Web-based dynamic evolution visualization method for spatio-temporal process of disaster monitoring and early warning

Abstract

The application discloses a Web-based disaster monitoring and early warning spatio-temporal process dynamic evolution visualization method, and specifically comprises the following steps: constructing an expression model to describe disaster component elements and disaster processes as the basis for constructing a disaster spatio-temporal expression ontology, then taking the expression ontology as the top-level guide for constructing a knowledge graph, extracting specific knowledge from a large number of real three-dimensional models to construct a disaster knowledge graph, and designing a disaster data self-adaptive organization mode and a visualization data scheduling optimization mechanism to improve the compactness and relevance of data in a Web partition. The expression model and the disaster knowledge graph enable the computer to understand disaster elements and the multidimensional relationship between them, thereby realizing dynamic adjustment and optimization of three-dimensional data, effectively reducing the overhead of complex disaster scene loading and rendering, and realizing high-fluency and low-time-consumption three-dimensional display of disaster spatio-temporal process dynamic evolution.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional visualization technology of geographic data, and in particular to a web-based method for visualizing the dynamic evolution of spatiotemporal processes in disaster monitoring and early warning. Background Technology

[0002] Visualizing the spatiotemporal processes of disasters is of great significance for disaster management, decision-making, and emergency response. By visually displaying the dynamic evolution of disaster spatiotemporal processes, disaster managers, decision-makers, and emergency response teams can better understand the spatiotemporal characteristics, trends, and impacts of disasters, thereby enabling them to take effective measures for disaster early warning, response, and management.

[0003] Traditional methods for displaying and analyzing disaster spatiotemporal data have some shortcomings. For example, traditional two-dimensional map displays cannot intuitively show the three-dimensional characteristics and dynamic evolution of disaster spatiotemporal processes; traditional data tables and charts cannot effectively express the complex relationships and evolutionary trends of disaster spatiotemporal processes; traditional data organization and management methods cannot efficiently process large-scale spatiotemporal data and lack explicit representation and utilization of relationships and semantic information between data.

[0004] In the visualization of the dynamic evolution of disaster spatiotemporal processes, it is necessary to process, organize, and optimize large amounts of spatiotemporal data for efficient display and analysis. Database, caching, and indexing technologies can be used to organize, store, and manage spatiotemporal data, thereby improving data retrieval and display efficiency. While traditional database and caching technologies can improve data retrieval and display efficiency, they struggle to meet the demands of real-time performance and dynamic updates. Therefore, it is necessary to combine the latest technologies, such as distributed computing, stream processing, and machine learning, to achieve real-time processing and analysis of disaster spatiotemporal data. Furthermore, optimizing the user experience can be achieved through interactive visualization and virtual reality technologies, providing more intuitive and vivid data presentation methods to help users better understand and analyze the dynamic evolution of disaster spatiotemporal processes. In conclusion, web-based methods for organizing and optimizing disaster spatiotemporal process dynamic evolution visualization data require the integration of multiple technologies to achieve efficient data management, real-time processing, and user-friendly display, providing strong support for disaster early warning, emergency response, and decision support. Summary of the Invention

[0005] The purpose of this invention is to propose a Web-based dynamic evolution visualization method for disaster monitoring and early warning processes, addressing the existing needs for disaster spatiotemporal visualization and the shortcomings of traditional visualization technologies. The key lies in innovatively constructing a three-dimensional entity representation model that uniformly describes and integrates multi-source, multi-scale disaster data through semantic modeling of knowledge graphs. Based on this, the invention utilizes the unique graph structure and explicit / implicit spatiotemporal relationship capabilities of knowledge graphs to organize and schedule spatial data according to spatiotemporal associations, thereby achieving dynamic visualization that adapts to the spatiotemporal evolution of disasters.

[0006] A web-based method for visualizing the dynamic evolution of spatiotemporal processes in disaster monitoring and early warning includes the following steps:

[0007] S1. Construct expression models describing the components of a disaster and the process of a disaster, respectively. Based on the expression models, construct a disaster spatiotemporal expression ontology. Guided by the disaster spatiotemporal expression ontology, extract specific knowledge from the three-dimensional model data to construct a disaster knowledge graph.

