Flood disaster avoidance route automatic generation system

Abstract

The invention relates to flood disaster emergency management, in particular to a flood disaster avoidance route automatic generation system, which is characterized in that a server confirms an unsubmerged area in a target area through an unsubmerged area confirmation module on the basis of real-time flood submerged data, performs disaster avoidance grid division on the unsubmerged area by using an unsubmerged area division module, and generates a disaster avoidance route according to the disaster avoidance grid division; the server generates a disaster avoidance route of the target person in the disaster avoidance grid through a disaster avoidance route generation module, and dynamically updates the disaster avoidance route generated by the disaster avoidance route generation module based on the real-time flood flooding data by using a disaster avoidance route automatic updating module; the server sends the disaster avoidance route generated by the disaster avoidance route generation module and the disaster avoidance route updated by the disaster avoidance route automatic updating module to target personnel in real time through the wireless communication module; according to the technical scheme provided by the invention, the defect that the disaster avoidance route cannot be accurately, efficiently and dynamically generated during flood invasion in the prior art can be effectively overcome.

Description

Technical Field

[0001] The invention relates to flood disaster emergency management, and in particular to a flood disaster avoidance route automatic generation system. Background Art

[0002] Flood disasters refer to the flooding caused by the increase of water volume and water level in rivers, lakes and coastal areas due to melting ice and snow, ice slush, dam breaches, storm surges, etc. Flood disasters are one of the most frequent natural disasters in my country. They occur frequently and have a wide range of impacts, posing a huge threat to people's lives and property. Flood disasters occur frequently in some areas with complex geographical conditions and changeable climates.

[0003] With the continuous development of science and technology, people are increasingly relying on scientific and technological means to deal with natural disasters. In the field of flood prevention and disaster reduction, the application of scientific and technological means has achieved remarkable results. By using remote sensing technology, geographic information system (GIS), big data and other advanced technologies, early warning, real-time monitoring and evaluation of flood disasters can be achieved, providing a scientific basis for disaster prevention and reduction.

[0004] In flood prevention and disaster reduction work, an automatic flood evacuation route generation system came into being. This system uses spatial information technologies such as GIS, combined with multi-source information such as real-time meteorological data, topographic data, historical flood data, etc., through complex algorithm models, to automatically generate evacuation routes, which can help people evacuate to safe areas before the flood comes, reducing casualties and property losses.

[0005] The existing automatic generation system of flood evacuation routes can only generate evacuation routes before the flood arrives. Due to the hysteresis of information transmission, some people may not find out until the flood has already invaded. At this time, due to the uncertainty of flood inundation, the flooded area changes dramatically over time. How to accurately, efficiently and dynamically generate evacuation routes in this situation has become a technical problem that needs to be solved urgently. Summary of the invention

[0006] 1. Technical issues to be resolved

[0007] In view of the above-mentioned shortcomings of the prior art, the present invention provides a flood evacuation route automatic generation system, which can effectively overcome the defect of the prior art that it is unable to accurately, efficiently and dynamically generate evacuation routes when floods invade.

[0008] (II) Technical solution

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0010] A flood evacuation route automatic generation system includes a server, wherein the server constructs a flood evolution model of a target area through a flood evolution model construction module, and uses a flood inundation data set construction module to establish a flood inundation data set containing flood inundation data at different spatial resolutions, wherein the server processes flood inundation data in the flood inundation data set through a flood inundation data processing module, and uses a mapping model training module to perform model training on a mapping model constructed by the mapping model construction module based on the processed flood inundation data set, and wherein the server obtains real-time flood inundation data based on the trained mapping model through a real-time flood inundation data acquisition module;

[0011] The server confirms the non-flooded area in the target area based on the real-time flood inundation data through the non-flooded area confirmation module, and divides the non-flooded area into disaster avoidance grids using the non-flooded area division module. The server generates a disaster avoidance route for the target personnel in the disaster avoidance grid through the disaster avoidance route generation module, and dynamically updates the disaster avoidance route generated by the disaster avoidance route generation module based on the real-time flood inundation data using the disaster avoidance route automatic update module. The server sends the disaster avoidance route generated by the disaster avoidance route generation module and the disaster avoidance route updated by the disaster avoidance route automatic update module to the target personnel in real time through the wireless communication module.

