Urban waterlogging resident emergency evacuation path guiding method for smart community

Abstract

The invention discloses an urban waterlogging resident emergency evacuation path guiding method for a smart community, and relates to the technical field of smart cities and emergency management. The method comprises the following steps: firstly, constructing a community digital twinborn base fusing geographic information, building information, Internet of Things sensing data and a hydrodynamic model; dynamic inversion and prediction of the waterlogging situation are realized by assimilating a multi-source observation data driving model; calculating the space-time dynamic risk of the evacuation road network based on the prediction result, and generating a risk map; calculating a space-time evacuation scheme including optimal departure time and a specific path sequence for residents or families by combining resident personalized constraints and adopting a double-layer optimization model combining macroscopic flow distribution and microscopic path planning; finally, instructions are issued through multiple channels such as an intelligent terminal and broadcast, and dynamic adjustment is performed based on real-time monitoring, so that accurate perception, prospective prediction, personalized planning and collaborative guidance of waterlogging evacuation are realized, and the overall efficiency of community emergency evacuation is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart city, emergency management and Internet of Things, and particularly relates to a city waterlogging resident emergency evacuation path guiding method for a smart community. BACKGROUND

[0002] In recent years, extreme rainfall events occur frequently, and urban waterlogging seriously threatens the life and property safety of residents in low-lying communities. The traditional emergency evacuation scheme is usually based on static emergency plans and limited real-time information, and has obvious deficiencies in the rapidly evolving urban waterlogging disasters.

[0003] The prior art mainly has the following defects: Dependence on meteorological warnings and a small number of water level monitoring points makes it difficult to accurately perceive the dynamic evolution process of the community internal water depth and flow rate in real time, and the availability of the evacuation road network is not accurately judged; Usually, a few fixed evacuation routes are planned based on historical data or preset scenarios, or simple avoidance is performed according to real-time water accumulation point information, without considering the dynamics of the waterlogging process (water accumulation rising, peak, subsiding) and the differentiated needs of different groups of people (such as the elderly, children, disabled people, and vehicles) for paths; Most of them are passive response guiding, that is, indicating the "current" optimal path after the disaster occurs, and cannot predict the submergence risk of the road network in the future (such as the next 30-60 minutes) based on a water dynamics model to conduct forward-looking and preventive evacuation guidance; Lack of macro control of overall evacuation demand, which can easily lead to a large number of residents rushing to a few "safe" exits or paths, forming new congestion points or crowd gathering risks, and reducing the overall evacuation efficiency.

[0004] Therefore, an intelligent guiding method is urgently needed that can integrate multi-source real-time data, dynamically predict waterlogging evolution, and provide personalized, executable, and dynamically adjustable evacuation paths for different groups of residents. SUMMARY

[0005] The purpose of the present application is to overcome the deficiencies of the prior art and provide a city waterlogging resident emergency evacuation path guiding method that can realize accurate perception, dynamic prediction, personalized decision-making, and overall coordination.

[0006] To achieve the above purpose, the present application adopts the following technical solutions: A city waterlogging resident emergency evacuation path guiding method for a smart community, comprising the following steps: Constructing a community multi-layer fusion digital twin basement containing geographic information, building information, real-time perception data, and a water dynamics model; Real-time multi-source observation data is aggregated and fused, and the water dynamic model is driven by data assimilation technology to dynamically invert and predict the three-dimensional situation of community waterlogging; Based on the real-time and predicted data of the waterlogging situation, the dynamic traffic risk of each road segment in the evacuation path network in multiple future time windows is calculated, and a spatio-temporal dynamic risk map is generated; The individual attributes and constraint conditions of the residents to be evacuated are identified, and a personalized evacuation demand model is constructed with the goal of the shortest safe arrival time; Based on the dynamic risk map, a double-layer optimization model combining macroscopic evacuation flow distribution and microscopic personalized path planning is used to calculate the spatio-temporal evacuation scheme for residents or family groups, including recommended departure time and specific path sequence, to meet their needs; The spatio-temporal evacuation scheme is pushed to the target residents, and dynamic evaluation and scheme adjustment are carried out based on real-time monitoring and prediction deviation during the evacuation process.

[0007] Preferably, the construction of the community multi-layer fusion digital twin basement specifically includes: integrating high-precision geographic information system data, building information model data of main buildings, Internet of Things sensing layer data composed of water level meters, flow meters and video monitoring deployed in the community, and surface and pipe network coupled water dynamic model data calibrated by local parameters, forming a computable digital twin platform in a unified spatio-temporal coordinate system.