[0008] S2, establish a consistent association between the nodes of the disaster knowledge graph and the three-dimensional model data, calculate the semantic similarity between the nodes, and set an adaptive organization mode for disaster data based on the semantic similarity;

[0009] S3, based on the disaster data adaptive organization mode, obtain a three-dimensional model dataset and set a visual data scheduling optimization mechanism for the disaster evolution process; the visual data scheduling optimization mechanism divides the three-dimensional model dataset according to the centrality of the nodes and realizes dynamic migration;

[0010] S4. Create a composite database, and schedule the composite database based on the adaptive organization mode of the disaster data and the visualization data scheduling optimization mechanism to realize the dynamic evolution visualization of the spatiotemporal process of disaster based on Web.

[0011] As a preferred embodiment of the present invention, the specific steps for constructing the disaster spatiotemporal representation ontology in step S1 include:

[0012] S11, Construct an expression model O describing the components of a disaster. The components of a disaster include the disaster-inducing environment, disaster-causing factors, and disaster-bearing bodies. The disaster-inducing environment includes stratum lithology and topographic relief. The disaster-causing factors include rainfall and human-made projects. The disaster-bearing bodies include land, farmland, houses, and roads. Each of the components of a disaster should have temporal, spatial, and attribute characteristics.

[0013] S12, Construct an expression model P describing the disaster process, wherein the disaster process is the process by which the spatial and attribute characteristics of the disaster components change over time, and consists of all the disaster components themselves participating in the process and their state sequence within a certain time range;

[0014] S13, Construct a disaster spatiotemporal process expression ontology. The disaster spatiotemporal process expression ontology includes spatiotemporal entities and spatiotemporal entity relationships. Construct the disaster spatiotemporal process expression ontology using a resource description framework according to the spatiotemporal entity division rules and the spatiotemporal entity relationship rules.

[0015] S14. Construct a disaster knowledge graph guided by the disaster spatiotemporal process expression ontology. Based on the spatiotemporal entity division and spatiotemporal entity relationship rules abstracted by the disaster spatiotemporal process expression ontology, extract specific knowledge from the table structure information stored in a large number of real three-dimensional models to construct the disaster knowledge graph.

[0016] As a preferred embodiment of the present invention, the object classification and relationship rules of the spatiotemporal entities include: the spatiotemporal entities are divided into disaster process entities with spatiotemporal changes and disaster component entities with spatiotemporal attributes; the relationship rules of the spatiotemporal entities are: the spatiotemporal entity relationships include the temporal order relationship and spatial position relationship between the disaster process entities, the inclusion relationship between the disaster process entities and the disaster component entities participating in the disaster process, and the temporal order relationship and spatial position relationship between the disaster component entities.

[0017] As a preferred embodiment of the present invention, the specific steps of setting the adaptive organization mode of disaster data based on the semantic similarity in step S2 include: calculating the semantic similarity between the same type of nodes based on the attribute information of the three-dimensional model data associated with the same type of nodes in the disaster knowledge graph, wherein the nodes of the disaster knowledge graph include disaster component node and disaster process node; and completing the same-area storage of the disaster knowledge graph entities, indexing and optimization of the associated semantic information, and compression of the associated data based on the semantic similarity between the same type of nodes.

[0018] As a preferred embodiment of the present invention, the specific steps for performing same-area storage of the disaster knowledge graph entities, indexing and optimizing the associated semantic information, and compressing the associated data based on the semantic similarity between the nodes of the same type include: performing same-area storage of the nodes of the same type with semantic similarity higher than the same-area storage threshold; performing indexing and optimization of the semantic information associated with the nodes of the same type with semantic similarity lower than the optimization threshold based on the graph path search algorithm; and compressing the large-scale data carried by the nodes of the same type with semantic similarity lower than the compression threshold using a data compression algorithm.