[0012] Preferably, the flood inundation dataset establishing module establishes a flood inundation dataset containing flood inundation data at different spatial resolutions, including:

[0013] Grids with different spatial resolutions are established in the target area, and the flood evolution model is used to simulate the flood evolution process in the target area under various conditions to obtain flood inundation data at different spatial resolutions and establish a flood inundation dataset.

[0014] Preferably, the flood inundation data processing module processes the flood inundation data in the flood inundation data set, including:

[0015] S11. Use the empirical orthogonal function analysis method to perform data dimension reduction and feature extraction on the flood inundation data in the flood inundation data set;

[0016] S12, interpolating low-precision flood inundation data in high spatial resolution grids;

[0017] Among them, the high-dimensional flood inundation data are processed by empirical orthogonal function analysis to obtain a spatial function that does not change with time and a temporal function that only depends on time.

[0018] Preferably, interpolating low-precision flood inundation data in a high spatial resolution grid in S12 includes:

[0019] The nearest neighbor interpolation method is used to interpolate low-precision flood inundation data in high spatial resolution grids, so that the high-precision flood inundation data in each high spatial resolution grid is equal to the low-precision flood inundation data in the nearest low spatial resolution grid, ensuring that the two have the same spatial function and the time function forms a function pair.

[0020] Preferably, the mapping model training module performs model training on the mapping model constructed by the mapping model construction module based on the processed flood inundation data set, including:

[0021] S21, dividing a time function pair set formed by high-precision flood inundation data and its corresponding low-precision flood inundation data in the flood inundation data set into a training set, a validation set, and a test set according to a preset ratio;

[0022] S22, setting the loss function and optimizer of the mapping model, and inputting the training set into the mapping model for model training;

[0023] S23, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information;

[0024] S24. If the loss value is less than the preset threshold, the model training ends, and the current mapping model is the trained mapping model. Otherwise, the process returns to S22 and continues to use the training set for model training.

[0025] S25. Input the validation set into the trained mapping model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model;

[0026] S26. Input the test set into the tuned mapping model to evaluate the model performance.

[0027] Preferably, the real-time flood inundation data acquisition module acquires the real-time flood inundation data based on the trained mapping model, including:

[0028] S31, loading the corresponding low spatial resolution grid and moving boundary conditions in the flood evolution model, performing real-time flood evolution process simulation, and obtaining real-time low-precision flood inundation data;

[0029] S32, interpolating real-time low-precision flood inundation data in a high spatial resolution grid, and using an empirical orthogonal function analysis method to perform data dimension reduction and feature extraction to obtain spatial functions and temporal functions of the real-time low-precision flood inundation data;

[0030] S33, inputting the time function of the real-time low-precision flood inundation data into the trained mapping model to obtain the corresponding time function of the real-time high-precision flood inundation data;

[0031] S34. According to the time function of the real-time high-precision flood inundation data and the space function of the real-time low-precision flood inundation data, the real-time high-precision flood inundation data is obtained by using the inverse analysis of the empirical orthogonal function.

[0032] Preferably, the non-flooded area confirmation module confirms the non-flooded area in the target area based on the real-time flood inundation data, and the non-flooded area division module divides the non-flooded area into disaster avoidance grids, including:

[0033] The non-flooded area confirmation module determines the flooded areas within the target area based on real-time high-precision flood inundation data, and considers the remaining areas as non-flooded areas;

[0034] The non-flooded area division module selects appropriate spatial resolution grids based on the geographical environment of the non-flooded area and divides the non-flooded area into disaster avoidance grids;

[0035] Among them, the nodes of the disaster avoidance grid are passable nodes.