[0008] Preferably, the multi-source observation data at least includes: meteorological radar and grid rainfall forecast data, community Internet of Things sensor data, anonymous crowdsourcing mobile speed and location data from mobile terminals, and text or image waterlogging report data extracted from social media; The data assimilation technology uses Kalman filtering or ensemble Kalman filtering method to fuse observation data and model simulation results to continuously correct the model state and improve the accuracy of situation awareness.

[0009] Preferably, the calculation formula of the dynamic traffic risk R(t) is: R(t) = α * Rs + β * Rh(t) + γ * Ru(t), where Rs is a static infrastructure risk factor, Rh(t) is a dynamic hydraulic risk factor calculated based on the predicted water depth and flow rate at time t, Ru(t) is a prediction risk factor calculated based on model prediction uncertainty, and α, β, γ are configurable weight coefficients.

[0010] Preferably, the specific steps of generating the spatio-temporal dynamic risk map include: Discretize all potential evacuation channels in the community into road segment units; For each road segment unit, predict its water depth and flow rate at each future time step based on the water dynamic model; In combination with the static attribute of the road section, the risk value of each time step is calculated according to the calculation formula of the dynamic passing risk R(t); and the change of the risk value with time is visualized and presented in the form of a heat map on the digital twin basement.

[0011] Preferably, the individual attributes and constraint conditions at least include: movement mode, movement speed, whether assistance is needed, accompanying personnel, available vehicle type, and family gathering preference; and in the shortest safe arrival time target, "safety" is defined as the dynamic passing risk value of each road section on the planned path at the resident passing time, which is lower than the preset risk tolerance threshold of the resident corresponding to the crowd-vehicle combination.

[0012] Preferably, the macroscopic evacuation flow distribution model takes each building or community in the community as a source and each emergency shelter as a sink, constructs a time-varying impedance network based on the dynamic risk map, adopts the dynamic user equilibrium or system optimization principle, and solves the flow distribution scheme and the time period departure rate of each source for minimizing the overall evacuation time; and the microscopic individualized path planning, under the framework of the macroscopic flow distribution scheme, for each resident or family group, in the extended space-time network, adopts the time-dependent shortest path algorithm or space-time A* search algorithm to solve the specific path sequence and departure time from the starting point to the shelter that meet the individual constraints.

[0013] Preferably, the way of pushing the space-time evacuation scheme includes one or more combinations of sending graphical navigation information to residents through a smart phone application, playing voice guidance through community building broadcasting, publishing text prompts through community variable information signboards, and air shouting and guiding through a drone.

[0014] Preferably, the trigger conditions for dynamic evaluation and scheme adjustment include any one or more of the following: receiving a higher level of rainstorm warning signal, the deviation of the monitoring data and the predicted value of the key position sensor exceeding the preset threshold, significant adverse changes in the dynamic risk map, and the deviation of the actual monitored evacuation flow / vehicle flow density and the planned prediction value exceeding the preset threshold.

[0015] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of the preceding embodiments when executing the program.

[0016] Compared with the prior art, the present application has the following remarkable beneficial effects: through constructing a multi-layer fusion digital twin basement and using multi-source data assimilation technology, real-time accurate perception and high-confidence prediction of the waterlogging situation in the community are realized; then, based on the spatiotemporal dynamic risk map, a double-layer optimization model combining macro-flow distribution and micro-individual path planning is used to generate differentiated and forward-looking evacuation plans for different groups of residents, taking into account the overall evacuation efficiency and individual safety constraints; finally, through multi-channel coordinated release and closed-loop dynamic adjustment mechanism based on real-time monitoring, the reliable delivery of evacuation instructions and the agile adaptation to disaster evolution are ensured, thereby systematically solving the problems of situation perception lag, static and homogeneous path planning, lack of prediction ability and easy to cause evacuation conflicts in traditional methods, and significantly improving the safety, efficiency and fairness of emergency evacuation under urban waterlogging disasters.

[0017] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0019] Figure 1 Flow chart of the method for guiding the emergency evacuation path of urban waterlogging residents in a smart community according to the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] EMBODIMENT REFERENCE Figure 1This is a flowchart of the method for guiding emergency evacuation routes for urban residents in smart communities during urban flooding, as described in this invention. The hardware foundation of this invention includes: a network of water level, rainfall, and video sensors deployed at key locations in the community; a community data center or cloud platform; residents' smartphone terminals; a large screen in the community emergency command center; and optional drones, variable message signs, etc. The software system includes: a data fusion and assimilation module, a hydraulic simulation and prediction module, an evacuation route optimization calculation module, and an instruction generation and issuance module.