[0019] As a preferred embodiment of the present invention, the specific steps for setting the same-area storage threshold, index optimization threshold and compression threshold based on the semantic similarity include: respectively calculating the semantic similarity of the nodes of the same type in each group, calculating the intra-group semantic similarity standard deviation of the nodes of the same type in each group, and selecting the same-area storage threshold, the index optimization threshold and the compression threshold based on a specified multiple of the intra-group semantic similarity standard deviation.

[0020] As a preferred embodiment of the present invention, step S3 specifically includes:

[0021] S31, Calculate the centrality of the nodes with related relationships in the disaster knowledge graph;

[0022] S32, Based on the disaster data adaptive organization mode, process the target three-dimensional model data in the existing large-scale database, obtain the three-dimensional model dataset, map the centrality of the node to the three-dimensional model dataset, and divide the three-dimensional model dataset associated with the node into a disaster body change dataset that needs to be updated frequently and a backup dataset of the disaster environment that needs to be displayed in a fixed manner.

[0023] S33, Based on the three-dimensional data corresponding to the changed dataset, the dynamic migration between the Web cache partition and the loading partition is specifically achieved by extracting the target three-dimensional data of the cache partition and writing it into the loading partition through a synchronous dual-write mechanism. This releases the cache space and accelerates the visualization of the three-dimensional model data in the loading partition, which is a data scheduling optimization mechanism for the visualization of disaster evolution process.

[0024] As a preferred embodiment of the present invention, the specific steps of dividing the three-dimensional model dataset associated with the node into a disaster body change dataset that needs to be updated frequently and a backup dataset of the disaster environment that needs to be displayed permanently, based on the centrality, in step S32 include: calculating the standard deviation of the node centrality, and using a specified multiple of the standard deviation of the centrality as the centrality threshold for dividing the disaster body change dataset and the backup dataset, and dividing the three-dimensional model dataset associated with the node into a disaster body change dataset that needs to be updated frequently and a backup dataset of the disaster environment that needs to be displayed permanently according to the centrality threshold.

[0025] As a preferred embodiment of the present invention, the specific steps of dynamically migrating the three-dimensional data corresponding to the changed dataset between the Web cache partition and the loading partition in step S33 include: the dynamic migration specifically extracts the target three-dimensional data of the cache partition and writes it into the loading partition through a synchronous dual-write mechanism, thereby releasing the cache space and accelerating the visualization of the three-dimensional model data in the loading partition.

[0026] As a preferred embodiment of the present invention, step S4 specifically includes:

[0027] S41, Build a Web visualization framework, which includes different levels of cache rendering partitions and key request interfaces for connecting backend data;

[0028] S42, Based on the spatiotemporal elements of disaster, a composite database is created for the storage, organization, and scheduling of multi-source disaster data. The composite database includes a spatial database and a graphical database. The composite database is processed based on the disaster data adaptive organization mode and the visualization data scheduling optimization mechanism to complete data caching.

[0029] Specifically, the three-dimensional data and attribute information of the spatiotemporal process elements of the disaster are input into a composite database. Based on the composite database, the nodes associated with the processed three-dimensional model data are updated synchronously. In the cache rendering partition, the updated composite database is distinguished between the backup dataset and the changed dataset, and the dynamic migration is realized. The visualization data scheduling optimization mechanism realizes knowledge-driven dynamic scheduling of three-dimensional visual data. At this time, only data that needs to be displayed in a fixed manner and data that needs to be updated frequently will be parsed and loaded, thereby reducing memory consumption.

