Rural refugee evacuation simulation method based on simulation model

Abstract

The invention provides a rural refugee evacuation simulation method based on a simulation model, and belongs to the technical field of evacuation methods, and the method comprises the steps: S1, collecting design indexes and influence factors of a rural refuge space; s2, constructing a rural refuge space planning design database; s3, four intelligent agents including resident, refuge site, refuge road and environment dynamic change are constructed; s4, constructing an environment model according to the actual condition of the target rural area, and embedding the four agents into the environment model to form a complete rural refuge emergency agent simulation model; s5, designing an interaction mechanism among the four types of agents; s6, setting a simulation scene according to the actual condition of the target rural area; s7, starting the rural refuge emergency agent simulation model, and simulating the evacuation process of the rural refuge personnel under the emergency condition; and S8, evaluating a simulation result. According to the invention, the emergency evacuation strategy of the rural area can be optimized, and the efficiency and safety of coping with emergencies such as natural disasters can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of evacuation methods, and in particular, to a simulation method for evacuating rural refuge seekers based on a simulation model. Background Art

[0002] In the context of the increasing safety needs in rural areas where natural disasters occur frequently, traditional emergency evacuation planning and management methods have been difficult to meet the increasingly complex evacuation needs. Due to the diverse geographical environment, relatively weak infrastructure, and wide distribution of residents in rural areas, the emergency evacuation work in rural areas faces many challenges.

[0003] To better address these challenges and improve the emergency response ability and evacuation efficiency in rural areas, using modern simulation technology to simulate the emergency evacuation process has become an effective solution. By constructing a virtual environment, simulation technology can simulate the flow and distribution of people and materials in emergency situations, helping decision-makers evaluate the effects of different evacuation strategies and emergency plans before real events occur. This method can not only reduce the risks and uncertainties in actual evacuations but also provide a scientific basis for optimizing evacuation routes, improving the utilization efficiency of shelters, and strengthening the allocation of emergency resources.

[0004] Therefore, based on the above requirements, it is of great significance to research and develop a simulation model that can accurately simulate the evacuation process of rural refuge seekers. Such a model will combine the actual situation in rural areas, considering various factors such as topography, population distribution, and traffic conditions, providing comprehensive and reliable simulation results and data analysis for decision-makers, and providing strong support for emergency management and public safety work in rural areas. Summary of the Invention

[0005] The purpose of the present invention is to provide a simulation method for evacuating rural refuge seekers based on a simulation model, which can optimize the emergency evacuation strategy in rural areas and improve the efficiency and safety in dealing with emergencies such as natural disasters.

[0006] To achieve the above purpose, the present invention provides a simulation method for evacuating rural refuge seekers based on a simulation model, including the following steps: Step S1, collect the design indicators and influencing factors of rural refuge spaces; Step S2, construct a database for the planning and design of rural refuge spaces according to the design indicators and influencing factors; Step S3, construct resident agents, shelter site agents, evacuation road agents, and environmental dynamic change agents; Step S4, construct an environmental model according to the actual situation of the target rural area, and embed the resident agents, shelter site agents, evacuation road agents, and environmental dynamic change agents into it to form a complete rural refuge emergency agent simulation model; Step S5: Design the interaction mechanism among the four types of agents; Step S6: Set the simulation scenario according to the actual situation of the target rural area; Step S7: Start the rural evacuation emergency agent simulation model, simulate the evacuation process of rural evacuees in an emergency, and record the performance behaviors of each agent; Step S8: Evaluate the simulation results, and optimize the agent behavior rules and environmental models according to the evaluation results.

[0007] Preferably, in Step S1, the design indicators include evacuation shelters and evacuation roads; the influencing factors include residents' behavior factors, spatial environment characteristics, infrastructure characteristics, and natural environment characteristics.

[0008] Preferably, in Step S2, construct a rural evacuation space planning and design database according to the design indicators and influencing factors. The specific operation is as follows: conduct a Pearson correlation analysis on the design indicators and influencing factors, identify the design indicators and influencing factors that have a significant impact on rural evacuation space planning and design, and construct a rural evacuation space planning and design database.