[0022] Step S1: Construct a multi-layered, integrated digital twin foundation for the community, including geographic information, building information, real-time sensing data, and hydrodynamic models. Specifically, this includes: integrating high-precision geographic information system data, building information model data of major buildings, IoT sensing layer data consisting of water level gauges, flow meters, and video surveillance deployed within the community, and surface and pipeline network coupled hydrodynamic model data calibrated with local parameters, forming a computable digital twin platform under a unified spatiotemporal coordinate system.

[0023] Taking a smart community as an example, firstly, a 1:500 high-precision GIS map of the community is acquired, containing the precise location and elevation information of all roads, green spaces, squares, waterways, and underground pipe networks (drainage pipes, cable trenches). Secondly, BIM models of major residential buildings and public buildings are imported to clarify the precise three-dimensional coordinates and elevations of building entrances and underground parking garage entrances. Then, deployed IoT sensors are marked on the GIS map (e.g., ultrasonic water level gauges are installed at 10 historically flood-prone points, flow meters are installed at drainage pumping stations, and 50 community surveillance videos are connected). Finally, based on the SWMM model, a coupled hydraulic model of surface runoff and pipe network drainage for the community is constructed using GIS topographic data and pipe network data. The model parameters (such as roughness and infiltration rate) are calibrated and verified using historical rainfall and water accumulation data. These four layers of data are correlated and fused in a unified spatial coordinate system to form a digital twin foundation that can be simulated, calculated, and visualized.

[0024] Step S2: Real-time aggregation and fusion of multi-source observation data, driving the hydrodynamic model through data assimilation technology to dynamically invert and predict the three-dimensional situation of community flooding; the multi-source observation data includes at least: meteorological radar and gridded rainfall forecast data, community IoT sensor data, anonymous crowdsourced mobile speed and location data from mobile terminals, and text or image flooding report data extracted from social media; the data assimilation technology uses Kalman filtering or ensemble Kalman filtering to fuse the observation data with the model simulation results in order to continuously correct the model state and improve the accuracy of situational awareness.

[0025] When the meteorological department issues a rainstorm warning, the system is activated, and the real-time data stream includes: Refined gridded rainfall forecasts and radar echo data from the Municipal Meteorological Bureau; The community water level gauge reports water level data every minute; With the resident's authorization, anonymous mobile phone signaling or APP backend reporting of movement speed data (such as the average movement speed on a certain road dropping from the normal 20km / h to 5km / h, which may indicate water accumulation or congestion); The web crawler extracts information containing keywords such as "flooded" and "flooded" along with location data from community WeChat groups and Weibo. The data fusion center compares these observational data (considered as "real" observations) with the simulation results of the hydraulic model; it uses an ensemble Kalman filter to assimilate the observational data into the model; specifically, it runs multiple model replicas (ensembles) with subtle parameter perturbations, compares the simulated values ​​of each replica at the observation points with the actual observation values, calculates the error covariance, and then adjusts the state variables (such as water depth at various locations) of all model replicas in reverse to make the average state of the model ensemble closer to the actual situation; after assimilation, the model can more accurately reflect the real-time water depth and flow velocity of each road and each area within the current community, and presents it in a three-dimensional visualization on the command screen.

[0026] Step S3: Based on the real-time and predicted data of the urban flooding situation, calculate the dynamic traffic risk of each road segment in the evacuation route network in multiple future time windows, and generate a spatiotemporal dynamic risk map; Based on the current assimilated model state, water accumulation prediction simulations are performed 30 minutes, 60 minutes, and 90 minutes in advance. All roads in the community that can be evacuated (including driveways, sidewalks, plaza passages, etc.) are divided into several road segments. For each road segment, the dynamic passage risk value R(t) for each future time step (e.g., 5 minutes per step) is calculated.

[0027] For example, consider an east-west branch road: Static risk Rs: Considering its road grade (sub-road), lowest point elevation (lower), and whether there is a drainage outlet (yes), a basic risk value of 0.3 (range 0-1) is assigned. Dynamic hydraulic risk Rh(t): The model predicts that the water depth at minute t (e.g., t=20) will be 25cm and the flow velocity will be 0.3m / s. According to the preset depth-risk comparison table (e.g., 15-30cm is high risk), Rh(20) is calculated to be 0.8; Prediction uncertainty Ru(t): Due to the uncertainty in rainfall forecast, the predicted water depth at this point in the ensemble forecast at the 20th minute has a large dispersion between 20-30cm, and Ru(20) is calculated to be 0.2; Let the weights be α=0.2, β=0.6, γ=0.2, then R(20) = 0.2*0.3 + 0.6*0.8 + 0.2*0.2 =0.58.