[0030] S43 enables interconnection between the data storage module and the data visualization module through key request interfaces, achieving smoother, lower-cost, and more demand-compliant visualization of the dynamic evolution of disaster spatiotemporal processes.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] This invention innovatively constructs a three-dimensional entity representation model for unified description and integration of multi-source, multi-scale disaster data. Utilizing the unique graph structure and key expressive capabilities of explicit and implicit spatiotemporal relationships inherent in knowledge graphs, spatial data is organized and scheduled according to spatiotemporal associations, thereby achieving dynamic visualization that adapts to the spatiotemporal evolution of disasters. This enables efficient management, real-time processing, and user-friendly display of disaster data, providing strong support for disaster early warning, emergency response, and decision support. Simultaneously, the adaptive disaster data organization mode and visualization data scheduling optimization mechanism designed in this invention can improve the density and relevance of data within Web partitions, enabling dynamic adjustment and optimization of three-dimensional data. Ultimately, this effectively reduces the overhead of loading and rendering complex disaster scenes, achieving a highly smooth and low-time-consuming three-dimensional display of the dynamic evolution of disaster spatiotemporal processes. Attached image description:

[0033] Figure 1 This is a schematic diagram of the process of organizing and optimizing data based on the dynamic evolution of disaster spatiotemporal processes according to the present invention;

[0034] Figure 2This is a schematic diagram of the steps and data integration process of a Web-based method for organizing and optimizing dynamic evolution visualization data of disaster spatiotemporal processes according to the present invention. Detailed Implementation

[0035] The present invention will be further described in detail below with reference to experimental examples and specific embodiments. However, this should not be construed as limiting the scope of the above-mentioned subject matter of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0036] Example 1

[0037] like Figure 1 and Figure 2 The diagram shown illustrates the process of a web-based disaster monitoring and early warning system that dynamically visualizes the spatiotemporal evolution of disaster events. The specific steps are as follows:

[0038] S1. Construct expression models describing the components of a disaster and the process of a disaster, respectively. Based on the expression models, construct a spatiotemporal expression ontology of a disaster. Guided by the spatiotemporal expression ontology of a disaster, extract specific knowledge from the three-dimensional model data to construct a disaster knowledge graph. The spatiotemporal expression ontology of a disaster includes spatiotemporal entities and spatiotemporal entity relationships.

[0039] Specifically, S1 includes the following steps:

[0040] S11, Construct an expression model O describing the components of a disaster. These components include the disaster-inducing environment, disaster-causing factors, and disaster-bearing bodies. The disaster-inducing environment includes geological lithology and topographic relief. The disaster-causing factors include rainfall and man-made engineering projects. The disaster-bearing bodies include land, farmland, houses, and roads. Each disaster component should have temporal (T), spatial (S), and attribute (A) characteristics, and can be formally represented as:

[0041] O =<T,S,A> ;

[0042] S12, Construct an expression model P describing the disaster process, whereby the disaster process is the process by which the spatial and attribute characteristics of disaster components change over time, consisting of all the disaster components themselves participating in the process and their characteristics within a certain time range (t0, t...). n ) state sequence Composition, which can be formally represented as:

[0043]

[0044]

[0045] Among them, f t It is an unordered set that records all characteristic information of the disaster components at time t in the time series;

[0046] S13, Construct a disaster spatiotemporal process representation ontology. The disaster spatiotemporal process representation ontology includes spatiotemporal entities and spatiotemporal entity relationships. The spatiotemporal entities are divided into disaster process entities E that have spatiotemporal changes. P And disaster component entity E with spatiotemporal attributes O The spatiotemporal entity relationship includes the disaster process entity E. P The temporal and spatial relationships between them, and the disaster process entity E P With the disaster component entity E involved in the disaster process O The inclusion relationship between them, and the disaster component entity E O Based on the temporal and spatial relationships between them, and in accordance with the spatiotemporal entity division rules and spatiotemporal entity relationships described above, a disaster spatiotemporal process expression ontology is constructed using a resource description framework;

[0047] S14. Construct a disaster knowledge graph guided by the disaster spatiotemporal process expression ontology. This process requires a large number of 3D model data samples for support. The disaster spatiotemporal process expression ontology is used as the top-level guide for constructing the knowledge graph. That is, based on the spatiotemporal entity division and spatiotemporal entity relationship rules abstracted from the disaster spatiotemporal process expression ontology, specific knowledge is extracted from the table structure information stored in a large number of real 3D models to construct the knowledge graph. For example, if the spatiotemporal entity of the disaster spatiotemporal process expression ontology is a landslide disaster entity, then knowledge of a landslide disaster entity including a certain landslide body and the landslide body's temporal attributes, spatial location, and feature information can be extracted.