[0009] Preferably, in Step S3, the basic attributes of the resident agent include age, gender, mobility, and mental state; the behavior rules of the resident agent include environmental perception strategy, evacuation strategy, path selection strategy, exit selection strategy, and action execution strategy; The environmental perception strategy is as follows: ; Among them, represents the perception distance; represents the position of the resident agent; represents the position of the evacuation site or obstacle; The evacuation strategy is as follows: ; Among them, represents the risk assessment; represents the risk threshold; represents the evacuation decision; ; Among them, represents the distance of the resident agent to the nearest disaster point; represents the score of the current traffic condition; represents the impact score of environmental obstacles; , , all represent weights; The path selection strategy is as follows: ; Among them, represents the optimal path; represents the weight of the th segment in the path; represents the length of the path segment; represents the total number of all segments in the path; ; Among them, represents the best exit; represents the maximum number of people that the exit of the shelter site can accommodate; represents the proportion of the number of people currently using this exit; The action execution strategy is as follows: ; Among them, represents the action speed; represents the remaining distance to the shelter site; represents the remaining time; The basic attributes of the shelter road agent include road width, road length, and road passing capacity; the behavior rules of the shelter road agent include simulating traffic conditions, calculating passing time, and obstacle avoidance and path optimization; Simulating traffic conditions is as follows: ; Among them, represents the traffic flow; represents the speed of the th vehicle; represents the proportion of the road occupied by this vehicle; ; Among them, represents the passing time; represents the road length; represents the average vehicle speed; Obstacle avoidance and path optimization are as follows: ; Among them, represents the optimal obstacle avoidance path; represents the influence degree of the obstacle; represents the weight coefficient; The basic attributes of the shelter site agent include location, capacity, and accessible roads; the behavior rules of the shelter site agent include receiving shelter personnel strategy, evaluating accommodation capacity, and material distribution strategy; The receiving shelter personnel strategy is as follows: ; Among them, represents the capacity allocation; represents the number of resident agents attempting to enter the shelter site; represents the number of people that the shelter site can accommodate; The accommodation capacity is evaluated as follows: ; Among them, represents the accommodation capacity; represents the number of people already in the shelter site; represents the maximum number of people that the shelter site can accommodate; The material distribution strategy is as follows: ; Among them, represents the material allocation; represents the total amount of materials in the shelter site; represents the number of resident agents in the shelter site; The basic attributes of the environmental dynamic change agent include perception ability, data processing ability, prediction ability, and communication ability; the behavior rules of the environmental dynamic change agent include real-time monitoring of the external environment, updating of environmental data, analysis of environmental data, and sharing of environmental data.

[0010] Preferably, a learning mechanism is also introduced into the resident agent, specifically: Define the state space : Define various situations encountered by the resident agent as states, and various situations include the distance from the shelter site, road congestion conditions, and environmental obstacles; Design the action space : Design possible actions for the resident agent, including choosing different evacuation routes and adjusting the walking speed; Set the reward function : Give corresponding rewards or punishments according to the behavior results of the resident agent, and the behavior results include the time to successfully reach the shelter site and the number of times of avoiding collisions; ; Among them, represents the reward for reaching the shelter site, represents the reward for avoiding collisions; Training process: Let the resident agent continuously try different actions in the simulated environment and adjust the strategy according to the reward function until the optimal evacuation path and behavior pattern are found; ; Among them, represents taking action in state 's value represents the learning rate represents the discount factor represents the new state after taking the action represents the action taken in the new state represents the updated value.

[0011] Preferably, in step S4, an environment model is constructed according to the actual situation of the target rural area, and the resident agent, the refuge site agent, the refuge road agent and the environmental dynamic change agent are embedded therein to form a complete rural refuge emergency agent simulation model. The specific operation is as follows: Integrate the influencing factors collected in step S1 and perform processing operations, including data cleaning, data transformation and data verification; Use 3D modeling software to construct a geographical environment model of the target rural area according to the processed data, including terrain, building structure and road network; Based on the geographical environment model, layout the refuge sites and refuge roads; Design a set of interaction interfaces for the interaction between each agent and the environment model. The interaction interfaces include data interfaces, control interfaces and visualization interfaces. The data interfaces define the formats and protocols for data transmission between each agent and the environment model. The control interfaces provide interfaces for control commands for each agent. The visualization interfaces are used to display the real-time status of the environment model and each agent. After the environment model is constructed, perform model verification and adjustment, including logical verification, parameter adjustment and test run.

[0012] Preferably, in step S5, the interaction mechanisms between the four types of agents include: the interaction mechanism between resident agents, the interaction mechanism between resident agents and refuge site agents, the interaction mechanism between resident agents and refuge road agents, the interaction mechanism between resident agents and environmental dynamic change agents, the interaction mechanism between refuge site agents and refuge road agents, the interaction mechanism between refuge site agents and environmental dynamic change agents, and the interaction mechanism between refuge road agents and environmental dynamic change agents.

[0013] Preferably, in step S6, set the simulation scenario according to the actual situation of the target rural area. The specific operation is as follows: First, set the simulation scenario parameters, including rural area characteristics, refuge site layout, resident distribution, refuge road network, emergency event type and occurrence time; Then, import geographical data into the simulation system, and set the attributes, location layout, and interaction mechanism of each agent. Finally, set the simulation conditions, including the simulation time length, data recording frequency, and output format.