[0028] This calculation is performed for all future time steps of all road segments to generate a dynamic risk heat map that changes over time; road segments with a risk value higher than 0.7 (threshold adjustable) will be marked as "about to be impassable" or "high risk".

[0029] Step S4: Identify the individual attributes and constraints of the residents to be evacuated, and construct a personalized evacuation demand model with the shortest safe arrival time as the objective; the specific steps for generating the spatiotemporal dynamic risk map include: Discretize all potential evacuation routes within the community into road segment units; For each road segment unit, the water depth and flow velocity at each future time step are predicted based on the hydrodynamic model. Based on the static attributes of the road segment, the risk value at each time step is calculated according to the formula for calculating the dynamic traffic risk R(t); the change of the risk value over time is visualized on the digital twin substrate in the form of a heat map.

[0030] The individual attributes and constraints include at least: movement mode, movement speed, whether assistance is required, accompanying persons, available vehicle types, and family grouping preferences; in the shortest safe arrival time target, "safety" is defined as the dynamic traffic risk value of each road segment on the planned route at the time of resident passage, which is lower than the preset risk tolerance threshold of the corresponding population-vehicle combination.

[0031] The system allows residents to pre-enter information via a community app or, in emergencies, fill in their own information, for example: Resident A: Lives on the 5th floor of Unit 1, Building 3, in good health, walking speed 1.2m / s, selected "walking evacuation"; Resident B (family): lives on the first floor of Unit 2, Building 5 (elderly), has a wheelchair, moves at a speed of 0.5 m / s, selects "requires assistance, gathers with son C who works near Building 8, and then evacuates by car"; Resident D: There is a car in the underground parking garage. Select "Evacuate by car".

[0032] The system establishes a demand model for each resident / family group, with the objective function being "shortest safe arrival time". However, the constraints are different. For pedestrians, "safety" means that there cannot be any sections of the path with water depth exceeding 15cm. For drivers, there cannot be any sections of the path with water depth exceeding 30cm. For wheelchair users, not only is the water depth required to be extremely shallow, but the path is also required to have a gentle slope and no steps.

[0033] Step S5: Based on the dynamic risk map, a two-layer optimization model combining macro-level evacuation flow allocation and micro-level personalized path planning is adopted to calculate a spatiotemporal evacuation plan that meets the needs of residents or family groups, including suggested departure times and specific path sequences. The macro-level evacuation flow allocation model uses each building or community within the community as the source and each emergency shelter as the sink. Based on the dynamic risk map, a time-varying impedance network is constructed. The dynamic user equilibrium or system optimality principle is adopted to solve for the flow allocation plan that minimizes the overall evacuation time and the departure rate of each source in different time periods. The micro-level personalized path planning, within the framework of the macro-level flow allocation plan, uses a time-dependent shortest path algorithm or a spatiotemporal A* search algorithm in the extended spatiotemporal network for each resident or family group to solve for the specific path sequence and departure time from the starting point to the shelter that meets their individual constraints. The methods for pushing the spatiotemporal evacuation plan include one or more combinations of sending graphical navigation information to residents through a smartphone application, playing voice guidance through community building broadcasts, publishing text prompts through community variable information signs, and using drones for aerial announcements and guidance.

[0034] First, macro-level traffic allocation is performed. Assuming the community has 1000 households distributed across 20 buildings (sources) and 3 designated emergency shelters (consolidations), based on the current and future dynamic risk map, the shortest path (considering risk-adjusted time costs) from each building to each shelter is calculated. The goal is to minimize the overall evacuation time for everyone while avoiding all vehicles converging on the same exit. The system might calculate that residents of buildings 1-10 should primarily travel to the North Shelter in the first 20 minutes; residents of buildings 11-20 should primarily travel to the East Shelter in 20-40 minutes; and drivers should depart 10 minutes later than pedestrians to reduce the risk of mixed pedestrian and vehicle traffic. This forms a macro-level scheduling scheme. The macro-level optimization results are shown in Table 1. Table 1