[0048] First, this invention designs an expression model for the components of a disaster and an expression model for describing the disaster process. Second, based on the above expression models, a disaster spatiotemporal expression ontology is constructed. Finally, the disaster spatiotemporal expression ontology is used as the top-level guide for constructing a knowledge graph. That is, based on the element entities and relationship rules abstracted from the disaster spatiotemporal expression ontology, specific knowledge is extracted from the table structure information stored in a large number of real three-dimensional models. For example, if the ontology entity is a landslide disaster entity, knowledge such as a certain landslide body, time attributes, spatial location and feature information can be extracted.

[0049] S2, establish a consistent association between the nodes of the disaster knowledge graph and the 3D model data, calculate the semantic similarity between the nodes, and set an adaptive organization mode for disaster data based on the semantic similarity; the adaptive organization mode for disaster data includes adaptive organization modes for data storage, data retrieval, and data transmission;

[0050] Specifically, S2 includes the following steps:

[0051] S21, establish a consistent association between the nodes of the disaster knowledge graph and the 3D model data, and use semantic mapping to associate the concepts in the knowledge base with the data elements in the database. The nodes include disaster component nodes and disaster process nodes. Calculate the semantic similarity between the nodes based on the attribute information of the 3D model data associated with the nodes. Statistically calculate the semantic similarity of each group of nodes of the same type, and calculate the standard deviation of the intra-group semantic similarity of nodes of the same type within each group. Select the same-area storage threshold, the index optimization threshold, and the compression threshold based on a specified multiple of the intra-group semantic similarity standard deviation. The semantic similarity calculation refers to the Jaccard algorithm for text semantic computation in the NLP field. Semantic similarity is a necessary algorithm for mapping knowledge nodes to a computable space. Semantic similarity calculation is performed on the attribute information of the 3D model data carried by the entities.

[0052]

[0053] The E x E y For any two nodes to be computed in the knowledge graph, e x ,e y The vectorized features of the spatiotemporal and attribute information of the three-dimensional model data linked by the two nodes are called dif(e), which is used to calculate the deviation value of the vector itself.

[0054] S22, Store the nodes of the same type that have a semantic similarity higher than the same-area storage threshold in the same area;

[0055] Specifically, the same-area storage threshold is selected based on a specified multiple of the standard deviation of the semantic similarity within the group. Disaster component node nodes with semantic similarity higher than the same-area storage threshold are stored in the same area. For example, for a disaster body model with similar geological and topographical attributes, the semantic similarity of the disaster component node associated with the disaster body model is all higher than the same-area storage threshold. The same-area storage method for disaster process nodes in the disaster knowledge graph is consistent with that for disaster component nodes. Disaster process nodes with semantic similarity higher than the same-area storage threshold are stored in the same area. For example, for disaster processes in adjacent time periods or spatial locations, the semantic similarity of the disaster process nodes associated with the disaster processes in adjacent time periods or spatial locations is all higher than the same-area storage threshold. Completing the same-area storage of the disaster component node and the disaster process node can reduce data access overhead. It also facilitates the use of caching technology to cache popular data in memory, thereby accelerating data access speed.

[0056] S23, based on the graph path search algorithm, the semantic information associated with nodes whose semantic similarity is lower than the optimization threshold is indexed and optimized. Through the graph path search algorithm and the breadth-first search pruning strategy, the invalid path expansion for the nodes whose semantic similarity is lower than the optimization threshold is reduced.

[0057] S24, use a data compression algorithm to compress large-scale data carried by nodes of the same type with semantic similarity below the compression threshold; this can reduce data transmission overhead, thereby reducing data transmission bandwidth and latency, and improving data transmission speed and efficiency.