[0014] Preferably, in step S7, start the rural evacuation emergency agent simulation model, simulate the evacuation process of rural evacuees in an emergency, and record the performance behaviors of the agents, including the following steps: Load the rural evacuation emergency agent simulation model constructed in step S4, and initialize all agents. Start the engine of the rural evacuation emergency agent simulation model, trigger the emergency time, and according to the type and occurrence time of the emergency event set in step S6, make the resident agents perceive the emergency situation. Conduct simulation runs, including simulation of resident agent behaviors, shelter site agent behaviors, evacuation route agent behaviors, and environmental dynamic change agent behaviors. Data recording and monitoring, including recording the evacuation time, path selection, and moving speed of resident agents; recording the traffic conditions and passing time of evacuation routes; monitoring the accommodation situation, utilization rate, and possible congestion points of shelter sites. When all resident agents reach the shelter site or reach the preset simulation duration, the simulation ends.

[0015] Preferably, in step S7, during the simulation process, an abnormal situation detection mechanism is also set, and the detection mechanism is as follows: During the simulation process, the system monitors the status of each agent, environmental changes, and traffic flow data in real time. Set thresholds based on historical data to judge the occurrence of abnormal situations. When the analyzed real-time data exceeds the threshold, trigger the abnormal situation detection mechanism. Calculate the Z - score of each index to judge the abnormal threshold: ; Wherein, represents the observed value; represents the historical mean; represents the historical standard deviation.

[0016] Therefore, the present invention adopts the above - mentioned method for simulating the evacuation of rural evacuees based on a simulation model, and the beneficial technical effects are as follows: (1) Precise simulation of complex environments: By constructing detailed environmental models, including terrain, building structures, road networks, and dynamically changing environmental factors (such as weather, progress of natural disasters, etc.), the simulation model can accurately simulate the complex environments in rural areas. This high-precision simulation enables decision-makers to more accurately evaluate the feasibility of evacuation strategies under different environments, providing a basis for formulating adaptable emergency plans.

[0017] (2) Optimization of agent behavior simulation: The introduction of resident agents, shelter site agents, evacuation route agents, and environmental dynamic change agents enables the simulation model to simulate the behavioral responses of various agents in emergency situations. Especially the learning mechanism in resident agents can find the optimal evacuation routes and behavioral patterns through continuous trial and error and optimization, improving evacuation efficiency and reducing risks. This agent behavior simulation technology helps to identify and optimize key issues at the individual behavior level.

[0018] (3) Data-driven decision support: A large amount of data collected during the simulation process, such as evacuation time, route selection, traffic conditions, and utilization rate of shelter sites, provides rich data support for decision-makers. Through data analysis, bottlenecks and key nodes in the evacuation process can be identified, and then targeted improvement measures can be formulated. In addition, the simulation model can also conduct sensitivity analysis and scenario comparison to help decision-makers evaluate the effects of different strategies and select the optimal solution.

[0019] (4) Real-time feedback and dynamic adjustment: During the simulation process, the set anomaly detection mechanism can monitor the states of various agents and environmental changes in real time. Once an anomaly is detected, the feedback mechanism is immediately triggered. This real-time feedback ability enables decision-makers to adjust strategies in a timely manner during the simulation, optimize resource allocation, and improve the flexibility and adaptability of emergency responses. Brief Description of the Drawings

[0020] Figure 1 It is a flowchart of a method for simulating the evacuation of rural refuge seekers based on a simulation model according to the present invention; Figure 2 It is a schematic diagram of the interaction mechanism of four types of agents. Detailed Embodiments

[0021] The technical solutions of the present invention will be further described below with reference to the drawings and embodiments.

[0022] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs.

[0023] Embodiment 1 As Figure 1As shown in the figure, it is a flowchart of a method for simulating the evacuation of rural refuge seekers based on a simulation model according to the present invention, including the following steps: Step S1: Collect the design indicators and influencing factors of rural refuge spaces.

[0024] The design indicators include refuge sites and refuge roads; the influencing factors include residents' behavioral factors, spatial environmental characteristics, infrastructure characteristics, and natural environmental characteristics.

[0025] More specific influencing factors include the location selection of refuge sites, the capacity of refuge sites, the topology of road networks, road widths, road accessibility, individual characteristics of residents, refuge responses, evacuation action patterns, main disaster sources, topography, building distribution, weather conditions, etc.

[0026] Step S2: Construct a database for the planning and design of rural refuge spaces based on the design indicators and influencing factors.

[0027] Specifically: Conduct statistical analysis and correlation analysis on the design indicators and influencing factors, and construct a database for the planning and design of rural refuge spaces.

[0028] Step S3: Construct resident agents, refuge site agents, refuge road agents, and environmental dynamic change agents.