[0035] Then, micro-path planning is carried out. Under the guidance of the macro plan, a specific route is planned for resident A: starting from the entrance of Building 3 → turn right through the stone slab road in the central garden (predicted to be free of water accumulation in the next 30 minutes) → pass through the west gate of the community (pedestrian passage) → arrive at the North Shelter. The estimated walking time is 12 minutes. It is recommended to start immediately. For resident family B / C: Son C walks from Building 8 to Building 5 to meet his parents (planning a low-risk walking route), with an estimated assembly time of 8 minutes. Then the whole family drives from the underground garage of Building 5 → detours around the main road outside the community (although it is a bit far, the risk of flooding is predicted to be low and is dissipating) → drives out from the south gate → arrives at the East Shelter, with an estimated driving time of 15 minutes. It is recommended that C set off to assemble now, and the whole family drive off in 10 minutes. The system uses the spatiotemporal A* algorithm for searching. This algorithm unfolds the road network in the time dimension, with each node having a state of (location, time). When searching for a path from the starting point (home) to the ending point (shelter), it considers not only spatial distance but also the risk value of each road segment at that moment. The algorithm will proactively choose to "pass through a road segment when the risk is low," even if it requires waiting for a moment or taking a detour. The results of the micro-path planning are shown in Table 2.

[0036] Table 2

[0037] Step S6: Push the spatiotemporal evacuation plan to the target residents, and dynamically evaluate and adjust the plan based on real-time monitoring and prediction deviations during the evacuation process; the triggering conditions for the dynamic evaluation and plan adjustment include any one or more of the following: receiving a higher-level rainstorm warning signal, the deviation between the monitoring data and the predicted value of key location sensors exceeding a preset threshold, a major adverse change in the dynamic risk map, and the deviation between the actual monitored evacuation flow / vehicle flow density and the planned predicted value exceeding a preset threshold.

[0038] Resident A's mobile app immediately received a push notification: a clear evacuation map highlighting their walking route, with the text message: "Dear resident, please immediately walk along the highlighted route to the North Shelter. It is estimated to take 12 minutes. Please do not go to the garage to drive your car." At the same time, the building's public address system in Building 3 also broadcast: "Residents of Building 3, please note that you should immediately go downstairs via the stairs and follow the instructions on your mobile phone or the staff to evacuate to the North Shelter on foot." In resident family B / C, C's mobile phone will receive the route to the assembly point in Building 5, while B (or a family member)'s mobile phone will receive a message saying, "Please get ready and wait for family member C. We are expected to leave together from the south gate in 8 minutes," along with driving directions.

[0039] During the evacuation, the system continuously monitored the situation. If a newly installed sensor near the North Gate suddenly reported a rapid rise in water level exceeding the predicted value, this deviation triggered the system to recalculate. After the model re-assimilated the new data, it predicted a surge in risk at the North Gate exit within the next 10 minutes. The system immediately replanned: it calculated new routes for residents who had not yet departed and originally planned to use the North Gate (such as some residents of Building 7), potentially changing their route to the East Shelter. It also issued updated instructions via the app and broadcast: "Residents of Building 7, please note that the original North Gate route has increased in risk. Please follow the new highlighted route to the East Shelter." For residents already en route, the system determined their location based on their mobile phones. If they were close to the North Gate and the risk was still acceptable, they were advised to speed up their journey. If they were still far away, they were advised to temporarily take refuge in a safe location and wait for the new route.

[0040] Through the above closed-loop process, precise, efficient, safe and humane emergency evacuation guidance for community residents is achieved in complex and dynamic urban flooding disasters.

[0041] This invention achieves real-time, accurate perception and high-confidence prediction of community flooding by constructing a multi-layered fusion digital twin foundation and utilizing multi-source data assimilation technology. Furthermore, based on a spatiotemporal dynamic risk map, it employs a two-layer optimization model combining macro-level flow allocation and micro-level personalized path planning. This model generates differentiated and forward-looking evacuation plans for different resident groups, balancing overall evacuation efficiency with individual safety constraints. Finally, through multi-channel collaborative dissemination and a closed-loop dynamic adjustment mechanism based on real-time monitoring, it ensures reliable delivery of evacuation instructions and agile adaptation to the evolving disaster situation. This systematically solves the problems of lagging situational awareness, static and homogeneous path planning, lack of predictive ability, and susceptibility to evacuation conflicts inherent in traditional methods, significantly improving the safety, efficiency, and fairness of emergency evacuation during urban flooding disasters.