[0058] S3, based on the disaster data adaptive organization mode, obtain a three-dimensional model dataset and set a visual data scheduling optimization mechanism for the disaster evolution process; the visual data scheduling optimization mechanism divides the three-dimensional model dataset according to the centrality of the nodes, and drives the dynamic migration of the three-dimensional model data between Web caching and loading based on the divided dataset;

[0059] Specifically, the steps in S3 include:

[0060] S31, Calculate the centrality of nodes with relationships in the disaster knowledge graph. Centrality is a necessary algorithm for mapping the relationships between knowledge nodes to computational relationships. The centrality of a node is an indicator that measures the importance of a labeled graph node in the network. It is calculated using the general algorithm PageRank.

[0061]

[0062] Wherein, M(v j ) indicates pointing to node v j The set of nodes, L(v) j ) represents node v j The number of directed edges that connect to the target;

[0063] S32, based on the adaptive organization mode of disaster data, the target three-dimensional model data in the existing large-scale database is processed to obtain the three-dimensional model dataset. The centrality of the nodes is mapped to the three-dimensional model dataset, and the durability and reliability of the three-dimensional model dataset are evaluated. That is, the calculated graph node centrality is sorted according to the statistical feature strategy. The statistical features include extreme values, medians, normal distributions, etc. In this embodiment, the standard deviation of the node centrality is calculated, and a specified multiple of the standard deviation of the centrality is used as the centrality threshold to divide the disaster body change dataset and the backup dataset. Then, the three-dimensional model dataset associated with the nodes with centrality higher than the centrality threshold is regarded as the disaster body change dataset that needs to be updated frequently, and the three-dimensional model dataset associated with the nodes with centrality lower than the centrality threshold is regarded as the backup dataset of the disaster environment that needs to be displayed permanently. Thus, the disaster body change dataset that needs to be updated frequently and the backup dataset of the disaster environment that needs to be displayed permanently are distinguished.

[0064] S33, based on the three-dimensional data corresponding to the changed dataset, the web cache partition and the loading partition are dynamically migrated. The dynamic migration is specifically achieved by extracting the target three-dimensional data of the cache partition and writing it into the loading partition through a synchronous dual-write mechanism. This releases the cache space and accelerates the visualization of the three-dimensional model data in the loading partition, which is the data scheduling optimization mechanism for the visualization of the disaster evolution process.

[0065] S4. Create a composite database, and schedule the composite database based on the adaptive organization mode of the disaster data and the visualization data scheduling optimization mechanism to realize the dynamic evolution visualization of the spatiotemporal process of disaster based on Web.

[0066] Specifically, the steps in S4 include:

[0067] S41, Build a Web visualization framework capable of creating, loading and rendering large 3D datasets. The Web visualization framework includes different levels of cached rendering partitions and key request interfaces for connecting backend data.

[0068] S42, Based on the spatiotemporal elements of disaster, a composite database is created for the storage, organization, and scheduling of multi-source disaster data. The composite database includes a spatial database and a graphical database. The composite database is processed based on the disaster data adaptive organization mode and the visualization data scheduling optimization mechanism to complete data caching.

[0069] Specifically, the three-dimensional data and attribute information of the spatiotemporal process elements of the disaster are input into a composite database. Based on the composite database, the nodes associated with the processed three-dimensional model data are updated synchronously. In the cache rendering partition, the updated composite database is distinguished between the backup dataset and the changed dataset, and the dynamic migration is realized. The visualization data scheduling optimization mechanism realizes knowledge-driven dynamic scheduling of three-dimensional visual data. At this time, only data that needs to be displayed in a fixed manner and data that needs to be updated frequently will be parsed and loaded, thereby reducing memory consumption.

[0070] S43 enables interconnection between the data storage module and the data visualization module through key request interfaces, achieving smoother, lower-cost, and more demand-compliant visualization of the dynamic evolution of disaster spatiotemporal processes.