[0036] Preferably, the disaster avoidance route generation module generates a disaster avoidance route for target personnel in the disaster avoidance grid, including:

[0037] S41. Create an empty undirected graph G(V,E), where V represents a vertex set and E represents a collision-free edge set, and sequentially add the nodes in the disaster avoidance grid to the vertex set V in the undirected graph G(V,E);

[0038] S42. For each newly added vertex, calculate the distance between it and other neighbor vertices in the neighborhood, and use line segments to connect the newly added vertex and other neighbor vertices in the neighborhood, and add the collision-free lines to the edge set E in the undirected graph G(V,E);

[0039] S43, constructing a probabilistic route map consisting of vertices and collision-free edges that can reflect potential route information in the disaster avoidance grid;

[0040] S44. Determine the location of the target person and the safe end point in the probability route map, and use the improved A* algorithm to search for the disaster evacuation route from the target person to the safe end point.

[0041] Preferably, in S44, the positions of the target personnel and the safe end point in the probability route map are determined, and the improved A* algorithm is used to search for the disaster avoidance route from the target personnel to the safe end point, including:

[0042] S441, taking the position of the target person in the probability route map as the starting point, and taking the position of the safe end point in the probability route map as the target point;

[0043] S442, determine the target vertex in the probability roadmap as the starting point of the current search, define a safe area with other neighbor vertices in the neighborhood of the target vertex as the center and a preset safety distance as the radius, and set the screening conditions:

[0044]

[0045] in, is the safe area of ​​the i-th neighbor vertex in the neighborhood of the target vertex, is the unflooded area at time t;

[0046] If the above filtering conditions are met, and the corresponding neighbor vertex is neither in the OPEN list nor in the CLOSED list, then the neighbor vertex is stored in the OPEN list;

[0047] If the above filtering conditions are met and the corresponding neighbor vertex is already in the CLOSED list, the neighbor vertex is ignored;

[0048] If the above filtering conditions are not met, the corresponding neighbor vertices are stored in the CLOSED list;

[0049] S443, determine the evaluation function f(n) of the improved A* algorithm:

[0050] f(n)=g(n)+h(n);

[0051] Where g(n) is the actual cost from the starting point to the neighbor vertex n in the OPEN list, and h(n) is the estimated cost from the neighbor vertex n in the OPEN list to the target point calculated using the heuristic function;

[0052] S444, select the neighbor vertex with the smallest value of the evaluation function f(n) from the OPEN list. If the neighbor vertex is the target point, the algorithm ends, and the disaster avoidance route can be obtained by backtracking the parent nodes in sequence, and enter S446; otherwise, store the neighbor vertex in the CLOSED list and enter S445;

[0053] S445, taking the neighbor vertex with the smallest value of the evaluation function f(n) in the OPEN list as the target vertex of the current search starting point in the probability roadmap, and returning to S442;

[0054] S446. Optimize the disaster evacuation route to further reduce the route length and improve the route smoothness.

[0055] Preferably, the disaster avoidance route automatic updating module dynamically updates the disaster avoidance route generated by the disaster avoidance route generating module based on the real-time flood inundation data, including:

[0056] If it is within the preset time interval, the disaster avoidance route automatic updating module will dynamically analyze the non-flooded area, and fine-tune and optimize the disaster avoidance route generated by the disaster avoidance route generating module according to the dynamic analysis result, so as to dynamically update the disaster avoidance route;

[0057] If the preset time interval is exceeded, the disaster evacuation route automatic update module notifies the disaster evacuation route generation module to regenerate the disaster evacuation route. The disaster evacuation route generation module generates a new disaster evacuation route based on the location of the target personnel and the safe end point in the probability route map to dynamically update the disaster evacuation route.

[0058] (III) Beneficial effects

[0059] Compared with the prior art, the automatic generation system of flood evacuation routes provided by the present invention has the following beneficial effects:

[0060] 1) Construct a flood evolution model for the target area, establish a flood inundation data set containing flood inundation data at different spatial resolutions, perform data dimension reduction, feature extraction and interpolation processing on the flood inundation data in the flood inundation data set, use the processed flood inundation data set to train the mapping model constructed by the mapping model construction module, obtain real-time flood inundation data based on the trained mapping model, so that the obtained real-time low-precision flood inundation data can be inverted into real-time high-precision flood inundation data, accurately and efficiently simulate the real-time flood evolution process, and provide data support for the subsequent accurate, efficient and dynamic generation of disaster avoidance routes;

