A natural disaster emergency road network map service method and system

Abstract

The present invention discloses a natural disaster emergency road network map service method and system. The method includes: designing a disaster emergency road network organization model based on dynamic segmentation, and constructing a road network database based on the model; extracting disaster information from multi-source data and storing it in the disaster information database; matching the extracted disaster information with the road network data in the constructed road network database; modifying the road traffic weights of the matched damaged road sections, and using the weighted shortest path algorithm to provide an emergency path planning scheme after the disaster occurs; and visualizing the generated disaster emergency path on a map. The system includes a hardware infrastructure module, a data storage module, a disaster multi-source data access module, a disaster emergency road network map service module, and an access interface module. The method and system provided by the present invention have the characteristics of high efficiency, accuracy, timeliness, and multi-source, and can be used for disaster emergency road network management, disaster emergency road network map services, and other aspects.

Description

Technical Field

[0001] The present invention belongs to the technical field of disaster information services, and in particular relates to a natural disaster emergency road network map service method and system. Background Art

[0002] Disasters can bring about a series of challenges for road networks in disaster-stricken areas. On the one hand, after a natural disaster, road networks are often severely damaged or even paralyzed. For example, earthquakes can cause road collapses, and torrential rain and flooding can destroy roads or cause flooding. On the other hand, the demands of timely evacuation of personnel from disaster-stricken areas to safe areas, the timely delivery of supplies to key disaster areas, and the prompt and effective transfer and treatment of rescued and injured personnel all lead to a sharp increase in the use of road networks in disaster-stricken areas, further increasing the load on the road networks. To address these issues and needs, scholars at home and abroad have been developing road network mapping services for natural disaster emergencies and building disaster emergency road network mapping service systems.

[0003] However, the current natural disaster emergency road network map service still has many shortcomings, which are mainly reflected in the following three aspects.

[0004] (1) The traditional “arc-node” road network model easily causes information redundancy and reduces storage and computing efficiency, making it impossible to provide accurate and fast services.

[0005] In natural disaster scenarios, linear features are not static but highly dynamic and complex. The traditional "arc-node" model struggles to store these dynamic and complex features, leading to a series of problems such as increased storage difficulty, reduced storage efficiency, and an inability to effectively integrate storage and computation. This is particularly evident in the following four scenarios: ① Disasters can occur anywhere on a road; ② Disasters are located using linear descriptions; ③ Multiple disasters occur on the same road; and ④ The same road can have both point and linear disasters. For scenario 1, the "arc-node" model must interrupt the road at the disaster node and generate new arcs to accurately represent the disaster information. For scenario 2, the "arc-node" model requires extensive computation to determine the disaster's location (converting linear location to coordinate location). For scenario 3, an arc segment has segmentation attributes, requiring further segmentation to store the disaster. For scenario 4, an arc segment has multiple attributes, resulting in redundant storage of arc segments that lack certain attributes.

[0006] (2) Unable to provide real-time map service.

[0007] Delayed information services can create time lags and pose risks to rescue operations. In the race against time, delayed services offer little practical value. For example, if the planned route doesn't account for road damage during the delay, rescue operations could be disrupted, leading to serious consequences.

[0008] (3) The study did not utilize multi-source data.

[0009] Disaster information that is too sparse can easily lead to omissions, resulting in inaccurate services. Disaster information comes from a variety of sources, including social media data, remote sensing imagery, and data from professional institutions. Combining these multiple data sources helps provide a comprehensive and detailed understanding of the disaster. However, using a single source of disaster information can lead to inaccurate disaster information services and low service value. Summary of the Invention

[0010] In response to the shortcomings of existing technologies, the present invention provides a natural disaster emergency road network map service method and system, which can meet the challenges of rapid update of road network information and timely service after a disaster occurs.

[0011] In order to achieve the above-mentioned object, the technical solution provided by the present invention is a natural disaster emergency road network map service method, comprising the following steps:

[0012] Step 1: Design a disaster emergency road network organization model based on dynamic segmentation and build a road network database based on the model;

[0013] Step 1.1: Use the linear referencing algorithm to dynamically store and express any point on the road;

[0014] Step 1.2: Use the linear referencing algorithm to dynamically store and express any subset of roads;

[0015] Step 1.3, dynamically segmenting the roads using the dynamic storage representation of points and lines in steps 1.1 and 1.2, and constructing a disaster emergency road network model based on the dynamic segmentation;

[0016] Step 2: extract disaster information from multi-source data and store it in the disaster information database;

[0017] Step 3, matching the disaster information extracted in step 2 with the road network data in the road network database constructed in step 1;

[0018] Step 4: Modify the road traffic weights for the damaged road sections matched in step 3, and use the weighted shortest path algorithm to provide an emergency path planning solution after the disaster occurs;

[0019] Step 5: Map visualization of the disaster emergency path generated in step 4.