[0029] The basic attributes of resident agents include age, gender, mobility, and mental state; the behavioral rules of resident agents include environmental perception strategies, evacuation strategies, path selection strategies, exit selection strategies, and action execution strategies.

[0030] Environmental perception strategy: Define how resident agents perceive the location and status of emergency events, refuge sites, and refuge roads: ; Among them, represents the perception distance; represents the location of the resident agent; represents the location of the refuge site or obstacle.

[0031] Evacuation strategy: Determine whether to evacuate and the evacuation timing based on the severity of the emergency event and the individual's location: ; Among them, represents the risk assessment; represents the risk threshold; represents the evacuation decision.

[0032] Path selection strategy: Consider factors such as the shortest path, obstacle avoidance, and congestion avoidance, and select the best evacuation path: ; Among them, Indicates the optimal path; Indicates the weight of the th segment in the path; Indicates the length of the path segment;

[0033] Exit selection strategy: Considering the distance, capacity, and congestion of the exits comprehensively, select the most suitable exit of the shelter site: ; Among them, Indicates the best exit; Indicates the maximum number of people that the exit of the shelter site can accommodate; Indicates the proportion of the number of people currently using this exit.

[0034] Action execution strategy: According to the decision result, execute the movement operation and go to the shelter site: ; Among them, Indicates the action speed; Indicates the remaining distance to the shelter site; Indicates the remaining time.

[0035] The basic attributes of the shelter road agent include road width, road length, and road passing capacity; the behavior rules of the shelter road agent include simulating traffic conditions, calculating passing time, and obstacle avoidance and path optimization.

[0036] Simulating traffic conditions: According to the movement of the resident agent and the road passing capacity, simulate the traffic congestion situation: ; Among them, Indicates the traffic flow; Indicates the speed of the th vehicle;

[0037] Indicates the proportion of the road occupied by this vehicle. Calculating passing time: Provide the estimated passing time from the starting point to the ending point for the resident agent: Among them, Indicates the passing time; Indicates the road length; Indicates the average vehicle speed.

[0038] Obstacle avoidance and path optimization: Considering the obstacles and congestion points on the road, provide the best path advice for the resident agent: ; Among them, Represents the optimal obstacle avoidance path; Represents the influence degree of obstacles; Represents the weight coefficient.

[0039] The basic attributes of the refuge site agent include location, capacity, and accessible roads; the behavior rules of the refuge site agent include the strategy of receiving refuge seekers, evaluating the accommodation capacity, and the material distribution strategy.

[0040] Strategy of receiving refuge seekers: According to the capacity limit, receive and manage the resident agents entering the refuge site: ; Among them, Represents the capacity allocation; Represents the number of resident agents attempting to enter the refuge site; Represents the number of people that the refuge site can accommodate.

[0041] Evaluating the accommodation capacity: Real-time evaluate the remaining capacity of the refuge site to provide information for the evacuation decision-making of resident agents: ; Among them, Represents the accommodation capacity; Represents the number of people already in the refuge site; Represents the maximum number of people that the refuge site can accommodate.

[0042] The material distribution strategy is as follows: ; Among them, Represents the material distribution; Represents the total amount of materials in the refuge site; Represents the number of resident agents in the refuge site.

[0043] The basic attributes of the environmental dynamic change agent include perception ability, data processing ability, prediction ability, and communication ability; the behavior rules of the environmental dynamic change agent include real-time monitoring of the external environment, updating environmental data, analyzing environmental data, and sharing environmental data.

[0044] Real-time monitoring of the external environment: Continuously monitor the changes in the external environment, including the occurrence and development of disasters, as well as factors such as traffic and weather that may affect evacuation.

[0045] Updating environmental data: Regularly or real-time update environmental data to ensure the accuracy and timeliness of information.

[0046] Analyzing environmental data: Based on the collected data and analysis results, predict the future change trend of the environment and send early warning information to other agents in a timely manner.

[0047] Shared environmental data: Establish a communication connection with other agents to share environmental information and evacuation status, promoting collaborative decision-making and actions.

[0048] A learning mechanism is also introduced in the resident agent, specifically: Define the state space : Define various situations encountered by the resident agent as states, including the distance to the evacuation site, road congestion, and environmental obstacles. Design the action space : Design possible actions for the resident agent, including choosing different evacuation routes and adjusting walking speeds. Set the reward function : Give corresponding rewards or punishments according to the behavior results of the resident agent, including the time to successfully reach the evacuation site and the number of collisions avoided. ; Among them, represents the reward for reaching the evacuation site, represents the reward for avoiding collisions; Training process: Let the resident agent continuously try different actions in the simulated environment and adjust the strategy according to the reward function until the optimal evacuation path and behavior pattern are found. ; Among them, represents the value of taking action in state , represents the learning rate, represents the discount factor, represents the new state after taking action , represents the action taken in the new state , represents the updated value.