[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A Web-based method for visualizing the dynamic evolution of spatiotemporal processes in disaster monitoring and early warning, characterized in that: Includes the following steps: S1. Construct expression models describing the components of a disaster and the process of a disaster, respectively. Based on the expression models, construct a disaster spatiotemporal expression ontology. Guided by the disaster spatiotemporal expression ontology, extract specific knowledge from the three-dimensional model data to construct a disaster knowledge graph. S2, establish a consistent association between the nodes of the disaster knowledge graph and the three-dimensional model data, calculate the semantic similarity between the nodes, and set an adaptive organization mode for disaster data based on the semantic similarity; S3, Based on the adaptive organization mode of disaster data, obtain a three-dimensional model dataset and set up a data scheduling and optimization mechanism for the visualization of disaster evolution process; The visualization data scheduling optimization mechanism divides the 3D model dataset according to the centrality of the nodes and realizes dynamic migration; S4. Create a composite database, and schedule the composite database based on the adaptive organization mode of the disaster data and the visualization data scheduling optimization mechanism to realize the dynamic evolution visualization of the spatiotemporal process of disaster based on Web. S2 includes the following steps: S21, establish a consistent association between the nodes of the disaster knowledge graph and the 3D model data, and use semantic mapping to associate the concepts in the knowledge base with the data elements in the database. The nodes include disaster component nodes and disaster process nodes. Calculate the semantic similarity between the nodes based on the attribute information of the 3D model data associated with the nodes. Statistically calculate the semantic similarity of each group of nodes of the same type, and calculate the standard deviation of the intra-group semantic similarity of nodes of the same type within each group. Select a storage threshold, index optimization threshold, and compression threshold based on a specified multiple of the intra-group semantic similarity standard deviation. The semantic similarity calculation refers to the Jaccard algorithm for text semantic computation in the NLP field. Semantic similarity is a necessary algorithm for mapping knowledge nodes to a computable space. Semantic similarity calculation is performed on the attribute information of the 3D model data carried by the entities. In the formula, , Let be any two nodes to be computed in the disaster knowledge graph. The vectorized features of the spatiotemporal and attribute information of the 3D model data linked by the two nodes. and They represent and The deviation value; k is or The total number of dimensions; m is the total number of dimensions of the deviation vector; S22, Store the nodes of the same type that have a semantic similarity higher than the same-area storage threshold in the same area; S23, based on the graph path search algorithm, the semantic information associated with nodes whose semantic similarity is lower than the optimization threshold is indexed and optimized. Through the graph path search algorithm and the breadth-first search pruning strategy, the invalid path expansion for the nodes whose semantic similarity is lower than the optimization threshold is reduced. S24, use a data compression algorithm to compress the large-scale data carried by the same type of nodes whose semantic similarity is lower than the compression threshold.

2. The Web-based disaster monitoring and early warning spatiotemporal process dynamic evolution visualization method according to claim 1, characterized in that, The specific steps for constructing the disaster spatiotemporal representation ontology in step S1 include: after constructing representation models describing the components of a disaster and the process of a disaster, constructing spatiotemporal entities based on the representation models, classifying the objects of the spatiotemporal entities, defining the relationships between the spatiotemporal entities, and constructing the disaster spatiotemporal process representation ontology using a resource description framework according to the object classification and relationship rules of the spatiotemporal entities.

3. The Web-based disaster monitoring and early warning spatiotemporal process dynamic evolution visualization method according to claim 2, characterized in that, The object classification and relationship rules of the spatiotemporal entities include: the spatiotemporal entities are divided into disaster process entities with spatiotemporal changes and disaster component entities with spatiotemporal attributes; the relationship rules of the spatiotemporal entities are: the spatiotemporal entity relationships include the temporal order relationship and spatial position relationship between the disaster process entities, the inclusion relationship between the disaster process entities and the disaster component entities participating in the disaster process, and the temporal order relationship and spatial position relationship between the disaster component entities.

4. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 1, characterized in that, The specific steps of setting the adaptive organization mode of disaster data based on the semantic similarity in step S2 include: calculating the semantic similarity between nodes of the same type based on the attribute information of the three-dimensional model data associated with nodes of the same type in the disaster knowledge graph, wherein the nodes of the disaster knowledge graph include disaster component nodes and disaster process nodes; and completing the same-area storage of the entities in the disaster knowledge graph, indexing and optimization of the associated semantic information, and compression of the associated data based on the semantic similarity between the nodes of the same type.

5. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 4, characterized in that, Based on the semantic similarity between the nodes of the same type, the specific steps for storing the disaster knowledge graph entities in the same area, indexing and optimizing the associated semantic information, and compressing the associated data include: storing the nodes of the same type with semantic similarity higher than the same area storage threshold in the same area; indexing and optimizing the semantic information associated with the nodes of the same type with semantic similarity lower than the optimization threshold based on the graph path search algorithm; and compressing the large-scale data carried by the nodes of the same type with semantic similarity lower than the compression threshold using a data compression algorithm.

6. The Web-based disaster monitoring and early warning spatiotemporal process dynamic evolution visualization method according to claim 5, characterized in that, The specific steps for setting the same-area storage threshold, index optimization threshold, and compression threshold based on the semantic similarity include: calculating the semantic similarity of the nodes of the same type in each group, calculating the intra-group semantic similarity standard deviation of the nodes of the same type in each group, and selecting the same-area storage threshold, the index optimization threshold, and the compression threshold based on a specified multiple of the intra-group semantic similarity standard deviation.

7. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 1, characterized in that, Step S3 specifically includes: S31, Calculate the centrality of the nodes with related relationships in the disaster knowledge graph; S32, Based on the disaster data adaptive organization mode, process the target three-dimensional model data in the existing large-scale database, obtain the three-dimensional model dataset, map the centrality of the node to the three-dimensional model dataset, and divide the three-dimensional model dataset associated with the node into a disaster body change dataset that needs to be updated frequently and a backup dataset of the disaster environment that needs to be displayed in a fixed manner. S33, Based on the three-dimensional data corresponding to the changed dataset, the dynamic migration between the Web cache partition and the loading partition is specifically achieved by extracting the target three-dimensional data of the cache partition and writing it into the loading partition through a synchronous dual-write mechanism. This releases the cache space and accelerates the visualization of the three-dimensional model data in the loading partition, which is a data scheduling optimization mechanism for the visualization of disaster evolution process.

8. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 7, characterized in that, Step S32, which divides the 3D model dataset associated with the node into a disaster body change dataset that requires frequent updates and a backup dataset of the disaster environment that requires fixed display based on the centrality, specifically includes: calculating the standard deviation of the node centrality, and using a specified multiple of the standard deviation of the centrality as the centrality threshold for dividing the disaster body change dataset and the backup dataset, and dividing the 3D model dataset associated with the node into a disaster body change dataset that requires frequent updates and a backup dataset of the disaster environment that requires fixed display based on the centrality threshold.

9. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 8, characterized in that, The specific steps of dynamically migrating the 3D data corresponding to the changed dataset between the Web cache partition and the loading partition in step S33 include: the dynamic migration specifically extracts the target 3D data of the cache partition and writes it into the loading partition through a synchronous dual-write mechanism, thereby releasing cache space and accelerating the visualization of 3D model data in the loading partition.

10. The Web-based method for visualizing the spatiotemporal evolution of disaster monitoring and early warning processes according to claim 1, characterized in that, The specific steps of S4 include: S41, Build a Web visualization framework, which includes different levels of cache rendering partitions and key request interfaces for connecting backend data; S42, Create a composite database based on the spatiotemporal elements of the disaster. The composite database includes a spatial database and a graphical database. The composite database is processed based on the adaptive organization mode of the disaster data and the visualization data scheduling optimization mechanism to complete data caching. S43 enables interconnection between the data storage module and the data visualization module through key request interfaces, achieving smoother, lower-cost, and more demand-compliant visualization of the dynamic evolution of disaster spatiotemporal processes.

Images