An emergency material reserve site selection optimization method for sudden public health events

Abstract

This application provides a method and system for optimizing the location of emergency material reserve depots in response to public health emergencies, belonging to the field of emergency logistics and operations optimization. The method includes collecting epidemic data, geographic information, and population data to generate a reachability matrix to identify controlled road sections; using system clustering to perform cluster analysis on each area to initially determine the number of reserve depots; constructing a multi-objective optimization location model with the objectives of minimizing weighted transportation distance and minimizing the comprehensive epidemic risk index, constrained by controlled road section avoidance, service radius, and capacity; and solving the model to obtain the optimal coordinates, jurisdiction, and delivery priority of the reserve depots. This invention introduces an epidemic risk coefficient based on real-time infection data, enabling the location scheme to dynamically respond to the epidemic situation and prioritize the supply of materials to high-risk areas; the controlled road section constraint ensures the operability of the scheme under lockdown scenarios, significantly reducing the risk of cross-infection during material transportation.

Description

Technical Field

[0001] This application relates to the field of emergency logistics and operations optimization technology, and in particular to a method for optimizing the location of emergency material reserve depots for public health emergencies. Background Technology

[0002] In recent years, public health emergencies have occurred frequently worldwide. During pandemic lockdowns, ensuring the supply of basic necessities for residents has become one of the core issues in emergency response. The location of emergency material reserve depots directly determines the efficiency of material distribution, the rationality of coverage, and the risk of personnel contact during transportation.

[0003] Traditional methods for selecting locations for material storage depots are mainly based on the average allocation of administrative regions or empirical judgment, which have the following technical drawbacks:

[0004] (1) Lack of response capability to dynamic risks of the epidemic: Traditional site selection methods only consider static population distribution and distance factors, and cannot dynamically adjust the site selection plan according to real-time epidemic data (such as the number of new infections and the number of people in isolation), resulting in insufficient material support capacity in high-risk areas;

[0005] (2) Path constraints under lockdown conditions are not considered: During the lockdown, a large number of roads are physically isolated or restricted, and the traditional location selection model based on the reach of the entire road network fails in the lockdown scenario;

[0006] (3) Lack of quantification of the risks of “last mile” delivery: Traditional methods mainly focus on transportation costs and ignore the risk of cross-infection that may be caused by the movement of people during the transportation of goods.

[0007] In existing technologies, such as CN114492953A, an optimization decision-making method for the deployment of emergency resources in response to public emergencies is disclosed. However, this method is mainly geared towards general public emergencies and is not specifically designed for the dynamic infection risks and road constraints under epidemic control scenarios. Therefore, there is an urgent need for an intelligent site selection method for emergency material reserve depots that can comprehensively consider the dynamic risks of the epidemic, road constraints under control, and population coverage requirements. Summary of the Invention

[0008] The purpose of this application is to address the technical problem that existing emergency resource layout optimization strategies are generally geared towards sudden public events and are not specifically designed for dynamic infection risks and road closure constraints under epidemic control scenarios.

[0009] A method for optimizing the site selection of emergency material reserve depots for public health emergencies includes the following steps:

[0010] S1: Data Acquisition and Preprocessing: Acquire epidemic data, geographic information, and population data for the target area;

[0011] S2: Determine the number of reserve depots based on epidemic risk: Taking each community as the demand point, use the system clustering method to perform cluster analysis on the coordinate data of each community, divide the community into several clusters according to the clustering results, calculate the total population and cumulative number of infections of all communities in each cluster; based on the preset maximum service population threshold and maximum service radius threshold of a single reserve depot, combined with the population size and infection risk of each cluster, preliminarily determine the number K of reserve depots to be built;

[0012] S3: Construct a multi-objective optimal location selection model;

[0013] S4: Model Solving and Result Output: Solve the multi-objective optimization site selection model to obtain the Pareto optimal solution set; select the solution with the highest overall satisfaction from the Pareto optimal solution set as the final site selection scheme; output the optimal coordinates, jurisdiction, total population covered, and delivery priority ranking of the K reserve warehouses.