[0020] Moreover, the step 1.1 uses the linear reference algorithm to dynamically store and express any point on the road by storing and expressing the position of a point on the road by its distance or percentage to the starting point of the road. The specific operation is as follows: ① Let any point on the road R be A, and the starting point of the road R be R s , the end point is R e ;②Calculate the A and R on road R s The distance is recorded as L s , or calculate A and R e The distance is recorded as L e , and calculate the length of road R, recorded as L R ;③Calculate the length of A on road R as follows: or Will or It is called the measurement value of point A on road R, and is expressed as M or express.

[0021] Moreover, any subset R of the roads R in step 1.2 Sub Both have a starting point and an end point, denoted as point S and point E. The linear reference algorithm is used to calculate the measurement values ​​of the starting point and the end point, denoted as and Then the tuple A subset R that can uniquely and dynamically express any road R Sub ; The measured value of the starting point FromM or End point measurement ToM or

[0022] Furthermore, the disaster emergency road network model constructed in step 1.3 includes a road table, a point event table, a line event table, and a disaster table. All roads are stored in the road table, with the road ID as the primary key and the start and end coordinates. Disaster point events are stored in the point event table, with the point event ID as the primary key and the measurement value, the associated road ID, the associated disaster ID, and (road ID, disaster ID) as foreign keys. Additionally, some thematic attributes, such as the point event description, occurrence time, impact radius, and severity level, are also stored. Disaster line events are stored in the line event table, with the line event ID as the primary key and the start and end measurement value, the associated road ID, the associated disaster ID, and (road ID, disaster ID) as foreign keys. Additionally, some thematic attributes, such as the line event description, occurrence time, impact area, and severity level, are also stored. All disasters are stored in the disaster table, with the disaster ID as the primary key and the disaster type, occurrence time, end time, and disaster description as disaster attribute information.

[0023] Moreover, in step 3, the source disaster data is summarized into three types of data: raster data, coordinate data, and linear expression data. Different matching methods need to be selected for different types of data, namely: a method for extracting and matching damaged roads based on raster data, a method for extracting and matching damaged roads based on coordinate data, and a method for extracting and matching damaged roads based on linear expression data.

[0024] The method for extracting and matching damaged roads based on raster data is to extract the disaster area, convert the raster data of the disaster area into vector data, and generate Geohash codes for the disaster area vector data and road network data to obtain the number of disaster area vectors with Geohash indexes. The character length of the Geohash code for the disaster area and road network data is n. The Geohash code of a single disaster vector or a single road is the set of intersecting Geohash grids, that is: ,whereLength Geohash= n. After Geohash encoding the disaster vectors and the road vectors in the database, the Geohash grid is traversed to perform an overlay analysis of the disaster vectors and road vectors within the grid. After extracting the damaged roads, the disaster emergency road network model proposed in step 1 is used to match them with the road network in the database. The specific matching method is as follows: ① Calculate the length of the overlaid "disaster-road" vector data. If the length is less than the resolution of the raster data, the disaster event is considered a point event. If the length is greater than the resolution of the raster data, the disaster event is considered a line event. ② Use the coordinates to calculate the measurement value M of the point event or the measurement value (FromM, ToM) of the line event. ③ Obtain the road ID and disaster ID corresponding to the point event or line event, and assign an ID to the point event or line event. ④ Store all necessary fields and event attributes in the point time table or line event table.

[0025] There are two situations for extracting and matching damaged roads based on coordinate data: disaster information is recorded in the form of coordinates of a single point and disaster information is recorded in the form of coordinates of the starting point and the end point.

[0026] If the disaster information is recorded in the form of coordinates of a single point, the measurement value M is calculated and stored in the point event table. The calculation steps of the measurement value M are as follows: ① Let the disaster point be A, perform Geohash encoding on A, and record it as Geohash A ;② Filter out all road sets R whose road Geohash codes contain the Geohash of point A, that is: R={Road|Geohash A ∈Geohash Road}; ③ Calculate the road R closest to point A in the set R A , that is: R A =min Road∈R {Distance A,Road}; ④ Project point A onto road R A ⑤ Calculate the projection point A′ on the road R A The measured value M on the road is also the value of point A on the road R A The measurement value M on the road is also the measurement value M of point A in the entire road network. ⑥ Assign an ID to the point disaster, and the measurement value M, road R A The ID of the event, the ID of the disaster, and the thematic attributes of the point disaster are stored together in the point event table.

