with multiple stage disaster feature sets to ensure that each geographic feature data set can accurately match the corresponding disaster feature set, so that each sub-area can have relevant information such as disaster type, disaster intensity, and disaster urgency. After the mapping is completed, the system synchronizes the mapped data to the disaster escalation analysis model to make a disaster escalation compensation prediction. The disaster escalation analysis model compensates and adjusts the disaster urgency based on the geographic feature data to obtain multiple compensation disaster feature sets. For example, if transportation in a sub-area is inconvenient, it may cause the disaster impact time to be prolonged. The system will adjust the disaster urgency in the area and increase the priority of early warning and response. Finally, the system restores the data of multiple compensation disaster feature sets according to the region-stage time series identifier to ensure that the disaster feature sets are restored according to the corresponding region and stage time series, and output K groups of compensation disaster features. These compensation disaster features will cover information such as the type of disaster, scope of impact, and urgency, helping decision makers to take emergency response measures in a timely manner.

[0047] According to the connection relationship among the K disaster risk monitoring sub-areas, risk diffusion fitting is performed on the disaster risk characteristics of the K sub-areas to output disaster risk evolution characteristics.

[0048] In disaster risk monitoring, the connection relationship between K disaster risk monitoring sub-areas refers to the relationship between these sub-areas that influence or connect each other in geographic space. For example, neighboring sub-areas may share natural resources, transportation networks, infrastructure, etc. A disaster in a sub-area may affect neighboring sub-areas or even spread across multiple regions. The system can determine the connection relationship between these sub-areas by analyzing the geographic spatial layout, socioeconomic activities, and historical disaster data of the K sub-areas. For example, two sub-areas located in adjacent mountains may have a strong connection relationship due to flash floods, or two urban areas may have a strong influence relationship due to the close connection of the transportation network.

[0049] Based on the connection between K disaster risk monitoring sub-areas, the system uses the risk diffusion fitting method to calculate the intensity and time difference of disaster risk propagation from one sub-area to another. Then, the system aggregates K groups of stage disaster characteristics to obtain multiple stage disaster feature sets, and clusters the geographic feature data of K sub-areas according to the region-stage time series identifier to obtain multiple sub-area geographic feature data sets. Then, these sub-area geographic feature data sets are mapped with multiple stage disaster feature sets, and passed to the disaster upgrade analysis model for compensation prediction, and finally K groups of compensation disaster features are output.

[0050] Furthermore, according to the connection relationship of the K disaster risk monitoring sub-areas, risk diffusion fitting is performed on the disaster risk characteristics of the K sub-areas to output the disaster risk evolution characteristics, and the method includes:

[0051] According to the adjacent connection relationship of the K disaster risk monitoring sub-areas in the disaster risk monitoring range, the disaster evolution consistency of K groups of compensating disaster characteristics is judged, and according to the judgment result, the K disaster risk monitoring sub-areas are merged into H disaster risk monitoring sub-areas; in the GIS twin model, the disaster urgency vectors are marked for the H disaster risk monitoring sub-areas to obtain H disaster evolution vectors; by extending the H disaster evolution vectors, the disaster risk source and the disaster risk diffusion direction are located; with the disaster risk source as the starting point, the disaster risk attenuation characteristics are calculated according to the disaster risk diffusion direction and K groups of compensating disaster characteristics; the disaster risk source, the disaster risk diffusion direction and the disaster risk attenuation characteristics are stored in association, and the disaster risk evolution characteristics are output.

[0052] Among multiple disaster risk monitoring sub - regions, the system needs to determine whether there is consistency in the disaster risk characteristics among these sub - regions. Especially during the disaster propagation process, whether the disaster risk characteristics of adjacent sub - regions have similar evolution trends. If a disaster expands in one sub - region and affects adjacent sub - regions, the disaster characteristics of these sub - regions may show a consistent change pattern. Therefore, the system needs to judge the disaster evolution consistency of K groups of compensated disaster characteristics based on the connection relationships of K disaster risk monitoring sub - regions, to determine whether the disaster evolves step by step along adjacent sub - regions, generating similar risk characteristics. Among them, the connection relationships are determined based on factors such as the geographical locations, resource sharing, and transportation networks among the K sub - regions, reflecting the possibility and similarity of disaster propagation between adjacent sub - regions. Specifically, according to the geographical locations and disaster characteristics of the K disaster risk monitoring sub - regions, by analyzing the connection relationships between sub - regions, similarity calculations (such as Euclidean distance, Pearson correlation coefficient, etc.) are used to determine the similarity of the disaster risk characteristics of adjacent sub - regions. If the judgment results show that the disaster characteristics of multiple sub - regions are highly consistent, the system combines these sub - regions into one disaster risk monitoring partition. After the consistency judgment, the system combines the K disaster risk monitoring sub - regions into H disaster risk monitoring partitions, and each partition represents a group of sub - regions with similar risk characteristics. Such a combination can help simplify the subsequent disaster risk analysis and prediction processes.

