Prevention and control resource scheduling method based on group dynamic deduction mechanism

Abstract

The invention discloses a medical prevention and control resource scheduling method based on a group dynamic deduction mechanism, relates to the field of medical prevention and control resource scheduling, and aims to solve the problem that the medical prevention and control resource scheduling is not required under the conditions of meeting resource balance constraint and node resource demand constraint for a highly infectious outbreak region. Reasonable resource allocation needs to be carried out on each node by taking day as a period, so that the prevention and control resource usage amount, the prevention and control duration and the number of infected people are minimized. For the problem, the invention provides a method based on a group dynamic deduction mechanism, which covers the following aspects: abstracting characteristics and models of group dynamic deduction by integrating factors such as a continuous improvement mechanism of group dynamic deduction, matrix sequence coding of resources and the like. Based on the model, the invention provides an algorithm for processing the medical prevention and control resource scheduling problem. The algorithm design comprises individual self-deduction and collaborative deduction based on group dynamic deduction, continuous improvement processing and matrix sequence coding.

Description

Technical Field

[0001] The present invention relates to the field of medical prevention and control resource scheduling, and in particular to a prevention and control resource scheduling method based on a group dynamic deduction mechanism. Background Art

[0002] The problem of resource scheduling for prevention and control is one of the important challenges facing today's medical system, which involves how to efficiently allocate limited medical resources to meet the needs of patients. The problem of resource scheduling for prevention and control is an NP (non-deterministic polynomial time hard) problem. The methods for solving such problems are mainly divided into approximate algorithms and heuristic algorithms.

[0003] Approximate algorithms are algorithms designed for certain computational problems, especially those that are difficult to find exact solutions to. The main goal of this type of algorithm is to find a feasible solution that is close to the optimal solution in a reasonable amount of time. Approximate algorithms are often used to solve combinatorial optimization problems, such as the traveling salesman problem, the knapsack problem, and the graph coloring problem. These problems are characterized by a huge solution space, and exhaustive search is almost impossible.

[0004] Heuristic algorithms do not guarantee to find a near-optimal solution, but rather seek a "good enough" solution by exploiting the specific properties and structure of the problem. Such algorithms rely on empirical rules and heuristic strategies, and aim to explore the solution space through intelligent search. Heuristic algorithms can quickly converge to an acceptable solution when faced with complex problems. Summary of the invention

[0005] The present invention provides a medical prevention and control resource scheduling method based on a group dynamic deduction mechanism, which includes a continuous improvement mechanism of group dynamic deduction, individual self-deduction and collaborative deduction, and matrix sequence coding.

[0006] In order to solve the above-mentioned technical problems, the present invention provides a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism, which comprises the following steps:

[0007] Step 1: Define the dynamic spatiotemporal contact network of contacts, set each individual as a node, and determine the activity space of each node that changes over time;

[0008] Step 2: Matrix sequence encoding; the nodes in the dynamic spatiotemporal contact network and the prevention and control materials are used as the two dimensions of the matrix, the distribution quantity of the materials is used as the element value in the matrix sequence encoding, and the entire sequence is arranged in sequence according to the prevention and control period T;

[0009] Step 3: Read geographic information grid and prevention and control resource information;

[0010] Step 4: Set the population size to P, and generate a population P containing |P| individuals based on the initial solution. Each individual in the population corresponds to a matrix sequence code.

[0011] Step 5: Calculate the objective function of the minimum use of prevention and control resources, the minimum duration of prevention and control, and the minimum number of infections for each individual;

[0012] Step 6: Set the number of iterations i = 1 and start iteration;

[0013] Step 7: Implement the individual self-deduction operation: randomly select a fragment of the matrix sequence encoding of the individual to update, and perform the process α times;

[0014] Step 8: Implement individual collaborative deduction operation: randomly select two individuals, exchange a certain fragment of the matrix sequence encoding corresponding to the two individuals, and perform the process β times;

[0015] Step 9: Provide new individuals based on the continuous improvement mechanism of group dynamic deduction; calculate the objective function of each individual, sort the individuals according to the results, eliminate the last-ranked individuals, introduce new individuals, and repeat the set number of times;