[0061] 2) The probabilistic roadmap algorithm is used in conjunction with the improved A* algorithm to generate disaster evacuation routes. The probabilistic roadmap algorithm can assist in rapid path planning when environmental information is complex and computing resources are limited. The use of the improved A* algorithm enables it to accurately find the optimal disaster evacuation route based on the probabilistic roadmap algorithm, thereby accurately and efficiently generating disaster evacuation routes when floods invade;

[0062] 3) In the improved A* algorithm, a screening condition is added when screening neighbor vertices, which can automatically generate disaster evacuation routes according to the dynamic changes of the flooded area, so that the generated disaster evacuation routes can maintain a certain safe distance from the flooded area, buy precious time for the evacuation of the masses, and ensure the safety of the masses during evacuation;

[0063] 4) Within the preset time interval, the disaster avoidance route automatic update module dynamically analyzes the non-flooded area, and fine-tunes and optimizes the disaster avoidance route generated by the disaster avoidance route generation module based on the dynamic analysis results; if the preset time interval is exceeded, the disaster avoidance route automatic update module notifies the disaster avoidance route generation module to regenerate the disaster avoidance route. The disaster avoidance route generation module generates a new disaster avoidance route based on the location of the target personnel and the safe end point in the probability route map, so that the disaster avoidance route can be generated efficiently and dynamically when floods invade, further ensuring the safety of people's lives. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0065] Figure 1 It is a schematic diagram of the system of the present invention;

[0066] Figure 2 A schematic diagram of the process of obtaining real-time flood inundation data in the present invention;

[0067] Figure 3 The figure is a schematic diagram of the process of dynamically generating a disaster avoidance route in the present invention. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] A flood evacuation route automatic generation system, such as Figure 1 and Figure 2 As shown, it includes a server, which constructs a flood evolution model of the target area through a flood evolution model construction module, and uses a flood inundation dataset establishment module to establish a flood inundation dataset containing flood inundation data at different spatial resolutions. The server processes the flood inundation data in the flood inundation dataset through a flood inundation data processing module, and uses a mapping model training module to train the mapping model constructed by the mapping model construction module based on the processed flood inundation dataset. The server obtains real-time flood inundation data based on the trained mapping model through a real-time flood inundation data acquisition module.

[0070] ① The flood inundation dataset establishment module establishes a flood inundation dataset containing flood inundation data at different spatial resolutions, including:

[0071] Grids with different spatial resolutions are established in the target area, and the flood evolution model is used to simulate the flood evolution process in the target area under various conditions to obtain flood inundation data at different spatial resolutions and establish a flood inundation dataset.

[0072] ② The flood inundation data processing module processes the flood inundation data in the flood inundation data set, including:

[0073] S11. Use the empirical orthogonal function analysis method to perform data dimension reduction and feature extraction on the flood inundation data in the flood inundation data set;

[0074] S12, interpolating low-precision flood inundation data in high spatial resolution grids;

[0075] Among them, the high-dimensional flood inundation data are processed by empirical orthogonal function analysis to obtain a spatial function that does not change with time and a temporal function that only depends on time.

[0076] Specifically, S12 interpolates low-precision flood inundation data in high spatial resolution grids, including:

[0077] The nearest neighbor interpolation method is used to interpolate low-precision flood inundation data in high spatial resolution grids, so that the high-precision flood inundation data in each high spatial resolution grid is equal to the low-precision flood inundation data in the nearest low spatial resolution grid, ensuring that the two have the same spatial function and the time function forms a function pair.

[0078] ③ The mapping model training module performs model training on the mapping model constructed by the mapping model construction module based on the processed flood inundation dataset, including:

[0079] S21, dividing a time function pair set formed by high-precision flood inundation data and its corresponding low-precision flood inundation data in the flood inundation data set into a training set, a validation set, and a test set according to a preset ratio;

[0080] S22, setting the loss function and optimizer of the mapping model, and inputting the training set into the mapping model for model training;

[0081] S23, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information;

[0082] S24. If the loss value is less than the preset threshold, the model training ends, and the current mapping model is the trained mapping model. Otherwise, the process returns to S22 and continues to use the training set for model training.

[0083] S25. Input the validation set into the trained mapping model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model;

[0084] S26. Input the test set into the tuned mapping model to evaluate the model performance.