[0027] If the disaster is recorded in the form of coordinates of the starting point and the end point, it will be stored in the line event table, with the starting point and the end point being point S and point E respectively. The starting point S and the end point E can form a complete path through the connected roads. The linear reference algorithm can be used to obtain the coordinates of point S on the nearest road R. SThe measured value FromM on the point E and the point E on the nearest road R E If the starting point and the end point are on the same road, the line disaster corresponds to only one line event, and the measurement value of the line event is (FrommM, ToM); if the starting point and the end point are not on the same road, the line disaster corresponds to multiple elements, and the measurement value of the line event is (FrommM, ToM). S and R E All connected roads between them will be recorded as line events, with FromM=0,ToM=1. S On the road R, ToM = 1, E Finally, an ID is assigned to the line disaster and stored in the line event table along with the measurement value (FromM, ToM), the ID of the corresponding road, the ID of the disaster, and the thematic attributes of the line disaster.

[0028] The method for extracting and matching damaged roads based on linear representation data requires extracting four types of semantic information from the text describing the disaster: the name of the reference point, the name of the reference road, the reference direction, and the reference distance. The Levensh distance is used to compare the names of the reference points and reference roads with the names of points and roads in the road network database. The points and roads with the highest similarity are selected as the actual reference points and reference roads. A linear referencing algorithm is used to convert the reference point positions and reference distances into measurements of the reference roads. Using the reference direction, the measurements of the reference points and reference distances are added or subtracted to obtain the measurements of the point and line disasters. Finally, each point or line disaster is assigned an ID and stored in a point event table or line event table along with the measurement value, reference road ID, disaster ID, and thematic attributes of the point or line disaster. Linear disasters in linear representations must form a complete path through connected roads. Line events involving multiple roads and features are stored in the same manner as coordinate data.

[0029] Moreover, in step 4, by changing the weight of the road corresponding to the disaster event, real-time disaster emergency path planning can be carried out according to the disaster situation. In the real-time disaster emergency path planning service, three different factors of priority path planning are provided, namely distance priority, time priority, and road level priority. For distance priority, the length of the road can be directly used as the weight. If the length of the road is d, the weight is: w = d. For time priority, it is necessary to specify an expected speed, and then divide the length of the road by the minimum value of the expected speed and the speed limit as the weight. Let the expected speed be v e , the speed limit is v l , then the weight is: For road priority, the product of the road length and the road priority is used as the weight. If the road priority is P, the weight is: w = d × P. By defining different weights above, running the Dijkstra algorithm can obtain different desired shortest path planning results.

[0030] The present invention also provides a natural disaster emergency road network map service system, which includes a hardware infrastructure module, a data storage module, a disaster multi-source data access module, a disaster emergency road network map service module and an access interface module.

[0031] Hardware infrastructure module: The hardware uses storage devices with large storage space and high-performance servers to ensure the timeliness of services.

[0032] Data storage module: store existing basic geographic data and road network data in the database.

[0033] Disaster multi-source data access module: provides an upload entry for disaster multi-source data. The uploaded disaster multi-source data will be automatically processed by the system according to the metadata input by the user, generating point events and line events, and storing them in the corresponding point event table and line event table.

[0034] Disaster emergency road network map service module: provides four services: real-time disaster emergency route planning, road network publishing / sharing, road network disaster visualization, and disaster event management. Users can use the system to perform emergency route planning services based on the real-time disaster situation and can see real-time visualization information of the disaster situation. In addition, users can publish or share local road network sets according to their needs, and can also manage input disaster events and update their life cycle in real time.

[0035] Access interface module: provides two access modes, one is to call the service directly through the network API, and the other is to access it through the built Web platform.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] 1) Applying dynamic segmentation to disaster scenarios. Through dynamic segmentation of road network organization, the storage, calculation, and management efficiency of disaster events can be improved. Any disaster event with dynamically changing spatial and attribute information can be well matched with the road network through point events and line events.