[0016] The introduction of new individuals includes individual self-deduction operation and / or individual collaborative deduction operation and / or group dynamic deduction; the group dynamic deduction operation specifically includes randomly selecting T individuals, selecting fragments for the T individuals, and the sequences of the fragments selected by each individual do not overlap, and after obtaining T fragments, they are combined to form a new individual;

[0017] Step 10: i++, restart from step 7;

[0018] Step 11: If the number of iterations i = τ, where τ represents the maximum number of deductions, stop the calculation, otherwise go to step 7;

[0019] Step 12: Use the final resource allocation plan to schedule medical prevention and control resources.

[0020] Furthermore, the geographic information grid includes the contact time and closeness between nodes, and the prevention and control resource information includes the total number of resources, the maximum number of resources that can be allocated to each node, and the total number of nodes.

[0021] Furthermore, the specific process of generating the initial solution is as follows:

[0022] First, input the set of nodes, the set of consumable materials, the set of occupied materials, and the set of prevention and control cycles. Take one day as the unit for a single prevention and control cycle. Then, for the nodes in the current cycle, calculate the quantity of consumable materials and occupied materials to be allocated. If all nodes have recovered, there is no need to allocate materials anymore; otherwise, allocate random types and random quantities of epidemic prevention materials for each node and increase the duration of prevention and control. Finally, return the result of the allocation of epidemic prevention resources as the initial solution.

[0023] Further, the specific steps of the individual self-deduction operation are as follows:

[0024] Step 7.1: Randomly select an individual p from the population Ρ.

[0025] Step 7.2: Randomly select the nth row and mth column in the matrix sequence encoding of individual p; 0 < n < the total number of prevention and control resources, 0 < m < the total number of nodes.

[0026] Step 7.3: Recalculate the quantity of consumable materials and occupied materials to be allocated for the segment composed of m * n.

[0027] Step 7.4: Replace the original segment of m * n with the newly generated segment of m * n to form a new individual p'.

[0028] Step 7.5: Calculate the objective function of the new individual p'. If the objective function of p' can dominate the scores of some individuals in the population P, then replace the worst one of these individuals with p'; otherwise, repeat the self-deduction operation until a new individual p' that is better than some individuals in the population P is found, and replace the worst one of these individuals with p' or reach the upper limit of the execution times.

[0029] Step 7.6: Repeat steps 7.1 - 7.5 α times.

[0030] Further, the specific steps of the individual collaborative deduction operation are as follows:

[0031] Step 8.1: Randomly select two individuals q1 and q2 from Ρ.

[0032] Step 8.2: Randomly select the number of rows a (0 < a < the total number of prevention and control resources) and the number of columns b (0 < b < the total number of nodes) in the matrix sequence encoding.

[0033] Step 8.3: Exchange the segments composed of a * b in the matrix sequence encodings of q1 and q2 to form two new individuals q1' and q2'.

[0034] Step 8.4: Calculate the objective functions of the new individuals q1' and q2'. If the objective functions of q1' and q2' can dominate the scores of some individuals in the population P, then use q1' and q2' to replace the worst of these individuals respectively. Otherwise, repeat the collaborative deduction operation until a new individual q' that is better than some individuals in the population P is found, and use q' to replace the worst of these individuals or the upper limit of the number of executions is reached;

[0035] Step 8.5: Repeat steps 8.1-8.3 for β times.

[0036] Furthermore, the step 9 is specifically as follows:

[0037] Step 9.1: Let r = i*θ / (α+β), θ is an artificially adjusted parameter. If r is an integer and the number of individuals in the group is greater than the initial number P, then execute steps 9.2 to 9.10, that is, remove the worst individual and add new individuals;

[0038] Step 9.2: Calculate the objective function for each individual in the population P, and sort each individual from high to low according to the quality of the objective function;

[0039] Step 9.3: Remove r individuals from the end of the sorted population P set;

[0040] Step 9.4: Generate r / 2 new individuals using individual self-deduction operation and individual collaborative deduction, and add them to the population P;