[0085] ④The real-time flood inundation data acquisition module acquires real-time flood inundation data based on the trained mapping model, including:

[0086] S31, loading the corresponding low spatial resolution grid and moving boundary conditions in the flood evolution model, performing real-time flood evolution process simulation, and obtaining real-time low-precision flood inundation data;

[0087] S32, interpolating real-time low-precision flood inundation data in a high spatial resolution grid, and using an empirical orthogonal function analysis method to perform data dimension reduction and feature extraction to obtain spatial functions and temporal functions of the real-time low-precision flood inundation data;

[0088] S33, inputting the time function of the real-time low-precision flood inundation data into the trained mapping model to obtain the corresponding time function of the real-time high-precision flood inundation data;

[0089] S34. According to the time function of the real-time high-precision flood inundation data and the space function of the real-time low-precision flood inundation data, the real-time high-precision flood inundation data is obtained by using the inverse analysis of the empirical orthogonal function.

[0090] The above technical scheme constructs a flood evolution model for the target area, establishes a flood inundation data set containing flood inundation data at different spatial resolutions, performs data dimension reduction, feature extraction and interpolation processing on the flood inundation data in the flood inundation data set, uses the processed flood inundation data set to train the mapping model constructed by the mapping model construction module, and obtains real-time flood inundation data based on the trained mapping model, so that the obtained real-time low-precision flood inundation data can be inverted into real-time high-precision flood inundation data, accurately and efficiently simulates the real-time flood evolution process, and provides data support for the subsequent accurate, efficient and dynamic generation of disaster avoidance routes.

[0091] like Figure 1 and Figure 3As shown, the server confirms the non-flooded area in the target area based on the real-time flood inundation data through the non-flooded area confirmation module, and divides the non-flooded area into disaster avoidance grids using the non-flooded area division module. The server generates a disaster avoidance route for the target personnel in the disaster avoidance grid through the disaster avoidance route generation module, and dynamically updates the disaster avoidance route generated by the disaster avoidance route generation module based on the real-time flood inundation data using the disaster avoidance route automatic update module. The server sends the disaster avoidance route generated by the disaster avoidance route generation module and the disaster avoidance route updated by the disaster avoidance route automatic update module to the target personnel in real time through the wireless communication module.

[0092] ① The non-flooded area confirmation module confirms the non-flooded area in the target area based on the real-time flood inundation data, and the non-flooded area division module divides the non-flooded area into disaster shelter grids, including:

[0093] The non-flooded area confirmation module determines the flooded areas within the target area based on real-time high-precision flood inundation data, and considers the remaining areas as non-flooded areas;

[0094] The non-flooded area division module selects appropriate spatial resolution grids based on the geographical environment of the non-flooded area and divides the non-flooded area into disaster avoidance grids;

[0095] Among them, the nodes of the disaster avoidance grid are passable nodes.

[0096] ② The disaster avoidance route generation module generates the disaster avoidance route for the target personnel in the disaster avoidance grid, including:

[0097] S41. Create an empty undirected graph G(V,E), where V represents a vertex set and E represents a collision-free edge set, and sequentially add the nodes in the disaster avoidance grid to the vertex set V in the undirected graph G(V,E);

[0098] S42. For each newly added vertex, calculate the distance between it and other neighbor vertices in the neighborhood, and use line segments to connect the newly added vertex and other neighbor vertices in the neighborhood, and add the collision-free lines to the edge set E in the undirected graph G(V,E);

[0099] S43, constructing a probabilistic route map consisting of vertices and collision-free edges that can reflect potential route information in the disaster avoidance grid;

[0100] S44. Determine the location of the target person and the safe end point in the probability route map, and use the improved A* algorithm to search for the disaster evacuation route from the target person to the safe end point.