[0038] 2) Dynamic segmentation solves the storage problem of dynamic complex linear features. ① If a disaster occurs at any location on a road: use the dynamic expression of points and lines to calculate its measurement value, and store the measurement value and the disaster thematic attribute together as a feature; ② Disasters are located through linear description: In linear description, reference points and linear distances are involved. The measurement value of the reference point is calculated and converted into a measurement value. A simple algebraic operation is performed on the two measurement values ​​to obtain the measurement value of the described object, which can be stored; ③ There are many disasters on the same road: Since the dynamically expressed points and lines do not share the same table with the actual road, the same road can correspond to multiple point disasters and line disasters; ④ If the same road has both point disasters and line disasters: according to the different geometric expressions of the disasters, the measurement values ​​are calculated and stored separately.

[0039] 3) Damaged road sections are extracted from multi-source disaster data and matched to the road network in the database. Point events and line events are generated based on the dynamically segmented road network model and stored in the database. This breaks down the barriers between multi-source data and incorporates multiple data into the same road network model, providing a basis for timely disaster emergency map services. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Flowchart of an embodiment of the present invention.

[0041] Figure 2 This is a schematic diagram of a natural disaster emergency road network model provided by an example of the present invention.

[0042] Figure 3 This is a flow chart of GF-3 data preprocessing provided by an example of the present invention.

[0043] Figure 4 This is a workflow diagram for extracting and matching damaged road sections based on raster data provided by an example of the present invention.

[0044] Figure 5 This is a workflow diagram for extraction and matching of damaged road sections based on coordinate data provided by an example of the present invention.

[0045] Figure 6 This is a workflow diagram for extraction and matching of damaged road sections based on linear expression data provided by an example of the present invention.

[0046] Figure 7 This is a general architecture diagram of a natural disaster emergency road network map service system provided by an example of the present invention. DETAILED DESCRIPTION

[0047] The present invention provides a natural disaster emergency road network map service method and system. The following takes the emergency road network map service under the flood disaster caused by the heavy rain on July 20 in a certain city as an example, and further illustrates the technical solution of the present invention in conjunction with the accompanying drawings.

[0048] like Figure 1 As shown, the process of a natural disaster emergency road network map service method according to the first embodiment of the present invention includes the following steps:

[0049] Step 1: Design a disaster emergency road network organization model based on dynamic segmentation and build a road network database based on the model.

[0050] Step 1.1: Use the linear referencing algorithm to dynamically store and express any point on the road.

[0051] In the current mainstream "arc-node" model, nodes are stored by coordinates, and arcs are stored by coordinate sets. Using linear reference to dynamically store and express any point on the road is to store and express the position of a point on the road by its distance or percentage from the starting point of the road, rather than directly storing and expressing its coordinates. The specific algorithm of linear reference is: ① Let any point on road R be A, and the starting point of road R be R s , the end point is R e (The starting point and end point of the road are determined when it is physically stored); ② Calculate the distance between A and R on road R s The distance is recorded as L s , or calculate A and R e The distance is recorded as L e , and calculate the length of road R, recorded as L R ;③Calculate the length of A on road R as follows: or In linear referencing, or It is called the measurement value of point A on road R, and is expressed as M or express.

[0052] Step 1.2: Use the linear referencing algorithm to dynamically store and express any subset of roads.

[0053] Any subset R of roads R Sub There are starting points and ending points, denoted as points S and E. The linear referencing algorithm can be used to calculate the measurement values ​​of the starting point and the ending point, denoted as and Then the tuple A subset R of any road R can be uniquely and dynamically expressed Sub In linear referencing, the measured value of the starting point FromM or End point measurement ToM or

[0054] Step 1.3: Use the dynamic storage expression of points and lines in steps 1.1 and 1.2 to dynamically segment the road, and build a disaster emergency road network model based on the dynamic segmentation.

[0055] In dynamic segmentation, the geometric and attribute information of roads is stored in the road table; dynamically changing events related to roads are stored separately in the event table. In disaster scenarios, disasters can generally be categorized as point, linear, and area disasters. However, road-specific disasters are limited to point and linear disasters, as area disasters can always be refined to point or linear disasters through their intersection with roads. Based on this, point event tables and line event tables are designed based on dynamic segmentation. Point disasters are converted into point events and stored in the point event table, while linear disasters are converted into line events and stored in the line event table. The design of event tables should take into account the diverse nature of disasters (such as earthquakes and floods), and the same type of disaster should be differentiated based on its different time and location. Separate tables should be created for different types of disasters: a point event table and a line event table for each type. To account for the heterogeneity of the same type of disaster, disaster tables should be added to differentiate them.