[0053] The disaster urgency vector is a multi - dimensional vector, representing the degree of impact of a disaster on a certain area and the urgency of response. The system will generate a disaster urgency vector for each disaster risk monitoring partition, where each dimension represents a characteristic of the disaster (such as disaster intensity, impact range, urgency, etc.). The GIS twin model combines geographical information and disaster data. The system creates a disaster urgency vector for each disaster risk monitoring partition through this model. This vector includes the distribution characteristics and evolution characteristics of the disaster in space, providing basic data for subsequent disaster propagation and risk assessment.

[0054] During the disaster risk evolution process, the disaster risk vector will change with the passage of time and the expansion of the disaster impact range. The system extends the disaster evolution vector to simulate the propagation direction and speed of the disaster risk, thereby determining the risk source of the disaster (i.e., the location where the disaster initially occurred) and the propagation direction of the disaster. By extending the disaster evolution vector, the system can accurately locate the position of the disaster risk source, that is, the area where the disaster initially occurred; the disaster risk diffusion direction refers to the main direction of the disaster propagation from the source point. The system calculates the possible expansion direction of the disaster risk by analyzing the change trend of the disaster evolution vector.

[0055] The disaster risk attenuation characteristic refers to the phenomenon that as the distance of disaster propagation increases, its impact on each region gradually weakens. The system calculates the attenuation of disaster risk during the propagation process based on the disaster risk diffusion direction and the compensated disaster characteristics. The attenuation characteristics usually include the reduction of disaster impact intensity, the weakening of urgency, etc. Optionally, the attenuation of disaster risk over time and space is calculated through an attenuation model (such as an exponential attenuation model, a linear attenuation model, etc.) to predict the disaster risk intensity of each sub-region at different time nodes; generally, the farther the disaster is from the source point, the weaker the impact.

[0056] The system stores the disaster risk source, the disaster risk diffusion direction, and the disaster risk attenuation characteristic in an associated manner and outputs K sets of compensated disaster characteristics. These disaster risk evolution characteristics include the evolution process of the disaster, the propagation direction of the risk, the attenuation process, etc., providing data support for disaster emergency response, resource allocation, post-disaster recovery, etc.

[0057] Based on the position deviation between the disaster risk evolution characteristic and the real-time longitude and latitude coordinates, spatial consistency analysis is carried out, and real-time disaster risk warnings are output, and the real-time disaster risk warnings are conveyed to the risk monitoring users through a mobile terminal.

[0058] The system conducts spatial consistency analysis based on the deviation between the disaster risk evolution characteristic and the real-time longitude and latitude coordinates. The disaster risk evolution characteristics include the propagation direction, intensity, influence range, etc. of the disaster risk, which are obtained based on the risk evolution of the disaster from the source point to each sub-region; while the real-time longitude and latitude coordinates represent the location of the monitoring user. The system determines whether the area where the user is located is within the disaster risk range by comparing the predicted disaster influence range in the disaster risk evolution characteristic with the spatial deviation between the real-time user location. Specifically, the system will compare the longitude and latitude coordinates of the real-time user location with the geographical boundary in the disaster risk evolution characteristic. For example, by calculating the distance between the real-time location and the disaster influence range, it is determined whether the real-time location has entered the disaster influence area. If the position deviation is small and the real-time location has entered the disaster influence area, the system will determine that there is a relatively high disaster risk in the area where the user is located. Then, the system will output a real-time disaster risk warning according to the result of the spatial consistency analysis. If the risk level of the area where the user is located reaches the set threshold (such as high risk, very high risk, etc.), the system will generate corresponding disaster risk warning information, and these warning information include the disaster type, influence area, influence time, coping suggestions, etc., providing real-time and accurate disaster information for the user. Finally, the system will convey the real-time disaster risk warning information to the risk monitoring users through a mobile terminal. Through a mobile terminal (such as a smart phone, a tablet computer, etc.), the system will convey the disaster risk warning information to the user in the form of text messages, APP notifications, push messages, etc. In this way, users can receive disaster warnings in a timely manner and take corresponding preventive measures to minimize the losses caused by disasters.

