Village refuge space optimization method based on Monte Carlo method and agent model

Abstract

The invention provides a village refuge space optimization method based on a Monte Carlo method and an agent model, and relates to the field of space optimization, comprising the following steps: collecting and arranging multiple types of space data; the attributes of the agents serve as input of a simulation model, random sampling is conducted on the attribute range of the agents through a Monte Carlo method, analog simulation is conducted in the simulation model, constraint conditions are set, and optimal model output is obtained; the intelligent agent attributes corresponding to the optimal model output are analyzed to obtain an optimal shelter space optimization scheme; and setting a visual model, displaying the simulation process through the visual model, and outputting a simulation process video output to the optimal village refuge space optimization scheme. According to the village refuge space optimization method based on the Monte Carlo method and the intelligent agent model, the optimization effect and the use efficiency of the village refuge space can be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of space optimization, and in particular to a village refuge space optimization method based on Monte Carlo method and intelligent agent model. Background Art

[0002] The Monte Carlo method is a simulation method based on random numbers that can handle complex random processes and uncertainty problems. It has a wide range of applications in finance, physics, engineering, computer science and other fields. Its flexibility and robustness make it a powerful tool for solving complex problems. In the optimization of refuge space, the Monte Carlo method can be used to simulate the evacuation process and refuge effect under different scenarios, providing a scientific basis for the optimization plan.

[0003] The agent model can simulate the behavioral characteristics of people when disasters occur, such as evacuation speed, path selection, etc. By building environmental agents and crowd agents, the interaction between people and the environment can be simulated, such as the impact of road capacity on evacuation speed. The simulation results based on the agent model can more accurately predict the evacuation process and shelter effect, providing accurate data support for the optimization plan.

[0004] With the rapid development of computer technology and artificial intelligence technology, the Monte Carlo method and agent model are increasingly used in disaster response and shelter space optimization. The maturity of these technologies provides strong technical support for the implementation of the program.

[0005] In the field of village refuge space optimization, traditional planning methods often fail to fully consider various complex factors and uncertainties. The application of Monte Carlo method and intelligent agent model can make up for this deficiency and improve the scientificity and feasibility of the optimization scheme. Summary of the invention

[0006] The purpose of the present invention is to provide a village shelter space optimization method based on the Monte Carlo method and the intelligent agent model, which can significantly improve the optimization effect and utilization efficiency of the village shelter space through data-driven, model simulation, multi-round iterative optimization, multiple evaluation indicators, resource optimization, auxiliary decision-making and enhancing the community's disaster response capabilities.

[0007] To achieve the above object, the present invention provides a village refuge space optimization method based on Monte Carlo method and intelligent agent model, comprising the following steps: Collect and organize multiple types of spatial data, including road network layers, alternative shelter layers, and population distribution layers; Create a crowd agent based on the population distribution layer, and create an environmental agent of road grids and shelters based on the road network layer and the alternative shelter layer; Set attributes for each agent, define the attribute range of each agent, use the attributes of the agent as the input of the simulation model, use the Monte Carlo method to randomly sample the attribute range of the agent, simulate it in the simulation model, set constraints, and obtain the optimal model output; the agents include crowd agents and environmental agents; An evaluation model is set up to evaluate the simulation results, and the village shelter space optimization plan is updated according to the evaluation results. The agent attributes corresponding to the optimal model output are analyzed to obtain the optimal shelter space optimization plan; A visualization model is set up, and the simulation process is displayed through the visualization model. At the same time, the simulation process video of the optimal village shelter space optimization plan is output.

[0008] Preferably, the attribute characteristics of the crowd agent include location, speed, age, gender, and function; The behavioral rules of crowd agents include evacuation speed, path selection strategy, and preference for shelter selection; Attributes of the environment agent: road capacity, shelter capacity, location, and entrance location; The interaction between crowd agents and environmental agents includes entrance flow restriction and evacuation path guidance.

[0009] Preferably, the model output includes the agent's refuge time, the agent's average speed, and the covered area.