[0014] Preferably, the data collected in S1 is as follows:

[0015] The epidemic data includes at least the real-time number of new infections and the cumulative number of people in quarantine for each community; the geographic information includes at least the coordinate data and road network data for each community; the population data includes at least the number of households and the number of people in each community; based on the road network data and combined with real-time lockdown management information, an accessibility matrix is ​​generated to identify passable road sections and locked-down road sections.

[0016] Preferably, in step S2, when performing cluster analysis using hierarchical clustering, the inter-cluster distance is Euclidean distance, expressed as: ,in, and The first The and the first The coordinates of each community.

[0017] Preferably, the model of S3 includes: a first objective function, which minimizes the weighted transportation distance from the reserve depot to its covered community; a second objective function, which minimizes the comprehensive epidemic risk index within the coverage area of ​​the reserve depot, wherein the comprehensive epidemic risk index is calculated by weighting the real-time new infection number and the cumulative number of quarantined people in each community; the constraints of the model include: each community is covered by at least one reserve depot, the transportation path from the reserve depot to the community must avoid the locked-down sections, and the service radius and service capacity constraints of the reserve depot.

[0018] Preferably, the expression for the first objective function in step S3 is: ,in, For the first The weighting coefficient of each community is obtained by normalizing the population of that community; For the first The candidate reserve to the first The actual reachability distance of each cell is determined based on the reachability matrix; A 0-1 variable, representing a cell Whether by reserve warehouse Serve; The total number of communities. This represents the total number of candidate reserves.

[0019] Preferably, the expression for the second objective function in step S3 is: ,in, For the first The COVID-19 risk coefficient of a community is calculated by weighting the real-time number of new infections and the cumulative number of people in quarantine in that community.

[0020] Preferably, the epidemic risk coefficient The calculation formula is: ,in, For the first Real-time number of new infections in each community For the first The cumulative number of people under quarantine in each community. and These are the weighting coefficients, and .

[0021] Preferably, the constraints in step S3 further include:

[0022] (1) Each community is covered by at least one reserve warehouse: , ;

[0023] (2) The transportation route from the storage warehouse to the community must avoid the controlled road sections, i.e. It takes a finite value only if the path is reachable; otherwise, it takes infinity.

[0024] (3) Service radius constraint: , ,in, The preset maximum service radius;

[0025] (4) Service capacity constraints: , ,in, For the first The population of each community For the first The maximum service capacity of each reserve depot.

[0026] Preferably, step S5 is also included: dynamic updating and re-optimization; based on the real-time changes in epidemic data, steps S1-S4 are periodically re-run to dynamically adjust the site selection plan for the reserve warehouse.

[0027] This application also provides a site selection optimization system for emergency material reserve depots for public health emergencies, including:

[0028] The data acquisition module is used to acquire epidemic data, geographic information, and population data for the target area;

[0029] The data processing module is used to preprocess the collected data and generate a reachability matrix based on road network data and real-time lockdown management information;

[0030] The cluster analysis module is used to perform cluster analysis on the coordinate data of each cell using the hierarchical clustering method to determine the number of reserves.

[0031] A model building module is used to build a multi-objective optimal location selection model as described in any one of claims 1-8;

[0032] The model solving module is used to solve the model;

[0033] The results output module is used to output the location coordinates, jurisdiction, total population covered, and delivery priority of the reserve warehouse;

[0034] The dynamic update module is used to trigger re-optimization based on changes in epidemic data.