[0059] Furthermore, perform spatial consistency analysis based on the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, output real-time disaster risk warnings, and convey the real-time disaster risk warnings to the risk monitoring users through a mobile terminal. The method includes:

[0060] Perform spatial consistency calculation on the real-time longitude and latitude coordinates and the disaster risk source to obtain the spatial angle of the risk source; preset deviation scale constraints, and judge the consistency between the spatial angle of the risk source and the disaster risk diffusion direction according to the deviation scale constraints; when the spatial angle of the risk source is consistent with the disaster risk diffusion direction, output an extreme disaster risk warning as the real-time disaster risk warning; when the spatial angle of the risk source is not consistent with the disaster risk diffusion direction, calculate the disaster diffusion deviation angle between the spatial angle of the risk source and the disaster risk diffusion direction; traverse the risk warning strategy library using the disaster diffusion deviation angle and the disaster risk attenuation characteristics, and call to output the real-time disaster risk warning; convey the real-time disaster risk warning to the risk monitoring user through a mobile terminal.

[0061] The system performs spatial consistency calculation based on the positional relationship between the real-time longitude and latitude coordinates and the disaster risk source to obtain the spatial angle of the risk source, and the spatial angle of the risk source represents the angular difference between the disaster risk source and the user's position. Optionally, by comparing the real-time longitude and latitude coordinates with the geographical coordinates of the disaster risk source, the angle value is calculated using coordinate calculation formulas (such as the earth surface distance formula, trigonometric functions, etc.), so as to help the system judge whether the user is on the risk path of the disaster.

[0062] The system makes a consistency judgment on the spatial angle of the risk source and the direction of disaster risk dispersion according to the deviation scale constraint. The deviation scale constraint means setting a threshold range, stipulating that when the difference between the spatial angle of the risk source and the direction of disaster risk dispersion is within the preset range, they are considered to be consistent, that is, the disaster risk spreads in the expected direction. If in the consistency judgment, it is found that the spatial angle of the risk source and the direction of disaster risk dispersion are consistent, an extreme disaster risk warning is output as the real-time disaster risk warning. If in the consistency judgment, it is found that the spatial angle of the risk source and the direction of disaster risk dispersion are not consistent, the system will further calculate the disaster dispersion deviation angle, that is, the angular difference between the spatial angle of the risk source and the direction of disaster risk dispersion. Among them, the calculation of the disaster dispersion deviation angle will reveal whether the disaster risk has abnormal diffusion and will affect the direction and scope of disaster risk assessment. After calculating the disaster dispersion deviation angle, the system traverses the risk warning strategy library through the disaster risk attenuation characteristics and calls the most appropriate warning strategy from it for further analysis. The disaster risk attenuation characteristics reflect the degree of attenuation of the disaster risk with the change of distance. The system will adjust the warning level according to this characteristic to ensure that users receive accurate disaster information and adjust response measures in a timely manner. Finally, the system conveys the output real-time disaster risk warning information to the risk monitoring users through the mobile terminal. The mobile terminal may include smartphones, tablets, etc. The system will notify users in a timely manner through text messages, APP push notifications, voice announcements, etc., to ensure that users can take preventive measures before the disaster occurs, thereby minimizing the losses caused by the disaster to the greatest extent.

[0063] In summary, the embodiments of the present application at least have the following technical effects:

[0064] First, the monitoring range is screened according to the real-time latitude and longitude coordinates of the risk monitoring users to obtain K disaster risk monitoring sub-areas. Then, taking the K sub-area longitude and latitude spans of the K disaster risk monitoring sub-areas as the geographical boundaries, multi-source information cross-platform crawling is performed to obtain K multi-source data for disaster monitoring. At the same time, according to the K sub-area longitude and latitude spans, K sub-area geographical feature data are retrieved and output from the open GIS data platform. Next, the K multi-source data for disaster monitoring and the K sub-area geographical feature data are fused and analyzed to obtain the disaster risk characteristics of the K sub-areas. Then, according to the connection relationship of the K disaster risk monitoring sub-areas, risk dispersion fitting is performed on the disaster risk characteristics of the K sub-areas, and the disaster risk evolution characteristics are output. Finally, spatial consistency analysis is performed according to the position deviation between the disaster risk evolution characteristics and the real-time latitude and longitude coordinates, and the real-time disaster risk warning is output, and the real-time disaster risk warning is conveyed to the risk monitoring users through the mobile terminal. It solves the technical problems of lagging response and insufficient accuracy in disaster risk warning in the prior art, and achieves the technical effects of improving the timeliness and accuracy of disaster warning.

[0065] Embodiment 2 is based on the same inventive concept as the disaster risk early warning method combined with multi-source data analysis in the previous embodiment. Figure 2 As shown, the present application provides a disaster risk early warning system combining multi-source data analysis, wherein the system includes:

[0066] The screening module 11 is used to screen the monitoring range according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas; the information crawling module 12 is used to use the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries, perform multi-source information cross-platform crawling, and obtain K disaster monitoring multi-source data; the retrieval module 13 is used to retrieve and output K sub-area geographical feature data from the open GIS data platform according to the longitude and latitude spans of the K sub-areas; the fusion analysis module 14 is used to fuse and analyze the K disaster monitoring multi-source data and the K sub-area geographical feature data to obtain the disaster risk characteristics of the K sub-areas; the fitting module 15 is used to perform risk diffusion fitting on the disaster risk characteristics of the K sub-areas according to the connection relationship of the K disaster risk monitoring sub-areas, and output the disaster risk evolution characteristics; the analysis module 16 is used to perform spatial consistency analysis based on the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, output the real-time disaster risk warning, and convey the real-time disaster risk warning to the risk monitoring user through the mobile terminal.

[0067] Furthermore, the screening module 11 is used to perform the following method:

[0068] The real-time longitude and latitude coordinates of the risk monitoring user are obtained by using the geographic location tracking technology of the mobile terminal; the disaster risk monitoring range is framed in the GIS twin model with the real-time longitude and latitude coordinates as the center of the circle and the reachable range of the response materials as the radius; the monitoring range segmentation conditions are preset, wherein the monitoring range segmentation conditions include annular segmentation distance intervals and sector segmentation angle parameters; the disaster risk monitoring range is divided into multiple circles and angles using the monitoring range segmentation conditions to obtain the K disaster risk monitoring sub-areas.

[0069] Furthermore, the fusion analysis module 14 is used to perform the following method:

[0070] A data validity window is preset; based on the data validity window and timestamp, the K disaster monitoring multi-source data are divided into K groups of stage monitoring multi-source data; disaster feature extraction is performed on the K groups of stage monitoring multi-source data to obtain K groups of stage disaster features, wherein the stage disaster features include stage disaster types and stage disaster urgency; based on the K sub-area geographic feature data, the K groups of stage disaster features are upgraded and compensated to obtain the K sub-area disaster risk features, wherein the K sub-area disaster risk features correspond to K groups of compensated disaster features.

[0071] Furthermore, the fusion analysis module 14 is configured to execute the following method:

[0072] Interactively obtain multiple benchmark disaster urgency sets, multiple sample geographical feature data sets, and multiple upgraded disaster urgency sets for multiple sample disaster types; use the multiple benchmark disaster urgency sets, multiple sample geographical feature data sets, and multiple upgraded disaster urgency sets as training data to construct multiple disaster upgrade analysis models; associatively store the multiple sample disaster types and multiple disaster upgrade analysis models to obtain a disaster upgrade model library; according to the K sets of stage disaster characteristics and K sub-region geographical feature data, schedule a real-time model from the disaster upgrade model library to perform data analysis and output the K sets of compensated disaster characteristics.

[0073] Furthermore, the fusion analysis module 14 is configured to execute the following method:

[0074] Aggregate the K sets of stage disaster characteristics to obtain multiple stage disaster characteristic sets corresponding to the multiple sample disaster types, where each stage disaster characteristic in the multiple stage disaster characteristic sets has a region-stage time sequence identifier; re-cluster the K sub-region geographical feature data according to the region-stage time sequence identifier to obtain multiple sub-region geographical feature data sets, where the multiple sub-region geographical feature data sets are mapped to the multiple stage disaster characteristic sets; synchronously map the multiple sub-region geographical feature data sets and multiple stage disaster characteristic sets to the multiple disaster upgrade analysis models for disaster upgrade compensation prediction to obtain multiple compensated disaster characteristic sets; perform data restoration on the multiple compensated disaster characteristic sets according to the region-stage time sequence identifier and output the K sets of compensated disaster characteristics.