[0041] Calculate the objective function and sort each individual from high to low according to the quality of the objective function; Step 9.6: Perform group dynamics deduction on the population P;

[0042] Step 9.7: Starting from the first one in the sorted population P set, select T individuals in order;

[0043] Step 9.8: For the T individuals in the group dynamics deduction, select a subsequence in the matrix sequence encoding of each individual in order, that is, the first individual takes the 1 / Tth subsequence, the second individual takes the 2 / Tth subsequence, and so on, the λth individual takes the λ / Tth subsequence, and the T subsequences are fused to form a new individual t;

[0044] Step 9.9: Repeat step 9.8 r / 2 times;

[0045] Step 9.10: Add the newly generated r / 2 individuals to the population P.

[0046] The beneficial effects of the present invention are as follows:

[0047] Compared with the traditional population evolution calculation method, the present invention mainly improves the speed and quality of the limited resource scheduling optimization algorithm.

[0048] In terms of speed: This invention significantly improves the algorithm's operation speed and increases the frequency of group iterations through innovative non-dominated solution set screening and sorting methods. At the same time, a group dynamic deduction mechanism is introduced, which can continuously introduce excellent individuals and eliminate inferior individuals, effectively preventing the algorithm from converging to the local optimal solution too early, thereby enhancing the overall performance of the algorithm. At the same time, compared with traditional multi-objective optimization algorithms such as NSGA-II, SPEA-II and MOEA / D, the time complexity of this algorithm is significantly lower.

[0049] In terms of quality: The present invention designs a new algorithm - group dynamic deduction algorithm, designs individual self-deduction and collaborative deduction, continuous improvement mechanism and matrix sequence coding. The group dynamic deduction method is adopted, and the prevention and control resources are optimized and supplemented through the continuous improvement mechanism, so as to achieve rapid iteration of the algorithm, continuously discard the solutions with lower quality, and provide the motivation for group deduction by introducing the resources of high-quality gene fragments, so as to maintain the diversity and innovation motivation of the group. The Parote solution set finally obtained also achieved significant advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The figure is a flow chart of the group deduction mechanism process of the method of the present invention.

[0051] Figure 2 Schematic diagram of the matrix sequence encoding method of the present invention.

[0052] Figure 3 The figure is a flow chart of the initial solution generation process of the method of the present invention.

[0053] Figure 4 Schematic diagram of self-deduction in the group deduction mechanism of the method of the present invention.

[0054] Figure 5 Schematic diagram of collaborative deduction in the group deduction mechanism of the method of the present invention.

[0055] Figure 6 This is a line chart of the IGD index of the algorithm of the method of the present invention compared with other algorithms.

[0056] Figure 7 This is a line chart of the HV index of the algorithm of the method of the present invention compared with other algorithms. DETAILED DESCRIPTION

[0057] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings. The following embodiments or drawings are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0058] The prevention and control resource scheduling method based on the group dynamic deduction mechanism includes the following contents: The group dynamic deduction mechanism proposed by the present invention requires the entire population to continuously perform self-deduction and collaborative deduction, and each deduction will generate different types of high-quality individuals in the population. In order to further accelerate the population convergence speed and obtain better individuals, it is also necessary to continuously improve the mechanism. First, a part of the relatively inferior individuals in the population are eliminated, and then a matrix subsequence is intercepted from each of the remaining high-quality individuals to form new individuals to join the population. This ensures that the encoding of individuals with excellent performance is intercepted, so that the group can continuously generate better individuals in iteration.

[0059] Conventional multi-objective optimization algorithms such as NSGA-II and MOEA / D have problems such as insufficient evolutionary power, slow convergence speed, and insufficient diversity of solutions. The group dynamic deduction mechanism of the present invention timely eliminates solutions with insufficient potential and supplements new solutions with excellent gene fragments, and ensures the diversity of the population through a continuous improvement mechanism, so that individuals with excellent performance can be retained to the greatest extent, while timely introducing high-quality individuals to enrich the diversity of the population. The convergence effect is quickly achieved through a fast and sensitive mechanism.