[0010] Preferably, the attributes of the agent are used as inputs of the simulation model, the Monte Carlo method is used to randomly sample the attribute range of the agent, simulation is performed in the simulation model, constraints are set, and the optimal model output is obtained, including: Establish a Monte Carlo tree, initialize the root node, and initialize the positions of crowd agents and environment agents; Setting up a simulation model, including a dynamics layer, a pathfinding layer, and a behavior layer, setting each layer in the simulation model as a Monte Carlo tree node, and the information in each node includes a state, a quality value, a number of visits, a parent node, and a child node; The states include the actions of the crowd agent and the environment agent at the dynamics layer, pathfinding layer, and behavior layer, respectively; Within the set total number of training times, the Monte Carlo tree is trained to obtain a trained Monte Carlo tree; according to the tree structure of the trained Monte Carlo tree, a range is set for each model output, the range of the model output is used as the training target, and the model output that reaches the training target each time is evaluated, and the solution with the highest evaluation index is obtained as the optimal simulation model based on the Monte Carlo method under the current training situation.

[0011] Preferably, a simulation model based on the Monte Carlo method is used to simulate the evacuation behavior and refuge process of the crowd when a disaster occurs: First, grid division is performed, and then the position and area of ​​obstacles in each sub-grid are identified to obtain the position and area of ​​blank spaces, and then the positions of adjacent blank spaces are spliced ​​to obtain the first path; Match each path in the first path with the basic function, and then solve the length of each path in the form of integration; ; in, s represents the length of the first path, represents the path fitting function, represents the starting point coordinates of the first path, represents the coordinates of the end point of the first path; The shelter is set as the end point, and different crowd gathering places are set as the starting point. Based on the first path obtained, all path planning solutions from the starting point to the end point are obtained; According to the length of each solved path, the length of the paths in all path planning schemes is calculated, all the paths in the path planning schemes are sorted in ascending order according to the path length, and the path with the shortest path length is selected as the refuge path for the current crowd gathering point.

[0012] Preferably, when the crowd takes refuge along the refuge path, the density of the crowd in each path is counted, and the simulated test value is gradually corrected toward the expected density using the basic graph density correction algorithm: ; in, v is the calculated density velocity, v ( m ) is the free speed, u is the current density, u ( m ) is the maximum density, i.e. the density when the crowd is completely still; Indicates the current scenario.

[0013] Preferably, the correction scheme is as follows: Different carrying levels are set according to the number of people carried daily at the crowd gathering point: first carrying group, second carrying group, and third carrying group; The number of evacuation routes is increased according to different load levels: At the first level of people carrying capacity, an evacuation route is provided; At the second level of population carrying capacity, three refuge routes are set up; At the third level of population carrying capacity, six refuge routes are set up; In addition, different load-bearing levels are analyzed: When the sum of the number of people under 5 years old and over 60 years old in the carrying level accounts for more than 30% of the number of people in this carrying level and less than 60%, an additional refuge route shall be provided; When the sum of the number of people under 5 years old and over 60 years old within the carrying level accounts for more than 60% of the number of people in this carrying level, two additional evacuation routes will be added.

[0014] Preferably, an evaluation model is set to evaluate the simulation results, and the village refuge space optimization plan is updated according to the evaluation results, including From the simulation results, the time spent in the evacuation passage, the number of casualties, the time to enter the evacuation site, and the proportion of agents in the evacuation site were selected as evaluation indicators; ; in, w 1 is the weight of the agent’s refuge time, w 2 is the weight of the agent’s average speed, w 3 is the weight of the covered area, w 4 is the weight of the proportion of agents in the shelter during simulation, w 5 is the weight of casualties in simulation, w 1 + w 2 + w 3 + w 4 + w 5 =1; Indicates the time consumed by the refuge passage, represents the average speed of the agent, represents the model's environment, represents the proportion of agents in the shelter during simulation, Indicates the number of casualties during the simulation; The evaluation index is scored according to its weight and the actual simulation situation to obtain the score of the simulation result. When the score of the simulation result is equal to or higher than the set threshold, the optimization plan of the village refuge space is directly output; when the score of the simulation result is lower than the set threshold, each evaluation index is analyzed and the corresponding refuge space setting is adjusted according to the score of each evaluation index. The adjusted refuge space settings are used as the new village refuge space optimization plan, and the relevant data of the new village refuge space optimization plan are fed back to the intelligent agent model.