[0101] Specifically, in S44, the positions of the target personnel and the safe end point in the probability route map are determined, and the improved A* algorithm is used to search for the disaster avoidance route from the target personnel to the safe end point, including:

[0102] S441, taking the position of the target person in the probability route map as the starting point, and taking the position of the safe end point in the probability route map as the target point;

[0103] S442, determine the target vertex in the probability roadmap as the starting point of the current search, define a safe area with other neighbor vertices in the neighborhood of the target vertex as the center and a preset safety distance as the radius, and set the screening conditions:

[0104]

[0105] in, is the safe area of ​​the i-th neighbor vertex in the neighborhood of the target vertex, is the unflooded area at time t;

[0106] If the above filtering conditions are met, and the corresponding neighbor vertex is neither in the OPEN list nor in the CLOSED list, then the neighbor vertex is stored in the OPEN list;

[0107] If the above filtering conditions are met and the corresponding neighbor vertex is already in the CLOSED list, the neighbor vertex is ignored;

[0108] If the above filtering conditions are not met, the corresponding neighbor vertices are stored in the CLOSED list;

[0109] S443, determine the evaluation function f(n) of the improved A* algorithm:

[0110] f(n)=g(n)+h(n);

[0111] Where g(n) is the actual cost from the starting point to the neighbor vertex n in the OPEN list, and h(n) is the estimated cost from the neighbor vertex n in the OPEN list to the target point calculated using the heuristic function;

[0112] S444, select the neighbor vertex with the smallest value of the evaluation function f(n) from the OPEN list. If the neighbor vertex is the target point, the algorithm ends, and the disaster avoidance route can be obtained by backtracking the parent nodes in sequence, and enter S446; otherwise, store the neighbor vertex in the CLOSED list and enter S445;

[0113] S445, taking the neighbor vertex with the smallest value of the evaluation function f(n) in the OPEN list as the target vertex of the current search starting point in the probability roadmap, and returning to S442;

[0114] S446. Optimize the disaster evacuation route to further reduce the route length and improve the route smoothness.

[0115] The above technical solution adopts a probabilistic roadmap algorithm in combination with an improved A* algorithm to generate disaster avoidance routes. The probabilistic roadmap algorithm can assist in rapid path planning when environmental information is complex and computing resources are limited. The use of the improved A* algorithm enables it to accurately find the optimal disaster avoidance route based on the probabilistic roadmap algorithm, thereby being able to accurately and efficiently generate disaster avoidance routes when floods invade.

[0116] In addition, the improved A* algorithm adds screening conditions when screening neighbor vertices, which can automatically generate disaster evacuation routes based on the dynamic changes of the flooded areas, so that the generated disaster evacuation routes can maintain a certain safe distance from the flooded areas, buying precious time for the evacuation of the people and ensuring their safety during the evacuation.

[0117] ③ The disaster evacuation route automatic update module dynamically updates the disaster evacuation route generated by the disaster evacuation route generation module based on real-time flood inundation data, including:

[0118] If it is within the preset time interval, the disaster avoidance route automatic updating module will dynamically analyze the non-flooded area, and fine-tune and optimize the disaster avoidance route generated by the disaster avoidance route generating module according to the dynamic analysis result, so as to dynamically update the disaster avoidance route;

[0119] If the preset time interval is exceeded, the disaster evacuation route automatic update module notifies the disaster evacuation route generation module to regenerate the disaster evacuation route. The disaster evacuation route generation module generates a new disaster evacuation route based on the location of the target personnel and the safe end point in the probability route map to dynamically update the disaster evacuation route.

[0120] In the above technical solution, within a preset time interval, the disaster escape route automatic update module dynamically analyzes the non-flooded area, and fine-tunes and optimizes the disaster escape route generated by the disaster escape route generation module based on the dynamic analysis results; if the preset time interval is exceeded, the disaster escape route automatic update module notifies the disaster escape route generation module to regenerate the disaster escape route, and the disaster escape route generation module generates a new disaster escape route based on the location of the target personnel and the safe end point in the probabilistic route map, thereby being able to efficiently and dynamically generate disaster escape routes when floods invade, further ensuring the safety of people's lives.