[0075] Furthermore, the fitting module 15 is configured to execute the following method:

[0076] According to the adjacent connection relationship of the K disaster risk monitoring sub-regions within the disaster risk monitoring range, judge the consistency of disaster evolution for the K sets of compensated disaster characteristics, and merge the K disaster risk monitoring sub-regions into H disaster risk monitoring partitions according to the judgment result; in the GIS twin model, perform disaster urgency vector identification on the H disaster risk monitoring partitions to obtain H disaster evolution vectors; locate the disaster risk source and the disaster risk diffusion direction by extending the H disaster evolution vectors; starting from the disaster risk source, calculate the disaster risk attenuation characteristics according to the disaster risk diffusion direction and the K sets of compensated disaster characteristics; associatively store the disaster risk source, the disaster risk diffusion direction, and the disaster risk attenuation characteristics and output the disaster risk evolution characteristics.

[0077] Furthermore, the information crawling module 12 is configured to execute the following method:

[0078] Crawl news report information using the first filtering condition to obtain the first disaster feature information, where the first filtering condition includes a filtering keyword feature and a filtering time feature; crawl social media information using the second filtering condition to obtain the second disaster feature information, where the second filtering condition includes a filtering topic feature and the filtering time feature; use the K sub-region longitude and latitude spans of the K disaster risk monitoring sub-regions as geographical boundaries to perform weather data crawling to obtain K sub-region weather data; perform geographical feature keyword recognition on the first disaster feature information and the second disaster feature information, and traverse the K sub-region longitude and latitude spans by using the recognition results to decompose the first disaster feature information and the second disaster feature information into K integrated disaster information; store the K sub-region weather data and the K integrated disaster information in association with the K disaster risk monitoring sub-regions, and output the K disaster monitoring multi-source data.

[0079] Further, the analysis module 16 is used to execute the following method:

[0080] Perform spatial consistency calculation on the real-time longitude and latitude coordinates and the disaster risk source to obtain the risk source spatial angle; preset a deviation scale constraint, and perform consistency judgment on the risk source spatial angle and the disaster risk diffusion direction according to the deviation scale constraint; when the risk source spatial angle and the disaster risk diffusion direction are consistent, output an extreme disaster risk warning as the real-time disaster risk warning; when the risk source spatial angle and the disaster risk diffusion direction are not consistent, calculate the disaster diffusion deviation angle between the risk source spatial angle and the disaster risk diffusion direction; traverse the risk warning strategy library by using the disaster diffusion deviation angle and the disaster risk attenuation feature, and call to output the real-time disaster risk warning; convey the real-time disaster risk warning to the risk monitoring user through a mobile terminal.

[0081] Embodiment III Figure 3 It is a schematic structural diagram of an electronic device provided in Embodiment III of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 3 The displayed electronic device is only an example and should not bring any limitation to the functions and usage scopes of the embodiments of the present invention. As Figure 3 shown, the electronic device includes a processor 21, a memory 22, an input device 23, and an output device 24; the number of processors 21 in the electronic device can be one or more, Figure 3 Taking one processor 21 as an example, the processor 21, the memory 22, the input device 23, and the output device 24 in the electronic device can be connected through a bus or other means, Figure 3 Taking the connection through a bus as an example.

[0082] It should be noted that the above order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0084] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A disaster risk early warning method combining multi-source data analysis, characterized in that: The method comprises: The monitoring range is screened according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas; Taking the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographic boundaries, performing cross-platform crawling of multi-source information to obtain K disaster monitoring multi-source data; Retrieving and outputting geographic feature data of K sub-areas from an open GIS data platform according to the longitude and latitude spans of the K sub-areas; Fusion analysis of the K disaster monitoring multi-source data and the K sub-region geographic feature data to obtain disaster risk characteristics of the K sub-regions; According to the connection relationship of the K disaster risk monitoring sub-areas, risk diffusion fitting is performed on the disaster risk characteristics of the K sub-areas, and the disaster risk evolution characteristics are output; A spatial consistency analysis is performed based on the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, a real-time disaster risk warning is output, and the real-time disaster risk warning is communicated to the risk monitoring user via a mobile terminal.