[0049] Step S4: Construct an environmental model according to the actual situation of the target rural area, and embed the resident agent, evacuation site agent, evacuation road agent, and environmental dynamic change agent into it to form a complete rural evacuation emergency agent simulation model.

[0050] Specifically: Integrate the influencing factors collected in step S1 and perform processing operations, including data cleaning, data transformation, and data verification. Use 3D modeling software to construct a geographical environment model of the target rural area based on the processed data, including terrain, building structure, and road network. Topography: Based on elevation data or satellite imagery, simulate the natural features such as terrain undulations, rivers, and lakes in rural areas.

[0051] Building Structure: Based on the CAD drawings of buildings or on-site measurement data, construct 3D models of buildings, including buildings such as houses, warehouses, schools, etc. that may serve as shelters.

[0052] Road Network: Based on road planning data or on-site surveys, draw the road network in rural areas, including main roads, secondary roads, country lanes, etc., and label attributes such as the width, length, and traffic capacity of the roads.

[0053] Based on the geographical environment model, layout the shelter sites and evacuation roads; Shelter Site Location: Precisely locate the shelter sites (such as school playgrounds, open spaces, etc.) in the model and label their attributes such as capacity and facilities.

[0054] Road Connection: Ensure that all shelter sites are connected to the road network, and the road width and traffic capacity can meet the evacuation requirements.

[0055] Obstacle Setting: Set obstacles that may affect evacuation in the model, such as trees, walls, bridges, etc., and label their locations and attributes.

[0056] Design an interactive interface for the interaction between each intelligent agent and the environment model. The interactive interface includes a data interface, a control interface, and a visualization interface. The data interface defines the format and protocol for data transmission between each intelligent agent and the environment model. The control interface provides an interface for control commands for each intelligent agent. The visualization interface is used to display the real-time status of the environment model and each intelligent agent; The data interface is responsible for defining the format and protocol for data transmission between each intelligent agent and the environment model. Specifically, the data interface includes the following types of data transmission: Status Data: The position, speed, health status, etc. of the resident intelligent agent.

[0057] The capacity, current occupancy, etc. of the shelter site intelligent agent.

[0058] The traffic flow, congestion status, etc. of the evacuation road intelligent agent.

[0059] The weather conditions, disaster progress, etc. of the environmental dynamic change intelligent agent.

[0060] Perception Data: The perception information of the resident intelligent agent about the surrounding environment, such as the distance to the nearest shelter site, the location of surrounding obstacles, etc.

[0061] The perception information of the shelter site intelligent agent about the inflow and outflow of shelter seekers.

[0062] Request / response data: The evacuation request sent by the resident agent to the evacuation site agent.

[0063] The response of the evacuation site agent to the resident agent, such as accepting or rejecting the evacuation request.

[0064] Control Interface The control interface provides an interface for each agent to receive control commands. The control commands include: Movement command: The command for the resident agent to move in a specific direction, such as "Move 100 meters north".

[0065] Evacuation command: The command for the resident agent to search for an evacuation site, such as "Search for the nearest evacuation site and go there".

[0066] Resource allocation command: The command for the evacuation site agent to allocate resources (such as food and water).

[0067] Traffic control command: The command for the evacuation road agent to control the traffic flow, such as "Road closed" or "Change traffic lights".

[0068] Visualization Interface The visualization interface is used to display the environmental model and the real-time status of each agent. The specific information to be displayed includes: Environmental status view: Displays a map of the entire environmental model, including terrain, buildings, road networks, etc.

[0069] Agent status view: Realtime display of the location, movement direction and speed of the resident agent.

[0070] Displays the current capacity and the number of evacuees of the evacuation site agent.

[0071] Event view: Displays the location and scope of influence of emergency events, such as fires, floods, etc.

[0072] Traffic view: Displays the traffic flow and congestion on the evacuation roads.

[0073] After the environmental model is constructed, perform model verification and adjustment, including logical verification, parameter adjustment and test run.

[0074] Step S5: Design the interaction mechanism among the four types of agents.

[0075] As Figure 2 shown, the interaction mechanism among the four types of agents includes: the interaction mechanism between resident agents, the interaction mechanism between resident agents and shelter site agents, the interaction mechanism between resident agents and evacuation route agents, the interaction mechanism between resident agents and environmental dynamic change agents, the interaction mechanism between shelter site agents and evacuation route agents, the interaction mechanism between shelter site agents and environmental dynamic change agents, and the interaction mechanism between evacuation route agents and environmental dynamic change agents.

[0076] Resident agents and resident agents: Information exchange: Resident agents can exchange information such as evacuation directions, shelter site conditions, and personal status through wireless communication or visual perception. This helps to form group behaviors such as following leaders and avoiding congestion.