[0121] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A flood evacuation route automatic generation system, characterized by: The server comprises a server, wherein the server constructs a flood evolution model of a target area through a flood evolution model construction module, and constructs a flood inundation data set containing flood inundation data at different spatial resolutions through a flood inundation data set construction module, wherein the server processes the flood inundation data in the flood inundation data set through a flood inundation data processing module, and performs model training on a mapping model constructed by the mapping model construction module based on the processed flood inundation data set through a mapping model training module, and wherein the server acquires real-time flood inundation data based on the trained mapping model through a real-time flood inundation data acquisition module; The server confirms the non-flooded area in the target area based on the real-time flood inundation data through the non-flooded area confirmation module, and divides the non-flooded area into disaster avoidance grids using the non-flooded area division module. The server generates a disaster avoidance route for the target personnel in the disaster avoidance grid through the disaster avoidance route generation module, and dynamically updates the disaster avoidance route generated by the disaster avoidance route generation module based on the real-time flood inundation data using the disaster avoidance route automatic update module. The server sends the disaster avoidance route generated by the disaster avoidance route generation module and the disaster avoidance route updated by the disaster avoidance route automatic update module to the target personnel in real time through the wireless communication module.

2. The automatic generation system of flood evacuation routes according to claim 1, characterized in that: The flood inundation dataset establishment module establishes a flood inundation dataset containing flood inundation data at different spatial resolutions, including: Grids with different spatial resolutions are established in the target area, and the flood evolution model is used to simulate the flood evolution process in the target area under various conditions to obtain flood inundation data at different spatial resolutions and establish a flood inundation dataset.

3. The automatic generation system of flood evacuation routes according to claim 2, characterized in that: The flood inundation data processing module processes the flood inundation data in the flood inundation data set, including: S11. Use the empirical orthogonal function analysis method to perform data dimension reduction and feature extraction on the flood inundation data in the flood inundation data set; S12, interpolating low-precision flood inundation data in high spatial resolution grids; Among them, the high-dimensional flood inundation data are processed by empirical orthogonal function analysis to obtain a spatial function that does not change with time and a temporal function that only depends on time.

4. The automatic generation system of flood evacuation routes according to claim 3 is characterized by: S12 interpolates low-precision flood inundation data in a high spatial resolution grid, including: The nearest neighbor interpolation method is used to interpolate low-precision flood inundation data in high spatial resolution grids, so that the high-precision flood inundation data in each high spatial resolution grid is equal to the low-precision flood inundation data in the nearest low spatial resolution grid, ensuring that the two have the same spatial function and the time function forms a function pair.

5. The automatic generation system of flood evacuation routes according to claim 4 is characterized by: The mapping model training module performs model training on the mapping model constructed by the mapping model construction module based on the processed flood inundation data set, including: S21, dividing a time function pair set formed by high-precision flood inundation data and its corresponding low-precision flood inundation data in the flood inundation data set into a training set, a validation set, and a test set according to a preset ratio; S22, setting the loss function and optimizer of the mapping model, and inputting the training set into the mapping model for model training; S23, calculating the loss value based on the loss function, and the optimizer updates the model parameters according to the loss value and the network gradient information; S24. If the loss value is less than the preset threshold, the model training ends, and the current mapping model is the trained mapping model. Otherwise, the process returns to S22 and continues to use the training set for model training. S25. Input the validation set into the trained mapping model, evaluate the generalization ability of the model by observing the performance on the validation set, and tune the hyperparameters and structure of the model; S26. Input the test set into the tuned mapping model to evaluate the model performance.

6. The automatic generation system of flood evacuation routes according to claim 5, characterized in that: The real-time flood inundation data acquisition module acquires real-time flood inundation data based on the trained mapping model, including: S31, loading the corresponding low spatial resolution grid and moving boundary conditions in the flood evolution model, performing real-time flood evolution process simulation, and obtaining real-time low-precision flood inundation data; S32, interpolating real-time low-precision flood inundation data in a high spatial resolution grid, and using an empirical orthogonal function analysis method to perform data dimension reduction and feature extraction to obtain spatial functions and temporal functions of the real-time low-precision flood inundation data; S33, inputting the time function of the real-time low-precision flood inundation data into the trained mapping model to obtain the corresponding time function of the real-time high-precision flood inundation data; S34. According to the time function of the real-time high-precision flood inundation data and the space function of the real-time low-precision flood inundation data, the real-time high-precision flood inundation data is obtained by using the inverse analysis of the empirical orthogonal function.