[0056] Based on the above analysis, the disaster emergency road network model was designed, consisting of a road table, a point event table, a line event table, and a disaster table. All roads are stored in the road table, with the road ID as the primary key and road attribute information such as the start and end coordinates. Disaster point events are stored in the point event table, with the point event ID as the primary key and the measurement value, the associated road ID, and the associated disaster ID. (Road ID, Disaster ID) serves as a foreign key. Additionally, thematic attributes such as the point event description, occurrence time, impact radius, and severity level are also stored. Disaster line events are stored in the line event table, with the line event ID as the primary key and the start and end measurement values, the associated road ID, and the associated disaster ID. (Road ID, Disaster ID) serves as a foreign key. Additionally, thematic attributes such as the line event description, occurrence time, impact area, and severity level are also stored. All disasters are stored in the disaster table, with the disaster ID as the primary key and the disaster type, occurrence time, end time, and description as attributes.

[0057] Step 2: Extract disaster information from multi-source data and store it in the disaster information database.

[0058] Step 3: Match the disaster information extracted in step 2 with the road network data in the road network database constructed in step 1.

[0059] Although multi-source disaster data comes from different sources and is organized in various ways, it can be categorized into three types: raster data, coordinate data, and linear representation data. Different matching methods are required for each type of data. Using techniques such as change detection and text semantic extraction, damaged roads can be extracted from this data and matched with the traffic network data in the database.

[0060] Step 3.1: Extract and match damaged roads based on raster data.

[0061] Floods were extracted using GF-3 image data. GF-3 L1A-level HHHV dual-polarization FSII images taken before the disaster (July 15) and during the disaster (July 20 and July 22) were selected. First, the GF-3 image data was preprocessed, as follows: Figure 3 As shown in the figure, the preprocessing process is as follows: ① Radiometric correction; ② Complex data conversion; ③ Multi-look processing; ④ Geocoding; ⑤ EnLee filtering; ⑥ Conversion to dB imagery; ⑦ Geometric correction. After preprocessing, the HV imagery extracts water bodies more clearly and accurately than the HH imagery. Based on an empirical threshold, dB<-27 indicates water bodies, while dB≥-27 indicates non-water bodies. After obtaining the water bodies for July 15, July 20, and July 22, the flood extent on July 20 and July 22 relative to July 15 can be determined by subtracting the water body extents.

[0062] like Figure 4 As shown in the figure, the extracted flood range raster data is first converted into flood range vector data through a raster-to-vector operation. Then, Geohash codes are generated for the flood vectors and road network data to obtain the number of disaster range vectors with Geohash indexes. The Geohash code character length of the disaster range and road network data is 6. The Geohash code of a single disaster vector or a single road is the set of intersecting Geohash grids, that is: whereLength Geohash = 6. After the disaster vectors and the road vectors in the database are Geohash-encoded, the disaster vectors and road vectors in the grid are overlaid and analyzed by traversing the Geohash grid.

[0063] After extracting the damaged roads, the disaster emergency road network model proposed in step 1 is used to match them with the road network in the database. The specific matching method is as follows: ① Calculate the length of the superimposed "disaster-road" vector data. If the length is less than or equal to 10 meters (the resolution of the GF-3 satellite is 10 meters), the disaster event is considered a point event. If the length is greater than 10 meters, the disaster event is considered a line event. ② Use the coordinates to calculate the measurement value M of the point event or the measurement value (FromM, ToM) of the line event. ③ Obtain the road ID and disaster ID corresponding to the point event or line event, and assign an ID to the point event or line event (either a sequential ID or a UUID). ④ Store all necessary fields and the event attributes in the point time table or line event table.

[0064] Step 3.2: Extract and match damaged roads based on coordinate data.

[0065] If the disaster information is recorded in the form of coordinates of a single point, it is stored in the point event table by calculating the measurement value M. The calculation steps of the measurement value M are as follows: ① Let the disaster point be A, and perform Geohash encoding on A, which is recorded as Geohash A ;② Filter out all road sets R whose road Geohash codes contain the Geohash of point A, that is: R={Road|Geohash A ∈Geojasj Road}; ③ Calculate the road R closest to point A in the set R A , that is: R A =min Road∈R {Distance A,Road}; ④ Project point A onto road R A ⑤ Calculate the projection point A′ on the road R A The measured value M on the road is also the value of point A on the road R A The measurement value M on the road is also the measurement value M of point A in the entire road network. ⑥ Assign an ID to the point disaster, and the measurement value M, road R A The ID of the event, the ID of the disaster, and the point disaster are stored together in the point event table.