[0077] Collision avoidance mechanism: During movement, resident agents need to achieve collision avoidance to prevent collisions or congestion. This can be achieved by setting a safety distance, predicting path conflicts, and adjusting the movement direction.

[0078] Resident agents and shelter site agents: Matching and allocation: Resident agents select appropriate shelter sites for evacuation based on the information provided by shelter site agents (such as capacity, location, facilities, etc.). Shelter site agents then allocate and schedule resources according to the received requests.

[0079] Status feedback: Shelter site agents provide feedback on the current status (such as the number of people already accommodated, remaining capacity, facility usage, etc.) to resident agents to help resident agents make more reasonable decisions.

[0080] Resident agents and evacuation route agents: Path planning: Resident agents plan the optimal evacuation path based on the road network information and real-time traffic conditions provided by evacuation route agents. Evacuation route agents are responsible for updating the road status (such as congestion level, closure situation, etc.) to ensure the effectiveness of path planning.

[0081] Traffic control: During the emergency evacuation process, evacuation route agents may need to control traffic to a certain extent, such as setting temporary traffic signals and adjusting lane directions, to optimize the evacuation efficiency. Resident agents need to follow these traffic control instructions to move.

[0082] Resident agents and environmental dynamic change agents: Environmental perception: The resident agent obtains real-time environmental change information (such as fire spread, weather changes, seismic wave propagation, etc.) through sensors or simulated environmental perception functions. This information affects the decisions and actions of the resident agent.

[0083] Adaptation and response: Based on the environmental change information, the resident agent needs to adjust its behavioral strategies, such as changing the evacuation direction, accelerating the movement speed, finding a temporary shelter, etc. The environmental dynamic change agent is responsible for simulating and providing this change information.

[0084] Shelter site agent and evacuation route agent: Collaborative optimization: The shelter site agent and the evacuation route agent need to perform collaborative optimization to ensure the efficiency and smoothness of the evacuation process. For example, when a certain shelter site is approaching saturation, the evacuation route agent can adjust the traffic flow direction to guide the resident agent to other shelter sites.

[0085] Information sharing: Real-time information such as the capacity change of the shelter site and the road congestion situation is shared between the two to make more accurate decisions and scheduling.

[0086] Shelter site agent and environmental dynamic change agent: Risk assessment: The shelter site agent evaluates the safety and sustainability of the shelter site based on the information provided by the environmental dynamic change agent (such as natural disaster threats, environmental pollution, etc.).

[0087] Emergency response: In the face of serious environmental threats, the shelter site agent needs to activate the emergency response mechanism, such as strengthening protective measures, evacuating the internal residents, etc., and maintaining close communication with the environmental dynamic change agent to obtain the latest information.

[0088] Evacuation route agent and environmental dynamic change agent: Road condition update: The environmental dynamic change agent provides real-time road condition information (such as damage degree, water accumulation situation, etc.) to the evacuation route agent so that the evacuation route agent can update the road network diagram and adjust the path planning algorithm.

[0089] Traffic control suggestions: Based on the environmental change information, the evacuation route agent may need to put forward traffic control suggestions (such as blocking dangerous sections, opening temporary channels, etc.) to the traffic management department to ensure the safety and smoothness of the evacuation process.

[0090] Step S6. Set up a simulation scenario according to the actual situation of the target rural area.

[0091] First, set the simulation scenario parameters, including rural area characteristics, shelter site layout, resident distribution, evacuation road network, emergency event type and occurrence time; Then, import geographical data into the simulation system, and set the attributes, location layouts, and interaction mechanisms of each agent; Finally, set the simulation conditions, including the simulation time length, data recording frequency, and output format.

[0092] Step S7: Start the rural evacuation emergency agent simulation model, simulate the evacuation process of rural evacuees in an emergency, and record the performance behaviors of each agent.

[0093] Load the rural evacuation emergency agent simulation model constructed in Step S4, and initialize all agents; Start the engine of the rural evacuation emergency agent simulation model, trigger an emergency event, and according to the emergency event type and occurrence time set in Step S6, make the resident agents perceive the emergency situation; Conduct simulation runs, including resident agent behavior simulation, evacuation site agent behavior simulation, evacuation route agent behavior simulation, and environmental dynamic change agent behavior simulation; Data recording and monitoring, including recording the evacuation time, path selection, and moving speed of resident agents; recording the traffic conditions and passing times of evacuation routes; monitoring the accommodation situation, utilization rate, and possible congestion points of evacuation sites; When all resident agents have reached the evacuation site or the preset simulation duration is reached, the simulation ends.

[0094] During the simulation process, an abnormal situation detection mechanism is also set up.