2. The disaster risk early warning method combined with multi-source data analysis according to claim 1, characterized in that: The monitoring range is screened according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas. The method includes: Using the geographic location tracking technology of the mobile terminal to obtain the real-time latitude and longitude coordinates of the risk monitoring user; With the real-time longitude and latitude coordinates as the center of the circle and the reachable range of the response materials as the radius, the disaster risk monitoring range is framed in the GIS twin model; Preset monitoring range segmentation conditions, wherein the monitoring range segmentation conditions include annular segmentation distance intervals and sector segmentation angle parameters; The disaster risk monitoring range is divided into multiple layers according to the monitoring range segmentation conditions to obtain the K disaster risk monitoring sub-areas.

3. The disaster risk early warning method combined with multi-source data analysis as claimed in claim 2, characterized in that: The K disaster monitoring multi-source data and the K sub-area geographic feature data are integrated and analyzed to obtain the disaster risk characteristics of the K sub-areas. The method includes: Preset data validity window; According to the data validity window and timestamp, the K disaster monitoring multi-source data are divided into K groups of stage monitoring multi-source data; Extracting disaster features from the K groups of stage monitoring multi-source data to obtain K groups of stage disaster features, wherein the stage disaster features include stage disaster types and stage disaster urgency; The K groups of stage disaster characteristics are upgraded and compensated according to the geographical feature data of the K sub-areas to obtain the K sub-area disaster risk characteristics, wherein the K sub-area disaster risk characteristics correspond to K groups of compensation disaster characteristics.

4. The disaster risk early warning method combined with multi-source data analysis as claimed in claim 3, characterized in that: The K groups of stage disaster characteristics are upgraded and compensated according to the geographical feature data of the K sub-areas to obtain the disaster risk characteristics of the K sub-areas. The method includes: Interactively obtain multiple baseline disaster urgency sets, multiple sample geographic feature data sets and multiple upgraded disaster urgency sets for multiple sample disaster types; Using the multiple baseline disaster urgency sets, the multiple sample geographic feature data sets, and the multiple upgraded disaster urgency sets as training data, constructing multiple disaster upgraded analysis models; The plurality of sample disaster types and the plurality of disaster upgrade analysis models are stored in association to obtain a disaster upgrade model library; According to the K groups of stage disaster characteristics and the geographical feature data of K sub-areas, the real-time model is scheduled to perform data analysis from the disaster upgrade model library to output the K groups of compensation disaster characteristics.

5. The disaster risk early warning method combined with multi-source data analysis as claimed in claim 4, characterized in that: According to the K groups of stage disaster characteristics and K sub-area geographic feature data, the real-time model is dispatched from the disaster upgrade model library to perform data analysis, and the K groups of compensation disaster characteristics are output. The method includes: Aggregating the K groups of stage disaster features to obtain multiple stage disaster feature sets corresponding to the multiple sample disaster types, wherein each stage disaster feature in the multiple stage disaster feature sets has a region-stage time series identifier; Re-clustering the K sub-region geographic feature data according to the region-stage time series identifier to obtain a plurality of sub-region geographic feature data sets, wherein the plurality of sub-region geographic feature data sets are mapped to a plurality of stage disaster feature sets; Synchronize the mapping of the plurality of sub-area geographic feature data sets and the plurality of stage disaster feature sets to the plurality of disaster upgrade analysis models to perform disaster upgrade compensation prediction, and obtain a plurality of compensation disaster feature sets; Data restoration is performed on the multiple compensation disaster feature sets according to the region-stage time series identifiers, and the K groups of compensation disaster features are output.