[0042] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the present invention and embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0043] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0044] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for guiding residents' emergency evacuation routes during urban flooding in smart communities, characterized in that: Includes the following steps: Construct a multi-layered, integrated digital twin foundation for the community that includes geographic information, building information, real-time sensing data, and hydrodynamic models; Real-time aggregation and fusion of multi-source observation data, driven by data assimilation technology, dynamically inverts and predicts the three-dimensional situation of community flooding; Based on real-time and forecast data of the waterlogging situation, the dynamic traffic risk of each road segment in the evacuation route network in the future multiple time windows is calculated, and a spatiotemporal dynamic risk map is generated. Identify the individual attributes and constraints of residents to be evacuated, and construct a personalized evacuation demand model with the goal of the shortest safe arrival time; Based on the dynamic risk map, a two-layer optimization model combining macro-level evacuation flow allocation and micro-level personalized route planning is adopted to calculate a spatiotemporal evacuation plan that meets the needs of residents or family groups, including suggested departure times and specific route sequences. The spatiotemporal evacuation plan is pushed to the target residents, and dynamic evaluation and plan adjustment are carried out based on real-time monitoring and prediction deviations during the evacuation process.

2. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The construction of a multi-layered integrated digital twin foundation for the community specifically includes: integrating high-precision geographic information system data, building information model data of major buildings, IoT sensing layer data consisting of water level gauges, flow meters and video surveillance deployed in the community, and surface and pipeline coupled hydrodynamic model data calibrated with local parameters, forming a computable digital twin platform under a unified spatiotemporal coordinate system.

3. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The multi-source observation data includes at least: weather radar and gridded rainfall forecast data, community IoT sensor data, anonymous crowdsourced mobile speed and location data from mobile terminals, and text or image water accumulation report data extracted from social media. The data assimilation technique employs Kalman filtering or ensemble Kalman filtering to fuse observed data with model simulation results, thereby continuously correcting the model state and improving the accuracy of situational awareness.

4. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The dynamic traffic risk R(t) is calculated as follows: R(t) = α * Rs + β * Rh(t) + γ * Ru(t), where Rs is the static infrastructure risk factor, Rh(t) is the dynamic hydraulic risk factor calculated based on the predicted water depth and flow velocity at time t in the future, Ru(t) is the predicted risk factor calculated based on the uncertainty of model prediction, and α, β, and γ are configurable weight coefficients.

5. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 4, is characterized in that... The specific steps for generating the spatiotemporal dynamic risk map include: Discretize all potential evacuation routes within the community into road segment units; For each road segment unit, the water depth and flow velocity at each future time step are predicted based on the hydrodynamic model. Based on the static attributes of the road segment, the risk value at each time step is calculated according to the formula for calculating the dynamic traffic risk R(t); the change of the risk value over time is visualized on the digital twin substrate in the form of a heat map.

6. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The individual attributes and constraints include at least: movement mode, movement speed, whether assistance is required, accompanying persons, available vehicle types, and family grouping preferences; in the shortest safe arrival time target, "safety" is defined as the dynamic traffic risk value of each road segment on the planned path at the time of resident passage, which is lower than the preset risk tolerance threshold of the corresponding population-vehicle combination for that resident.

7. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The macro-level evacuation flow allocation model uses each building or community within the community as the source and each emergency shelter as the sink. Based on the dynamic risk map, it constructs a time-varying impedance network and uses the dynamic user equilibrium or system optimality principle to solve for the flow allocation scheme that minimizes the overall evacuation time and the departure rate of each source in different time periods. The micro-level personalized path planning, within the framework of the macro-level flow allocation scheme, uses a time-dependent shortest path algorithm or a spatiotemporal A* search algorithm in an extended spatiotemporal network for each resident or family group to solve for the specific path sequence and departure time from the starting point to the shelter that satisfies their individual constraints.

8. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The methods for pushing out the spatial and temporal evacuation plan include one or more combinations of sending graphical navigation information to residents via smartphone applications, playing voice guidance through community building broadcasts, publishing text prompts through community variable information signs, and using drones for aerial announcements and guidance.

9. The method for guiding emergency evacuation routes for residents in smart communities during urban flooding, as described in claim 1, is characterized in that... The triggering conditions for the dynamic assessment and plan adjustment include any one or more of the following: receiving a higher-level rainstorm warning signal, the deviation between the monitoring data and the predicted value of key location sensors exceeding a preset threshold, a significant adverse change in the dynamic risk map, and the deviation between the actual monitored evacuation flow / vehicle flow density and the planned predicted value exceeding a preset threshold.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1 to 9.

Images