[0015] Therefore, the present invention adopts the above-mentioned village refuge space optimization method based on the Monte Carlo method and the intelligent agent model, and the technical effects are as follows: 1. Improve the scientific nature of refuge space planning Data-driven: By collecting and organizing various types of spatial data, including road networks, alternative shelters, and population distribution, a solid data foundation is provided for the optimization process. The accuracy and completeness of these data ensure the scientificity and feasibility of the optimization plan.

[0016] Model simulation: The Monte Carlo method and agent-based model are used for simulation, which can simulate the evacuation situation and refuge process under different scenarios, so as to more accurately predict and evaluate the effect of refuge space.

[0017] 2. Enhance the flexibility of optimization solutions Multiple rounds of iteration: Through the continuous iteration of the optimization process, the shelter space optimization plan can be adjusted in time according to the evaluation results until the optimal solution is found. This flexibility enables the optimization plan to better adapt to changes in actual conditions.

[0018] Multiple evaluation indicators: Set multiple evaluation indicators to comprehensively evaluate the effectiveness of the optimization plan, including evacuation time, shelter utilization rate, crowd safety, etc. These indicators can be weighted according to actual needs to reflect the importance of different aspects.

[0019] 3. Improve the efficiency of using shelter space Reasonable layout: By optimizing the location and capacity of shelters, as well as the traffic capacity of the road network, shelter space can be used more effectively when disasters occur. This helps reduce evacuation time and improve the safety of people.

[0020] Resource optimization: The optimization plan can ensure that limited resources (such as shelter capacity, rescue forces, etc.) are reasonably allocated and utilized, thereby improving the overall ability to respond to disasters.

[0021] 4. Assist decision making Visualization: Displaying the optimization plan to decision makers in a visual way helps them understand the optimization process and results more intuitively, which helps improve the efficiency and accuracy of decision making.

[0022] Decision support: Through detailed reports or presentations, the optimization process and results are introduced to relevant departments or the public to provide strong support for decision making.

[0023] 5. Strengthen community resilience to disasters Improve residents' awareness: By optimizing shelter space, community residents' awareness of disasters and their ability to respond can be enhanced. During the optimization process, publicity and education can be combined to improve residents' awareness of disaster prevention and reduction.

[0024] Promote community participation: The implementation of the optimization plan requires the active participation and cooperation of community residents. By strengthening community participation, a good atmosphere of co-construction and sharing can be formed, and the overall disaster response capacity of the community can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The present invention is a flow chart of the village refuge space optimization method based on the Monte Carlo method and the intelligent agent model. DETAILED DESCRIPTION

[0026] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0027] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0028] Embodiment 1 like Figure 1 As shown, the present invention provides a village refuge space optimization method based on the Monte Carlo method and the intelligent agent model, comprising the following steps: Collect and organize multiple types of spatial data, including road network layers, alternative shelter layers, and population distribution layers; Create a crowd agent based on the population distribution layer, and create an environmental agent of road grids and shelters based on the road network layer and the alternative shelter layer; The attribute characteristics of crowd agents include location, speed, age, gender, and function; The behavioral rules of crowd agents include evacuation speed, path selection strategy, and preference for shelter selection; Attributes of the environment agent: road capacity, shelter capacity, location, and entrance location; Among them, the interaction methods between crowd agents and environmental agents include entrance flow restriction and evacuation path guidance.