[0035] Compared with the prior art, this application has at least the following beneficial effects:

[0036] This application introduces an epidemic risk coefficient based on real-time infection data into the site selection model, enabling the reserve warehouse site selection to dynamically respond to the development of the epidemic, prioritizing the supply of materials to high-risk areas, and reducing the risk of epidemic spread due to insufficient material distribution. It adapts to lockdown management scenarios: by generating a reachability matrix and introducing lockdown road segment constraints, the practical operability of the site selection scheme under road closure conditions is ensured, solving the problem of traditional site selection models failing in lockdown scenarios. By minimizing the comprehensive epidemic risk index, the risk of personnel contact during transportation is included in the optimization objective, minimizing the risk of cross-infection while ensuring the supply of materials. A two-stage strategy of "clustering first, then optimization" is adopted, ensuring the scientific nature of the site selection and enhancing the operability of the scheme by outputting the jurisdiction and delivery priority. Attached Figure Description

[0037] Figure 1 This is a logic diagram for cluster analysis;

[0038] Figure 2This is a cluster analysis phylogenetic diagram of 168 communities in Kuancheng District in this embodiment of the invention;

[0039] Figure 3 This is a schematic diagram showing the clustering results and reserve site selection for Kuancheng District and the other 8 districts of Changchun City in an embodiment of the present invention. Detailed Implementation

[0040] This application discloses a method for optimizing the location of emergency material reserve depots for public health emergencies, the steps of which are as follows:

[0041] S1: Data Acquisition and Preprocessing

[0042] Acquire epidemic data, geographic information, and population data for the target area; the epidemic data includes at least the real-time number of new infections and the cumulative number of people in quarantine for each community; the geographic information includes at least the coordinate data and road network data for each community; the population data includes at least the number of households and the number of people in each community; based on the road network data and combined with real-time lockdown management information, generate an accessibility matrix to identify passable road sections and locked-down road sections;

[0043] S2: Determine the number of reserve warehouses based on the risk of the epidemic.

[0044] Please see Figure 1 Each community is taken as a demand point, and the system clustering method is used to perform cluster analysis on the coordinate data of each community. Based on the clustering results, the community is divided into several clusters, and the total population and cumulative number of infections of all communities in each cluster are calculated. Based on the preset maximum service population threshold and maximum service radius threshold of a single reserve warehouse, combined with the population size and infection risk of each cluster, the number K of reserve warehouses to be built is initially determined.

[0045] In one implementation, when using hierarchical clustering for cluster analysis, the inter-cluster distance is expressed as Euclidean distance, as follows:

[0046] ,in, and The first The and the first The coordinates of each community.

[0047] S3: Constructing a multi-objective optimal location selection model

[0048] A site selection model based on multi-objective programming is established. The model includes: a first objective function, which minimizes the weighted transportation distance from the reserve depot to its covered community; and a second objective function, which minimizes the comprehensive epidemic risk index within the coverage area of ​​the reserve depot, wherein the comprehensive epidemic risk index is calculated by weighting the real-time new infection number and the cumulative number of quarantined people in each community. The constraints of the model include: each community is covered by at least one reserve depot; the transportation path from the reserve depot to the community must avoid the locked-down sections; and constraints on the service radius and service capacity of the reserve depot.

[0049] In one embodiment, the expression for the first objective function is: ,in, For the first The weighting coefficient of each community is obtained by normalizing the population of that community; For the first The candidate reserve to the first The actual reachability distance of each cell is determined based on the reachability matrix; A 0-1 variable, representing a cell Whether by reserve warehouse Serve; The total number of communities. This represents the total number of candidate reserves.

[0050] The expression for the second objective function is: ,in, For the first The COVID-19 risk coefficient of a community is calculated by weighting the real-time number of new infections and the cumulative number of people in quarantine in that community.

[0051] The epidemic risk coefficient The calculation formula is: ,in, For the first Real-time number of new infections in each community For the first The cumulative number of people under quarantine in each community. and These are the weighting coefficients, and .

[0052] In one embodiment, the constraint further includes:

[0053] (1) Each community is covered by at least one reserve warehouse: , ;

[0054] (2) The transportation route from the storage warehouse to the community must avoid the controlled road sections, i.e. It takes a finite value only if the path is reachable; otherwise, it takes infinity.