6. The disaster risk early warning method combined with multi-source data analysis according to claim 5, characterized in that: According to the connection relationship of the K disaster risk monitoring sub-areas, risk diffusion fitting is performed on the disaster risk characteristics of the K sub-areas, and disaster risk evolution characteristics are output. The method includes: According to the adjacent connection relationship of the K disaster risk monitoring sub-areas in the disaster risk monitoring range, the disaster evolution consistency of the K groups of compensation disaster characteristics is judged, and according to the judgment result, the K disaster risk monitoring sub-areas are merged into H disaster risk monitoring sub-areas; In the GIS twin model, the H disaster risk monitoring zones are identified by disaster urgency vectors to obtain H disaster evolution vectors; By extending the H disaster evolution vectors, the disaster risk source and the disaster risk diffusion direction are located; Taking the disaster risk source as the starting point, the disaster risk attenuation characteristics are calculated according to the disaster risk diffusion direction and K groups of compensating disaster characteristics; The disaster risk source, disaster risk diffusion direction and disaster risk attenuation characteristics are stored in association, and the disaster risk evolution characteristics are output.

7. The disaster risk early warning method combined with multi-source data analysis according to claim 1, characterized in that: The longitude and latitude spans of the K disaster risk monitoring sub-areas are used as geographic boundaries, and multi-source information cross-platform crawling is performed to obtain K disaster monitoring multi-source data. The method includes: Applying a first filtering condition to crawl news report information to obtain first disaster feature information, wherein the first filtering condition includes filtering keyword features and filtering time features; Applying a second filtering condition to crawl social media information to obtain second disaster feature information, wherein the second filtering condition includes filtering topic features and filtering time features; Taking the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries, performing weather data crawling to obtain weather data of the K sub-areas; Performing geographic feature keyword recognition on the first disaster feature information and the second disaster feature information, and traversing the longitude and latitude spans of the K sub-areas using the recognition results to decompose the first disaster feature information and the second disaster feature information into K integrated disaster information; The K sub-area weather data and K integrated disaster information are stored in association with the K disaster risk monitoring sub-areas, and the K disaster monitoring multi-source data are output.

8. The disaster risk early warning method combined with multi-source data analysis according to claim 6, characterized in that: Performing spatial consistency analysis based on the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, outputting a real-time disaster risk warning, and transmitting the real-time disaster risk warning to the risk monitoring user via a mobile terminal, the method comprising: Performing spatial consistency calculation on the real-time latitude and longitude coordinates and the disaster risk source to obtain the spatial angle of the risk source; Preset a deviation scale constraint, and make a consistency judgment on the spatial angle of the risk source and the disaster risk diffusion direction based on the deviation scale constraint; When the spatial angle of the risk source is consistent with the direction of disaster risk diffusion, an extreme disaster risk warning is output as a real-time disaster risk warning; In the case where the risk source spatial angle is inconsistent with the disaster risk diffusion direction, calculating the disaster diffusion deviation angle between the risk source spatial angle and the disaster risk diffusion direction; The disaster diffusion deviation angle and disaster risk attenuation characteristics are used to traverse the risk warning strategy library, and a real-time disaster risk warning is called and output; The real-time disaster risk warning is transmitted to the risk monitoring user via a mobile terminal.

9. A disaster risk early warning system combining multi-source data analysis, characterized in that: A disaster risk early warning method combined with multi-source data analysis for implementing any one of claims 1 to 8, the system comprising: A screening module is used to screen the monitoring range according to the real-time longitude and latitude coordinates of the risk monitoring user to obtain K disaster risk monitoring sub-areas; An information crawling module, used to use the longitude and latitude spans of the K sub-areas of the K disaster risk monitoring sub-areas as geographical boundaries, perform multi-source information cross-platform crawling, and obtain K disaster monitoring multi-source data; A retrieval module, used for retrieving and outputting geographic feature data of K sub-areas from an open GIS data platform according to the longitude and latitude spans of the K sub-areas; A fusion analysis module is used to fuse and analyze the K disaster monitoring multi-source data and the K sub-area geographic feature data to obtain the disaster risk characteristics of the K sub-areas; A fitting module, used to perform risk diffusion fitting on the disaster risk characteristics of the K sub-areas according to the connection relationship of the K disaster risk monitoring sub-areas, and output the disaster risk evolution characteristics; The analysis module is used to perform spatial consistency analysis based on the position deviation between the disaster risk evolution characteristics and the real-time longitude and latitude coordinates, output a real-time disaster risk warning, and convey the real-time disaster risk warning to the risk monitoring user through a mobile terminal.

10. An electronic device, characterized in that: The electronic device comprises: A memory for storing executable instructions; The processor is used to implement the disaster risk warning method combined with multi-source data analysis as described in any one of claims 1 to 8 when executing the executable instructions stored in the memory.

Images