[0029] Set attributes for each agent, define the attribute range of each agent, use the attributes of the agent as the input of the simulation model, use the Monte Carlo method to randomly sample the attribute range of the agent, perform simulation in the simulation model, set constraints, and obtain the optimal model output; the agents include crowd agents and environmental agents; the model output includes the agent's evacuation time, the agent's average speed, and the covered area. Use the attributes of the agent as the input of the simulation model, use the Monte Carlo method to randomly sample the attribute range of the agent, perform simulation in the simulation model, set constraints, and obtain the optimal model output, including: Establish a Monte Carlo tree, initialize the root node, and initialize the positions of crowd agents and environment agents; Setting up a simulation model, including a dynamics layer, a pathfinding layer, and a behavior layer, setting each layer in the simulation model as a Monte Carlo tree node, and the information in each node includes a state, a quality value, a number of visits, a parent node, and a child node; The states include the actions of the crowd agent and the environment agent at the dynamics layer, pathfinding layer, and behavior layer, respectively; Within the set total number of training times, the Monte Carlo tree is trained to obtain a trained Monte Carlo tree; according to the tree structure of the trained Monte Carlo tree, a range is set for each model output, the range of the model output is used as the training target, and the model output that reaches the training target each time is evaluated, and the solution with the highest evaluation index is obtained as the optimal simulation model based on the Monte Carlo method under the current training situation.

[0030] The simulation model based on Monte Carlo method is used to simulate the evacuation behavior and refuge process of people when disasters occur: First, grid division is performed, and then the position and area of ​​obstacles in each sub-grid are identified to obtain the position and area of ​​blank spaces, and then the positions of adjacent blank spaces are spliced ​​to obtain the first path; Match each path in the first path with the basic function, and then solve the length of each path in the form of integration; ; in, s represents the length of the first path, represents the path fitting function, represents the starting point coordinates of the first path, represents the coordinates of the end point of the first path; The shelter is set as the end point, and different crowd gathering places are set as the starting point. Based on the first path obtained, all path planning solutions from the starting point to the end point are obtained; According to the length of each solved path, the length of the paths in all path planning schemes is calculated, all the paths in the path planning schemes are sorted in ascending order according to the path length, and the path with the shortest path length is selected as the refuge path for the current crowd gathering point.

[0031] When people take refuge along the evacuation paths, the density of people in each path is counted, and the basic graph density correction algorithm is used to gradually correct the simulated test values ​​toward the expected density: ; in, v is the calculated density velocity, v ( m ) is the free speed, u is the current density (expected density), u ( m ) is the maximum density, i.e. the density when the crowd is completely still; γ Indicates the current situation, for example, the values ​​for regular and casual situations are , the value in the commuting case is , the value in emergency situation is The basic graph density correction algorithm is an algorithm that uses a greedy strategy to improve the graph density algorithm. The basic graph density correction algorithm can better optimize the village environment in the simulation of the evacuation behavior and refuge process of the crowd when a disaster occurs, so as to obtain a more complete village construction plan, reduce the time for refugees to reach the refuge site, and reduce casualties.

[0032] The correction plan is as follows: Different carrying levels are set according to the number of people carried daily at the crowd gathering point: first carrying group, second carrying group, and third carrying group; The number of evacuation routes is increased according to different load levels: At the first level of people carrying capacity, an evacuation route is provided; At the second level of population carrying capacity, three refuge routes are set up; At the third level of population carrying capacity, six refuge routes are set up; In addition, different load-bearing levels are analyzed: When the sum of the number of people under 5 years old and over 60 years old in the carrying level accounts for more than 30% of the number of people in this carrying level and less than 60%, an additional refuge route shall be provided; When the sum of the number of people under 5 years old and over 60 years old within the carrying level accounts for more than 60% of the number of people in this carrying level, two additional evacuation routes will be added.