[0055] (3) Service radius constraint: , ,in, The preset maximum service radius;

[0056] (4) Service capacity constraints: , ,in, For the first The population of each community For the first The maximum service capacity of each reserve depot.

[0057] S4: Model Solving and Result Output

[0058] The multi-objective optimization location model is solved to obtain the Pareto optimal solution set; the solution with the highest overall satisfaction is selected from the Pareto optimal solution set as the final location scheme; the optimal coordinates, jurisdiction, total population covered, and delivery priority ranking of the K reserve warehouses are output.

[0059] S5: Dynamic updates and re-optimization; Based on real-time changes in epidemic data, regularly trigger the re-run of steps S1-S4 to dynamically adjust the site selection plan for the reserve warehouse.

[0060] Based on the methods described above, this application also provides an emergency material reserve site selection optimization system for public health emergencies, which includes a data acquisition module for acquiring epidemic data, geographic information and population data of the target area;

[0061] The data processing module is used to preprocess the collected data and generate a reachability matrix based on road network data and real-time lockdown management information;

[0062] The cluster analysis module is used to perform cluster analysis on the coordinate data of each cell using the hierarchical clustering method to determine the number of reserves.

[0063] The model building module is used to build the multi-objective optimization location model described above.

[0064] The model solving module is used to solve the model;

[0065] The results output module is used to output the location coordinates, jurisdiction, total population covered, and delivery priority of the reserve warehouse.

[0066] The dynamic update module is used to trigger re-optimization based on changes in epidemic data.

[0067] The above content will be explained in conjunction with specific verification experiments:

[0068] Example 1: This example uses Kuancheng District of Changchun City as an example to illustrate the implementation process of the method of the present invention in detail.

[0069] Step S1: Data Acquisition and Preprocessing

[0070] Coordinate data, population data, and COVID-19 data from March 4th to March 25th, 2022, were collected from 168 residential communities in Kuancheng District. This included daily new infections and cumulative quarantine numbers. Based on road network data and the lockdown management information at the time, a reachability matrix was generated to identify passable and locked-down road sections. Data preprocessing included imputation of missing data and removal of outliers.

[0071] Step S2: Determining the number of reserve warehouses based on epidemic risk

[0072] Using 168 communities as demand points, hierarchical clustering was employed for cluster analysis. Euclidean distance was used to calculate the inter-cluster distance, with the following formula: .

[0073] Figure 2 This paper presents a clustering tree diagram of 168 residential communities in Kuancheng District. Based on the clustering tree diagram, the communities in Kuancheng District are divided into 6 categories, and the total population and cumulative number of infections in all communities within each category are calculated. Based on the preset maximum service population threshold (set to 50,000 people) and maximum service radius threshold (set to 8 units of distance) for a single reserve warehouse, combined with the population size and infection risk of each cluster, the number of reserve warehouses to be built in Kuancheng District is determined to be K=6. Taking 26 communities in category 1 as an example, the weights of each community are given, as shown in Table 1. Finally, the warehouse location is determined by solving the mathematical model.

[0074] Table 1. Weights of each residential community in Kuancheng District

[0075]

[0076]

[0077]

[0078]

[0079] Step S3: Construct a multi-objective optimal location selection model

[0080] The multi-objective programming location selection model is established as follows:

[0081] First objective function (minimize weighted transport distance):

[0082] Second objective function (minimizing the comprehensive epidemic risk index): Among them, the cell weight Obtained by normalizing the population; actual reachable distance The reachability matrix generated in step S1 is used to calculate a finite value that is only taken when the path is reachable; the epidemic risk coefficient. The calculation is as follows: The weighting coefficients are taken as follows: , This demonstrates that the number of newly infected individuals is more important than the cumulative number of people in quarantine.

[0083] The constraints include:

[0084] (1) Each community is covered by at least one reserve warehouse: , ;

[0085] (2) The transportation route must avoid the blocked sections; that is, if the route is inaccessible, then... ;

[0086] (3) Service radius constraint: ;

[0087] (4) Service capacity constraints: , .