[0066] If the disaster is recorded in the form of coordinates of the starting point and the end point, it is stored in the line event table. The starting point and the end point are respectively point S and point E. The linear reference algorithm can be used to obtain the coordinates of point S on the nearest road R. S The measured value FromM on the point E and the point E on the nearest road R EThe measurement value ToM on the line. It is worth noting that the starting point and the end point need to be able to form a complete path through connected roads, otherwise it will be considered an illegal disaster record, because disasters passing through non-connected roads can always be split into a set of disasters passing through connected roads. If the starting point and the end point happen to be on the same road, the line disaster corresponds to only one line event, and the measurement value of this line event is (FromM, ToM); if the starting point and the end point are not on the same road, the line disaster corresponds to multiple elements, and the road R where the starting point and the end point are located is the same. S and R E All connected roads between (S and E passing through non-connected roads will not be recorded) will be recorded as line events, with FromM=0, ToM=1. S On the road R, ToM = 1, E Finally, an ID is assigned to the line disaster and stored in the line event table along with the measurement value (FromM, ToM), the ID of the corresponding road, the ID of the disaster, and the thematic attributes of the line disaster.

[0067] Step 3.3: Extract and match damaged roads based on linear expression data.

[0068] In popular recording methods and media expressions, linear expressions are mostly used to describe disaster information, such as "road damage occurred approximately 350 meters west of a certain hotel along a certain road." In the linear expression text describing disaster information, four types of semantic information can be extracted: the name of the reference point, the name of the reference road, the reference direction, and the reference distance. Figure 6 As shown, the Levenshtein distance is used to compare the names of reference points and reference roads with those in the road network database. The points and roads with the highest similarity are selected as the actual reference points and reference roads. A linear referencing algorithm is used to convert the reference point positions and reference distances into measurements of the reference roads. Using the reference direction, the reference point and reference distance measurements are added or subtracted to obtain the measurements of point and line hazards. Finally, an ID is assigned to each point or line hazard and stored in a point event table or line event table along with the measurement value, reference road ID, hazard ID, and thematic attributes of the point or line hazard. Linearly expressed line hazards also need to form a complete path through connected roads. Line events involving multiple roads and features are stored in the same manner as coordinate data.

[0069] Step 4: Modify the weight of the road access for the damaged road section matched in step 3, and use the weighted shortest path algorithm to provide an emergency path planning solution after the disaster occurs.

[0070] The weighted shortest path planning will give the optimal path based on the minimum value of the sum of the weights. Therefore, by changing the weights of the roads corresponding to the disaster events, real-time disaster emergency path planning can be carried out according to the disaster situation. Setting the weight of the damaged road to infinity means that the damaged road is impassable. In the real-time disaster emergency path planning service, three different factors of priority path planning are provided, namely distance priority, time priority, and road level priority. For distance priority, the length of the road can be directly used as the weight. Let the length of the road be d, then the weight is: w = d. For time priority, an expected speed needs to be specified, because the time to travel the same section of road is different at different speeds, and each section of road will have a speed limit field. If the expected speed is greater than the speed limit, the speed limit is used, otherwise the expected speed is used. Therefore, the length of the road is divided by the minimum value of the expected speed and the speed limit as the weight, and the expected speed is set to v e , the speed limit is v l , then the weight is: For road priority, the weight is the product of the road's length and its priority. The higher the road's level, the smaller its priority. For example, national highways (w = 1) < provincial highways (w = 1.25) < rural highways (w = 2). Let's assume the road priority is P, and the weight is w = d × P. By defining these weights differently, running the Dijkstra algorithm yields different desired shortest path planning results.

[0071] Step 5: Map visualization of the disaster emergency path generated in step 4.

[0072] like Figure 7 As shown, a natural disaster emergency road network map service system according to a second embodiment of the present invention includes a hardware infrastructure module, a data storage module, a disaster multi-source data access module, a disaster emergency road network map service module and an access interface module.

[0073] Hardware infrastructure module: The hardware uses storage devices with large storage space and high-performance servers to ensure the timeliness of services.

[0074] Data storage module: stores existing basic geographic data and road network data into the database. In this embodiment, the basic geographic data and road network data are open source data from Open Street Map (OSM), and the database is a PostgreSQL open source relational database.

[0075] Disaster multi-source data access module: provides an upload entry for disaster multi-source data. The uploaded disaster multi-source data will be automatically processed by the system according to the metadata input by the user, generating point events and line events, and storing them in the corresponding point event table and line event table.