[0033] An evaluation model is set up to evaluate the simulation results, and the village shelter space optimization plan is updated according to the evaluation results. The agent attributes corresponding to the optimal model output are analyzed to obtain the optimal shelter space optimization plan; An evaluation model is set up to evaluate the simulation results, and the optimization plan for village refuge space is updated according to the evaluation results, including From the simulation results, the time spent in the evacuation passage, the number of casualties, the time to enter the evacuation site, and the proportion of agents in the evacuation site were selected as evaluation indicators; ; in, w 1 is the weight of the agent’s refuge time, w 2 is the weight of the agent’s average speed, w 3 is the weight of the covered area, w 4 is the weight of the proportion of agents in the shelter during simulation, w 5 is the weight of casualties in simulation, w 1 + w 2 + w 3 + w 4 + w 5 =1; Indicates the time consumed by the refuge passage, represents the average speed of the agent, represents the model's environment, represents the proportion of agents in the shelter during simulation, Indicates the number of casualties during the simulation; The evaluation index is scored according to its weight and the actual simulation situation to obtain the score of the simulation result. When the score of the simulation result is equal to or higher than the set threshold, the optimization plan of the village refuge space is directly output; when the score of the simulation result is lower than the set threshold, each evaluation index is analyzed and the corresponding refuge space setting is adjusted according to the score of each evaluation index. The adjusted refuge space settings are used as the new village refuge space optimization plan, and the relevant data of the new village refuge space optimization plan are fed back to the intelligent agent model.

[0034] A visualization model is set up, and the simulation process is displayed through the visualization model. At the same time, the simulation process video of the optimal village shelter space optimization plan is output.

[0035] The output video is transmitted to the corresponding village, which will play the video repeatedly in places where all villagers can see it, and organize villagers to repeatedly practice the evacuation route in the video, thereby improving the escape time and escape proficiency during evacuation and reducing the casualty rate during actual escape.

[0036] The present application also provides a storage medium on which a computer program is stored, and when the computer program is executed, the steps provided in the above embodiment can be implemented. The storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0037] The present application also provides an electronic device, which may include a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps provided in the above embodiment may be implemented. Of course, the electronic device may also include various network interfaces, power supplies and other components.

[0038] Therefore, the present invention adopts the above-mentioned village shelter space optimization method based on the Monte Carlo method and the intelligent agent model, which can significantly improve the optimization effect and utilization efficiency of the village shelter space through data-driven, model simulation, multi-round iterative optimization, multiple evaluation indicators, resource optimization, auxiliary decision-making, and enhancing the community's disaster response capabilities.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A village shelter space optimization method based on Monte Carlo method and agent model, characterized in that: The following steps are involved: Collect and organize multiple types of spatial data, including road network layers, alternative shelter layers, and population distribution layers; Create a crowd agent based on the population distribution layer, and create an environmental agent of road grids and shelters based on the road network layer and the alternative shelter layer; The agents include crowd agents and environmental agents; set attributes for each agent, define the attribute range of each agent, use the attributes of the agent as the input of the simulation model, use the Monte Carlo method to randomly sample the attribute range of the agent, perform simulation in the simulation model, set constraints, and obtain the optimal model output; An evaluation model is set up to evaluate the simulation results, and the village shelter space optimization plan is updated according to the evaluation results. The agent attributes corresponding to the optimal model output are analyzed to obtain the optimal shelter space optimization plan; A visualization model is set up, and the simulation process is displayed through the visualization model. At the same time, the simulation process video of the optimal village shelter space optimization plan is output.

2. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 1, characterized in that: The attribute characteristics of crowd agents include location, speed, age, gender, and function; The behavioral rules of crowd agents include evacuation speed, path selection strategy, and preference for shelter selection; Attributes of the environment agent: road capacity, shelter capacity, location, and entrance location; The interaction between crowd agents and environmental agents includes entrance flow restriction and evacuation path guidance.

3. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 2, characterized in that: The model output includes the agent’s refuge time, the agent’s average speed, and the area covered.

4. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 2, characterized in that: The attributes of the agent are used as the input of the simulation model, the Monte Carlo method is used to randomly sample the attribute range of the agent, the simulation is performed in the simulation model, and the constraints are set to obtain the optimal model output, including: Establish a Monte Carlo tree, initialize the root node, and initialize the positions of crowd agents and environment agents; Setting up a simulation model, including a dynamics layer, a pathfinding layer, and a behavior layer, setting each layer in the simulation model as a Monte Carlo tree node, and the information in each node includes a state, a quality value, a number of visits, a parent node, and a child node; The states include the actions of the crowd agent and the environment agent at the dynamics layer, pathfinding layer, and behavior layer, respectively; Within the set total number of training times, the Monte Carlo tree is trained to obtain a trained Monte Carlo tree; according to the tree structure of the trained Monte Carlo tree, a range is set for each model output, the range of the model output is used as the training target, and the model output that reaches the training target each time is evaluated, and the solution with the highest evaluation index is obtained as the optimal simulation model based on the Monte Carlo method under the current training situation.

5. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 4, characterized in that: The simulation model based on Monte Carlo method is used to simulate the evacuation behavior and refuge process of people when disasters occur: First, grid division is performed, and then the position and area of ​​obstacles in each sub-grid are identified to obtain the position and area of ​​blank spaces, and then the positions of adjacent blank spaces are spliced ​​to obtain the first path; Match each path in the first path with the basic function, and then solve the length of each path in the form of integration; ； in, s represents the length of the first path, represents the function of path fitting, represents the starting point coordinates of the first path, represents the coordinates of the end point of the first path; The shelter is set as the end point, and different crowd gathering places are set as the starting point. Based on the first path obtained, all path planning solutions from the starting point to the end point are obtained; According to the length of each solved path, the length of the paths in all path planning schemes is calculated, all the paths in the path planning schemes are sorted in ascending order according to the path length, and the path with the shortest path length is selected as the refuge path for the current crowd gathering point.

6. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 5, characterized in that: When people take refuge along the evacuation paths, the density of people in each path is counted, and the basic graph density correction algorithm is used to gradually correct the simulated test values ​​toward the expected density: ； in, v is the calculated density velocity, v ( m ) is the free speed, u is the current density, u ( m ) is the maximum density, i.e. the density when the crowd is completely still; Indicates the current scenario.

7. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 6, characterized in that: The correction plan is as follows: Different carrying levels are set according to the number of people carried daily at the crowd gathering point: first carrying group, second carrying group, and third carrying group; The number of evacuation routes is increased according to different load levels: At the first level of people carrying capacity, an evacuation route is provided; At the second level of population carrying capacity, three refuge routes are set up; At the third level of population carrying capacity, six refuge routes are set up; In addition, different load-bearing levels are analyzed: When the sum of the number of people under 5 years old and over 60 years old in the carrying level accounts for more than 30% of the number of people in this carrying level and less than 60%, an additional refuge route shall be provided; When the sum of the number of people under 5 years old and over 60 years old within the carrying level accounts for more than 60% of the number of people in this carrying level, two additional evacuation routes will be added.

8. The method for optimizing village refuge space based on Monte Carlo method and agent model according to claim 1, characterized in that: An evaluation model is set up to evaluate the simulation results, and the optimization plan for village refuge space is updated according to the evaluation results, including From the simulation results, the time spent in the evacuation passage, the number of casualties, the time to enter the evacuation site, and the proportion of agents in the evacuation site were selected as evaluation indicators; ； in, w 1 is the weight of the agent’s evacuation time, w 2 is the weight of the average speed of the agent, w 3 is the weight of the covered area, w 4 is the weight of the proportion of agents in the shelter during simulation, w 5 is the weight of casualties in simulation, Indicates the time consumed by the refuge passage, represents the average speed of the agent, represents the model's environment, represents the proportion of agents in the shelter during simulation, Indicates the number of casualties during the simulation; The evaluation index is scored according to its weight and the actual simulation situation to obtain the score of the simulation result. When the score of the simulation result is equal to or higher than the set threshold, the optimization plan of the village refuge space is directly output; when the score of the simulation result is lower than the set threshold, each evaluation index is analyzed and the corresponding refuge space setting is adjusted according to the score of each evaluation index. The adjusted refuge space settings are used as the new village refuge space optimization plan, and the relevant data of the new village refuge space optimization plan are fed back to the intelligent agent model.

Images