[0088] Step S4: Model Solving and Result Output

[0089] Based on the data in the table, the x-axis of the center of the large-scale material storage warehouse in Class 1 residential areas of Kuancheng District can be obtained. y-axis The warehouse governs 26 communities with a population of 78,092 within its jurisdiction, and has a location radius of 6.718091532 units.

[0090] Ultimately, the locations of the six large-scale material storage warehouse centers in Kuancheng District were obtained. Using the same method, the number and locations of the large-scale material storage warehouse centers in the other eight districts of Changchun City were obtained, as shown in Table 2.

[0091] Table 2 Specific locations of warehouses in each district

[0092]

[0093] Figure 3 The map shows the locations of the reserve warehouses in the nine districts of Changchun City and the distribution of their respective jurisdictions. Black dots represent the locations of the reserve warehouses, and red dots represent the jurisdictions of the covered areas. Specifically, A is the warehouse location in Kuancheng District, B is the warehouse location in Erdao District, C is the warehouse location in Chaoyang District, D is the warehouse location in Lvyuan District, E is the warehouse location in Nanguan District, F is the warehouse location in Jingkai District, G is the warehouse location in Changchun New Area, H is the warehouse location in Jingyue District, and I is the warehouse location in Qikai District.

[0094] Step S5: Dynamic Updates and Re-optimization

[0095] The system is set to trigger a re-optimization once a week. When the number of new infections changes significantly (such as the daily increase exceeding a threshold), the re-optimization is triggered immediately to dynamically adjust the location plan of the reserve warehouse.

[0096] Example 2

[0097] This embodiment provides an emergency material reserve site selection optimization system for public health emergencies, including:

[0098] The data acquisition module is used to acquire epidemic data, geographic information, and population data for the target area;

[0099] The data processing module is used to preprocess the collected data and generate a reachability matrix;

[0100] The cluster analysis module is used to perform cluster analysis on each cell using hierarchical clustering to determine the number of reserves.

[0101] The model building module is used to build the multi-objective optimal location selection model.

[0102] The model solving module is used to solve the model using a multi-objective evolutionary algorithm;

[0103] The results output module is used to output the location coordinates, jurisdiction, total population covered, and delivery priority of the reserve warehouse;

[0104] The dynamic update module is used to trigger re-optimization based on changes in epidemic data.

[0105] The specific implementation methods of each module correspond to the method steps in Example 1, and will not be repeated here.

[0106] This application introduces an epidemic risk coefficient based on real-time infection data into the site selection model, enabling the reserve warehouse site selection to dynamically respond to the development of the epidemic, prioritizing the supply of materials to high-risk areas, and reducing the risk of epidemic spread due to insufficient material distribution. It adapts to lockdown management scenarios: by generating a reachability matrix and introducing lockdown road segment constraints, the practical operability of the site selection scheme under road closure conditions is ensured, solving the problem of traditional site selection models failing in lockdown scenarios. By minimizing the comprehensive epidemic risk index, the risk of personnel contact during transportation is included in the optimization objective, minimizing the risk of cross-infection while ensuring the supply of materials. A two-stage strategy of "clustering first, then optimization" is adopted, ensuring the scientific nature of the site selection and enhancing the operability of the scheme by outputting the jurisdiction and delivery priority.

Claims

1. A method for optimizing the location of an emergency material reserve warehouse for a sudden public health event, characterized in that: Includes the following steps: S1: Data Acquisition and Preprocessing: Acquire epidemic data, geographic information, and population data for the target area; S2: Determine the number of reserve depots based on epidemic risk: Taking each community as the demand point, use the system clustering method to perform cluster analysis on the coordinate data of each community, divide the community into several clusters according to the clustering results, calculate the total population and cumulative number of infections of all communities in each cluster; based on the preset maximum service population threshold and maximum service radius threshold of a single reserve depot, combined with the population size and infection risk of each cluster, preliminarily determine the number K of reserve depots to be built; S3: Construct a multi-objective optimal location selection model; S4: Model Solving and Result Output: Solve the multi-objective optimization site selection model to obtain the Pareto optimal solution set; select the solution with the highest overall satisfaction from the Pareto optimal solution set as the final site selection scheme; output the optimal coordinates, jurisdiction, total population covered, and delivery priority ranking of the K reserve warehouses.