[0076] Disaster Emergency Road Network Map Service Module: This module provides four services: real-time emergency route planning, road network publishing / sharing, road network disaster visualization, and disaster event management. Users can plan emergency routes based on real-time disaster situations within the system and view real-time visualization of the disaster situation. Furthermore, users can publish or share local road network datasets based on their needs. Users can also manage input disaster events and update their lifecycles in real time.

[0077] Access interface module: provides two access modes, one is to call the service directly through the network API, and the other is to access it through the built Web platform.

[0078] The specific embodiments described herein are merely illustrative of the spirit of the present invention. Persons skilled in the art may make various modifications, additions, or substitutions to the described specific embodiments without departing from the spirit of the present invention or exceeding the scope of the appended claims.

Claims

1. A natural disaster emergency road network map service method, characterized in that: The following steps are involved: Step 1: Design a disaster emergency road network organization model based on dynamic segmentation and build a road network database based on the model; Step 1.1: Use the linear referencing algorithm to dynamically store and express any point on the road; The location of a point on the road is stored by its distance or percentage from the starting point of the road. The specific operation is as follows: ① Let any point on the road R be A, and the starting point of the road R be R s , the end point is R e ;②Calculate the A and R on road R s The distance is recorded as L s , or calculate A and R e The distance is recorded as L e , and calculate the length of road R, recorded as L R ;③Calculate the length of A on road R as follows: or Will or It is called the measurement value of point A on road R, and is expressed as M or express; Step 1.2: Use the linear referencing algorithm to dynamically store and express any subset of roads; Any subset R of roads R Sub Both have a starting point and an end point, denoted as point S and point E. The linear reference algorithm is used to calculate the measurement values ​​of the starting point and the end point, denoted as and Then the tuple A subset R that can uniquely and dynamically express any road R Sub ; The measured value of the starting point FromM or End point measurement ToM or Step 1.3, dynamically segmenting the roads using the dynamic storage representation of points and lines in steps 1.1 and 1.2, and constructing a disaster emergency road network model based on the dynamic segmentation; Step 2: extract disaster information from multi-source data and store it in the disaster information database; Step 3, matching the disaster information extracted in step 2 with the road network data in the road network database constructed in step 1; The source disaster data is classified into three types of data: raster data, coordinate data, and linear expression data. Different matching methods need to be selected for different types of data, namely: extraction and matching methods for damaged roads based on raster data, extraction and matching methods for damaged roads based on coordinate data, and extraction and matching methods for damaged roads based on linear expression data. There are two scenarios for extracting and matching damaged roads based on coordinate data: disaster information is recorded in the form of single-point coordinates, and disaster information is recorded in the form of start and end point coordinates; If the disaster information is recorded in the form of coordinates of a single point, the measurement value M is calculated and stored in the point event table. The calculation steps of the measurement value M are as follows: ① Let the disaster point be A, perform Geohash encoding on A, and record it as Geohash A ;② Filter out all road sets R whose road Geohash codes contain the Geohash of point A, that is: R={Road|Geohash A ∈Geohash Road }; ③ Calculate the road R closest to point A in the set R A , that is: R A =min Road∈R {Distance A,Road }; ④ Project point A onto road R A On, we get the projection point A ′ ; ⑤Calculate the projection point A ′ On the road R A The measured value M on the road is also the value of point A on the road R A The measurement value M on the road is also the measurement value M of point A in the entire road network. ⑥ Assign an ID to the point disaster, and the measurement value N, road R A The ID of the disaster, the ID of the point disaster, and the thematic attributes of the point disaster are stored in the point event table; if the disaster is recorded in the form of the coordinates of the starting point and the end point, it is stored in the line event table, and the starting point and the end point are respectively point S and point E. The starting point S and the end point E can form a complete path through the connected roads. The linear reference algorithm can be used to obtain the distance between point S and its nearest road R. S The measured value FromM on the point E and the point E on the nearest road R E If the starting point and the end point are on the same road, the line disaster corresponds to only one line event, and the measurement value of the line event is (FromM, ToM); if the starting point and the end point are not on the same road, the line disaster corresponds to multiple elements, and the measurement value of the line event is (FromM, ToM). S and R E All connected roads between them will be recorded as line events, with FromM=0,ToM=1. S On the road R, ToM = 1, E Above, FromM = 0; finally, an ID is assigned to the line disaster and stored in the line event table together with the measurement value (FromM, ToM), the ID of the corresponding road, the ID of the disaster, and the thematic attributes of the line disaster; Step 4: Modify the road traffic weights for the damaged road sections matched in step 3, and use the weighted shortest path algorithm to provide an emergency path planning solution after the disaster occurs; Step 5: Map visualization of the disaster emergency path generated in step 4.