[0095] The detection mechanism is as follows: During the simulation process, the system monitors the states of each agent, environmental changes, and traffic flow data in real time; Set thresholds based on historical data to judge the occurrence of abnormal situations. When the analyzed real-time data exceeds the threshold, trigger the abnormal situation detection mechanism; Calculate the Z - score of each index to judge the abnormal threshold: ; Among them, represents the observed value; represents the historical mean; represents the historical standard deviation.

[0096] Step S8: Evaluate the simulation results, and optimize the agent behavior rules and environmental models according to the evaluation results.

[0097] It should be noted that the content not elaborated in detail in the present invention is all prior art and is well known to those skilled in the art.

[0098] Therefore, by adopting the above-mentioned method for simulating the evacuation of rural refuge seekers based on a simulation model, the present invention can optimize the emergency evacuation strategy in rural areas and improve the efficiency and safety in dealing with emergencies such as natural disasters.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions of the present invention or make equivalent replacements, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A simulation method for evacuating rural refugees based on a simulation model, characterized in that: The following steps are involved: Step S1, collecting design indicators and influencing factors of rural refuge spaces; Step S2, constructing a rural refuge space planning and design database based on design indicators and influencing factors; Step S3, constructing resident agents, refuge site agents, refuge road agents and environment dynamic change agents; Step S4: construct an environmental model according to the actual situation of the target rural area, embed the resident agent, the refuge site agent, the refuge road agent and the environment dynamic change agent into it, and form a complete rural refuge emergency agent simulation model; Step S5: design the interaction mechanism between the four types of agents; Step S6: setting a simulation scenario according to the actual situation of the target rural area; Step S7, start the rural emergency refuge agent simulation model, simulate the evacuation process of rural refugees in an emergency, and record the performance of each agent; Step S8: Evaluate the simulation results, and optimize the agent behavior rules and environment model according to the evaluation results.

2. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 1, characterized in that: In step S1, the design indicators include refuge places and refuge roads; the influencing factors include resident behavior factors, spatial environment characteristics, infrastructure characteristics and natural environment characteristics.

3. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 1, characterized in that: In step S2, a rural refuge space planning and design database is constructed based on design indicators and influencing factors. The specific operations are: performing Pearson correlation analysis on design indicators and influencing factors, identifying design indicators and influencing factors that have a significant impact on rural refuge space planning and design, and constructing a rural refuge space planning and design database.

4. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 3, characterized in that: In step S3, the basic attributes of the resident agent include age, gender, mobility and psychological state; the behavior rules of the resident agent include environmental perception strategy, evacuation strategy, path selection strategy, exit selection strategy and action execution strategy; The environmental awareness strategies are as follows: ； in, Indicates perceived distance; represents the location of the resident agent; Indicates the location of a refuge area or obstacle; The evacuation strategy is as follows: ； in, represents risk assessment; represents the risk threshold; Indicates evacuation decision; ； in, Represents the distance from the resident agent to the nearest disaster point; A score indicating the current traffic conditions; represents the impact score of environmental barriers; , , All represent weights; The path selection strategy is as follows: ； in, represents the optimal path; Indicates the path The weight of the segment; Indicates the length of the path segment; Indicates the total number of all segments in the path; The export selection strategy is as follows: ； in, Indicates the best exit; Indicates the maximum number of people that the exit of the shelter can accommodate; Indicates the proportion of people currently using the exit; The action execution strategy is as follows: ； in, Indicates the speed of action; Indicates the remaining distance to the refuge site; Indicates the remaining time; The basic attributes of the refuge road agent include road width, road length and road capacity; the behavior rules of the refuge road agent include simulating traffic conditions, calculating travel time, and avoiding obstacles and optimizing paths; The simulated traffic conditions are as follows: ； in, Indicates traffic flow; Indicates The speed of the vehicle; Indicates the proportion of the road occupied by the vehicle; The travel time is calculated as follows: ； in, Indicates the travel time; Indicates the length of the road; Indicates the average vehicle speed; Obstacle avoidance and path optimization are as follows: ； in, represents the optimal obstacle avoidance path; Indicates the influence of obstacles; represents the weight coefficient; The basic attributes of the shelter agent include location, capacity and accessible roads; the behavior rules of the shelter agent include strategies for receiving refugees, evaluating capacity and material distribution strategies; The strategy for receiving refugees is as follows: ； in, Indicates capacity allocation; Represents the number of resident agents trying to enter the shelter; Indicates the number of people that the shelter can accommodate; The assessment capacity is as follows: ； in, Indicates capacity; Indicates the number of people already in the shelter; Indicates the maximum number of people that can be accommodated at the shelter; The material distribution strategy is as follows: ； in, Indicates the distribution of materials; Indicates the total amount of supplies in the shelter site; Represents the number of resident agents in the shelter; The basic attributes of the environment-dynamically changing intelligent agent include perception ability, data processing ability, prediction ability and communication ability; the behavioral rules of the environment-dynamically changing intelligent agent include real-time monitoring of the external environment, updating environmental data, analyzing environmental data and sharing environmental data.

5. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 4, characterized in that: A learning mechanism is also introduced into the resident agent, specifically: Defining the state space : Define various situations encountered by resident agents as states, including the distance from the shelter, road congestion, and environmental obstacles; Designing the Action Space : Design possible actions for resident agents, including choosing different evacuation routes and adjusting walking speed; Setting the reward function : Give corresponding rewards or penalties based on the behavior results of the resident intelligent agent, including the time to successfully reach the shelter and the number of collisions avoided; ； in, Rewards for reaching the evacuation site. represents the reward for avoiding collision; Training process: Let the resident agent continuously try different actions in the simulated environment and adjust the strategy according to the reward function until the optimal evacuation path and behavior pattern are found; ； in, Indicates in status Take action of value, represents the learning rate, represents the discount factor, Indicates taking action The new state after In the new state The following actions are taken, Indicates updated value.

6. A simulation method for evacuating rural refugees based on a simulation model according to claim 5, characterized in that: In step S4, an environmental model is constructed according to the actual situation of the target rural area, and the resident agent, the refuge site agent, the refuge road agent and the environment dynamic change agent are embedded in it to form a complete rural refuge emergency agent simulation model. The specific operations are as follows: Integrate the impact factors collected in step S1 and perform processing operations, including data cleaning, data conversion and data verification; Using 3D modeling software to construct a geographical environment model of the target rural area based on the processed data, including topography, building structure, and road network; Based on the geographical environment model, layout of refuge sites and refuge roads is carried out; A set of interactive interfaces is designed for the interaction between each intelligent agent and the environmental model. The interactive interfaces include data interface, control interface and visualization interface. The data interface defines the format and protocol of data transmission between each intelligent agent and the environmental model. The control interface provides an interface for control commands for each intelligent agent. The visualization interface is used to display the real-time status of the environmental model and each intelligent agent. After completing the construction of the environmental model, the model is verified and adjusted, including logic verification, parameter adjustment and test run.

7. A simulation method for evacuating rural refugees based on a simulation model according to claim 6, characterized in that: In step S5, the interaction mechanism between the four types of intelligent agents includes: the interaction mechanism between resident intelligent agents and resident intelligent agents, the interaction mechanism between resident intelligent agents and refuge site intelligent agents, the interaction mechanism between resident intelligent agents and refuge road intelligent agents, the interaction mechanism between resident intelligent agents and environmental dynamic change intelligent agents, the interaction mechanism between refuge site intelligent agents and refuge road intelligent agents, the interaction mechanism between refuge site intelligent agents and environmental dynamic change intelligent agents, and the interaction mechanism between refuge road intelligent agents and environmental dynamic change intelligent agents.

8. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 7, characterized in that: In step S6, the simulation scenario is set according to the actual situation of the target village, and the specific operations are as follows: First, the simulation scenario parameters were set, including rural area characteristics, shelter site layout, resident distribution, shelter road network, emergency event type and occurrence time; Then, geographical data is imported into the simulation system to set the properties, location layout and interaction mechanism of each agent; Finally, set the simulation conditions, including simulation time length, data logging frequency, and output format.

9. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 8, characterized in that: In step S7, the rural emergency refuge agent simulation model is started to simulate the evacuation process of rural refugees in an emergency situation and record the performance of the agent, including the following steps: Load the rural emergency response agent simulation model constructed in step S4 and initialize all agents; Starting the engine of the rural emergency refuge agent simulation model, triggering the emergency time, and making the resident agent perceive the emergency according to the emergency event type and occurrence time set in step S6; Conduct simulation operations, including resident agent behavior simulation, refuge site agent behavior simulation, refuge road agent behavior simulation, and environment dynamic change agent behavior simulation; Data recording and monitoring, including recording the evacuation time, path selection and movement speed of resident intelligent bodies; recording the traffic conditions and travel time of the refuge roads; monitoring the capacity, utilization rate and possible congestion points of the refuge sites; The simulation ends when all resident agents arrive at the shelter or the preset simulation time is reached.

10. The method for simulating the evacuation of rural refugees based on a simulation model according to claim 9, characterized in that: In step S7, during the simulation process, an abnormal situation detection mechanism is also provided, and the detection mechanism is as follows: During the simulation, the system monitors the status of each agent, environmental changes, and traffic flow data in real time; Thresholds are set based on historical data to determine the occurrence of abnormal situations. When the analyzed real-time data exceeds the threshold, the abnormal situation detection mechanism is triggered; Calculate the Z-score of each indicator and determine the abnormal threshold: ； in, represents the observed value; represents the historical mean; represents the historical standard deviation.

Images