2. The method of claim 1, wherein the method is characterized by: The data collected in S1 is as follows: The epidemic data includes at least the real-time number of new infections and the cumulative number of people in quarantine for each community; the geographic information includes at least the coordinate data and road network data for each community; the population data includes at least the number of households and the number of people in each community; based on the road network data and combined with real-time lockdown management information, an accessibility matrix is ​​generated to identify passable road sections and locked-down road sections.

3. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 1, characterized in that: The Euclidean distance is used for the distance between classes in the system clustering method in the step S2, and the expression is as follows: wherein, and are the coordinates of the first and the first cell, respectively.

4. The method of claim 1, wherein the method further comprises: The model of S3 includes: a first objective function, which minimizes the weighted transportation distance from the reserve depot to its covered communities; a second objective function, which minimizes the comprehensive epidemic risk index within the coverage area of ​​the reserve depot, wherein the comprehensive epidemic risk index is calculated by weighting the real-time new infection number and the cumulative number of people in isolation in each community; the constraints of the model include: each community is covered by at least one reserve depot, the transportation path from the reserve depot to the community must avoid the locked-down sections, and the service radius and service capacity constraints of the reserve depot.

5. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 4, characterized in that: The expression for the first objective function in step S3 is: ,in, For the first The weighting coefficient of each community is obtained by normalizing the population of that community; For the first The candidate reserve to the first The actual reachability distance of each cell is determined based on the reachability matrix; A 0-1 variable, representing a cell Whether by reserve warehouse Serve; The total number of communities. This represents the total number of candidate reserves.

6. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 4, characterized in that: The expression for the second objective function in step S3 is: ,in, For the first The COVID-19 risk coefficient of a community is calculated by weighting the real-time number of new infections and the cumulative number of people in quarantine in that community.

7. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 6, characterized in that: The epidemic risk coefficient The calculation formula is: ,in, For the first Real-time number of new infections in each community For the first The cumulative number of people under quarantine in each community. and These are the weighting coefficients, and .

8. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 7, characterized in that: The constraints in step S3 also include: (1) Each community is covered by at least one reserve warehouse: , ; (2) The transportation route from the storage warehouse to the community must avoid the controlled road sections, i.e. It takes a finite value only if the path is reachable; otherwise, it takes infinity. (3) Service radius constraint: , ,in, The preset maximum service radius; (4) Service capacity constraints: , ,in, For the first The population of each community For the first The maximum service capacity of each reserve warehouse.

9. The method for optimizing the site selection of emergency material reserve depots for public health emergencies according to claim 1, characterized in that: It also includes step S5: dynamic updating and re-optimization; based on the real-time changes in epidemic data, steps S1-S4 are periodically re-run to dynamically adjust the site selection plan for the reserve warehouse.

10. A site selection optimization system for emergency material reserve depots in response to public health emergencies, characterized in that: The data acquisition module is used to acquire epidemic data, geographic information, and population data for the target area; The data processing module is used to preprocess the collected data and generate a reachability matrix based on road network data and real-time lockdown management information; The cluster analysis module is used to perform cluster analysis on the coordinate data of each cell using the hierarchical clustering method to determine the number of reserves. A model building module is used to build a multi-objective optimal location selection model as described in any one of claims 1-8; The model solving module is used to solve the model; The results output module is used to output the location coordinates, jurisdiction, total population covered, and delivery priority of the reserve warehouse; The dynamic update module is used to trigger re-optimization based on changes in epidemic data.

Images