2. The natural disaster emergency road network map service method according to claim 1, characterized in that: The disaster emergency road network model constructed in step 1.3 includes a road table, a point event table, a line event table, and a disaster table. All roads are stored in the road table, with the road ID as the primary key and the start and end coordinates. Disaster point events are stored in the point event table, with the point event ID as the primary key and the measurement value, the associated road ID, and the associated disaster ID. The road ID and disaster ID are used as foreign keys. In addition, some thematic attributes, such as the description of the point event, the occurrence time of the point event, the impact radius of the point event, and the severity level of the point event, are also stored. Disaster line events are stored in the line event table, with the line event ID as the primary key and the start and end measurement value, the associated road ID, and the associated disaster ID. The road ID and disaster ID are used as foreign keys. In addition, some thematic attributes, such as the description of the line event, the occurrence time of the line event, the impact area of ​​the line event, and the severity level of the line event, are also stored. All disasters are stored in the disaster table, with the disaster ID as the primary key and the disaster type, occurrence time, end time, and disaster description as disaster attribute information.

3. The natural disaster emergency road network map service method according to claim 1, characterized in that: The method for extracting and matching damaged roads based on raster data in step 3 is to convert the raster data of the disaster area into vector data after extracting the disaster area, and generate Geohash codes for the disaster area vector data and road network data to obtain the number of disaster area vectors with Geohash indexes. The character length of the Geohash codes for the disaster area and road network data is n. The Geohash code of a single disaster vector or a single road is the set of intersecting Geohash grids, that is: where Length Geohash =n; after Geohash encoding the disaster vector and the road vector in the database, the Geohash grid is traversed to perform an overlay analysis on the disaster vector and road vector within the grid; after extracting the damaged roads, the disaster emergency road network model proposed in step 1 is used to match them with the road network in the database. The specific matching method is as follows: ① Calculate the length of the overlaid "disaster-road" vector data. If the length is less than the resolution of the raster data, the disaster event is considered a point event; if the length is greater than the resolution of the raster data, the disaster event is considered a line event; ② Use the coordinates to calculate the measurement value M of the point event or the measurement value (FromM, ToM) of the line event; ③ Obtain the road ID and disaster ID corresponding to the point event or line event, and assign an ID to the point event or line event; ④ Store all necessary fields and event attributes in the point event table or line event table.

4. The natural disaster emergency road network map service method according to claim 1, characterized in that: The method for extracting and matching damaged roads based on linear expression data in step 3 requires extracting four types of semantic information from the text describing the disaster information in the linear expression: the name of the reference point, the name of the reference road, the reference direction, and the reference distance. The Levenshtein distance is used to compare the similarity between the names of the reference points and reference roads and the names of the points and roads in the road network database, and the points and roads with the highest similarity are selected as the actual reference points and reference roads. The linear referencing algorithm is used to convert the position of the reference point and the reference distance into the measurement value of the reference road. With the help of the reference direction, the measurement value of the reference point and the measurement value of the reference distance are added or subtracted to obtain the measurement value of the point disaster and line disaster. Finally, an ID is assigned to the point disaster or line disaster and stored in the point event table or line event table together with the measurement value, reference road ID, disaster ID, and thematic attributes of the point disaster or line disaster. Linear disasters expressed in linear expression need to be able to form a complete path through connected roads. The storage method for line events involving multiple roads and multiple features is exactly the same as that for coordinate data.

5. The natural disaster emergency road network map service method according to claim 1, characterized in that: In step 4, by changing the weight of the road corresponding to the disaster event, real-time disaster emergency path planning can be carried out according to the disaster situation. In the real-time disaster emergency path planning service, three different factors of priority path planning are provided, namely distance priority, time priority, and road level priority. For distance priority, the length of the road can be directly used as the weight. Let the length of the road be d, then the weight is: w = d; For time priority, it is necessary to specify an expected speed, and then divide the length of the road by the minimum value of the expected speed and the speed limit as the weight. Let the expected speed be v e , the speed limit is v l , then the weight is: For road level priority, the product of the road length and the road priority is used as the weight. If the road priority is P, the weight is: w = d × P; through the different definitions of the above weights, the Dijkstra algorithm can be used to obtain the desired different shortest path planning results.

Images