7. The automatic generation system of flood evacuation routes according to claim 1, characterized in that: The non-flooded area confirmation module confirms the non-flooded area in the target area based on the real-time flood inundation data, and the non-flooded area division module divides the non-flooded area into disaster avoidance grids, including: The non-flooded area confirmation module determines the flooded areas within the target area based on real-time high-precision flood inundation data, and considers the remaining areas as non-flooded areas; The non-flooded area division module selects appropriate spatial resolution grids based on the geographical environment of the non-flooded area and divides the non-flooded area into disaster avoidance grids; Among them, the nodes of the disaster avoidance grid are passable nodes.

8. The automatic generation system of flood evacuation routes according to claim 7, characterized in that: The disaster avoidance route generation module generates a disaster avoidance route for target personnel in the disaster avoidance grid, including: S41. Create an empty undirected graph G(V,E), where V represents a vertex set and E represents a collision-free edge set, and sequentially add the nodes in the disaster avoidance grid to the vertex set V in the undirected graph G(V,E); S42. For each newly added vertex, calculate the distance between it and other neighbor vertices in the neighborhood, and use line segments to connect the newly added vertex and other neighbor vertices in the neighborhood, and add the collision-free lines to the edge set E in the undirected graph G(V,E); S43, constructing a probabilistic route map consisting of vertices and collision-free edges that can reflect potential route information in the disaster avoidance grid; S44. Determine the location of the target person and the safe end point in the probability route map, and use the improved A* algorithm to search for the disaster evacuation route from the target person to the safe end point.

9. The automatic generation system of flood evacuation routes according to claim 8, characterized in that: In S44, the positions of the target personnel and the safe end point in the probability route map are determined, and the improved A* algorithm is used to search for the disaster evacuation route from the target personnel to the safe end point, including: S441, taking the position of the target person in the probability route map as the starting point, and taking the position of the safe end point in the probability route map as the target point; S442, determine the target vertex in the probability roadmap as the starting point of the current search, define a safe area with other neighbor vertices in the neighborhood of the target vertex as the center and a preset safety distance as the radius, and set the screening conditions: in, is the safe area of ​​the i-th neighbor vertex in the neighborhood of the target vertex, is the unflooded area at time t; If the above filtering conditions are met, and the corresponding neighbor vertex is neither in the OPEN list nor in the CLOSED list, then the neighbor vertex is stored in the OPEN list; If the above filtering conditions are met and the corresponding neighbor vertex is already in the CLOSED list, the neighbor vertex is ignored; If the above filtering conditions are not met, the corresponding neighbor vertices are stored in the CLOSED list; S443, determine the evaluation function f(n) of the improved A* algorithm: f(n)=g(n)+h(n); Where g(n) is the actual cost from the starting point to the neighbor vertex n in the OPEN list, and h(n) is the estimated cost from the neighbor vertex n in the OPEN list to the target point calculated using the heuristic function; S444, select the neighbor vertex with the smallest value of the evaluation function f(n) from the OPEN list. If the neighbor vertex is the target point, the algorithm ends, and the disaster avoidance route can be obtained by backtracking the parent nodes in sequence, and enter S446; otherwise, store the neighbor vertex in the CLOSED list and enter S445; S445, taking the neighbor vertex with the smallest value of the evaluation function f(n) in the OPEN list as the target vertex of the current search starting point in the probability roadmap, and returning to S442; S446. Optimize the disaster evacuation route to further reduce the route length and improve the route smoothness.

10. The automatic generation system of flood evacuation routes according to claim 9, characterized in that: The disaster avoidance route automatic updating module dynamically updates the disaster avoidance route generated by the disaster avoidance route generating module based on the real-time flood inundation data, including: If it is within the preset time interval, the disaster avoidance route automatic updating module will dynamically analyze the non-flooded area, and fine-tune and optimize the disaster avoidance route generated by the disaster avoidance route generating module according to the dynamic analysis result, so as to dynamically update the disaster avoidance route; If the preset time interval is exceeded, the disaster evacuation route automatic update module notifies the disaster evacuation route generation module to regenerate the disaster evacuation route. The disaster evacuation route generation module generates a new disaster evacuation route based on the location of the target personnel and the safe end point in the probability route map to dynamically update the disaster evacuation route.

Images