[0060] The present invention provides a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism. Figure 1 Flow chart of the group deduction mechanism process of the method of the present invention, such as Figure 1 As shown, the method includes:

[0061] Step 1: Define the dynamic spatiotemporal contact network of contacts and determine the activity space of each node (i.e., individual) that changes over time;

[0062] Step 2: Matrix sequence encoding: Use a matrix to represent the nodes in the dynamic spatiotemporal contact network and the prevention and control materials as the two dimensions of the matrix. The distribution quantity of the materials is used as the element value in the matrix sequence encoding. The entire sequence is arranged in sequence according to the prevention and control period T; Figure 2 The schematic diagram of the matrix sequence encoding method of the method of the present invention is shown. In order to ensure the diversity of the population, the initial population of the group dynamic deduction algorithm is obtained by random generation. Corresponding to the matrix sequence encoding, the element value is randomly generated. In the matrix sequence, the horizontal axis of each matrix is ​​the individual number, the vertical axis is the prevention and control resources, and the element value is the specific allocation number of each material. The entire sequence is arranged in sequence according to the prevention and control cycle.

[0063] Step 3: Read the geographic information grid and prevention and control resource information. The geographic information grid includes the contact time and closeness between nodes. The prevention and control resource information includes the total number of resources, the maximum number of resources that can be allocated to each node, and the total number of nodes.

[0064] Step 4: Set the population size to P, and generate a population P containing |Ρ| individuals according to the initial solution generation operator. The encoding of each individual in the population corresponds to a matrix sequence encoding; as Figure 3 The flowchart of the initial solution generation process of the method of the present invention shown. First, it is necessary to input the node set, the consumable material set, the occupancy material set, and the prevention and control cycle set, and use days as the unit for a single prevention and control cycle. Then, for the nodes in the current cycle, calculate the quantity of consumable materials and occupancy materials to be allocated. If all nodes have recovered, there is no need to allocate materials anymore. Otherwise, it is necessary to allocate random types and random quantities of epidemic prevention materials for each node and increase the prevention and control duration. Finally, return the epidemic prevention resource allocation result as the initial feasible solution.

[0065] Step 5: Calculate the objective functions of the minimum prevention and control resource usage, the minimum prevention and control duration, and the minimum number of infected people for each individual;

[0066] Step 6: Let i = 1;

[0067] Step 7: Implement the individual self-deduction operation: randomly select a certain segment of the matrix sequence encoding corresponding to the individual for update, and the process is executed α times;

[0068] Step 7.1: Randomly select an individual p from the population Ρ;

[0069] Step 7.2: Randomly select the nth row and the mth column in the matrix sequence encoding of the individual p; 0 < n < the total number of prevention and control resources, 0 < m < the total number of nodes;

[0070] Step 7.3: Recalculate the quantity of consumable materials and occupancy materials to be allocated for the segment composed of m*n;

[0071] Step 7.4: Replace the original segment of m*n with the newly generated segment of m*n to form a new individual p';

[0072] Step 7.5: Calculate the objective function of the new individual p'. If the objective function of p' can dominate the scores of some individuals in the population P, then replace the worst one of these individuals with p'. Otherwise, repeat the self-deduction operation until a new individual p' better than some individuals in the population P is found, and replace the worst one of these individuals with p' or reach the execution times limit;

[0073] Step 7.6: Repeat Step 7 α times.

[0074] Figure 4This is the self-deduction schematic diagram in the group deduction mechanism of the method of the present invention. Through self-deduction of the individual matrix sequence encoding, new features are introduced or existing features are enhanced. Such an operation can enable an individual to have more possibilities during the search process, thereby improving the global search ability of the algorithm. The self-deduction operation can be carried out according to a preset probability or can be adaptively adjusted according to the fitness of the individual in the population.

[0075] Step 8: Implement the individual collaborative deduction operation; randomly select two individuals, and exchange a certain segment of the corresponding matrix sequence encodings of the two individuals, and the process is executed β times;

[0076] Step 8.1: Randomly select two individuals q1 and q2 from Ρ;

[0077] Step 8.2: Randomly select the number of rows a (0 < a < the total number of prevention and control resources) and the number of columns b (0 < b < the total number of nodes) in the matrix sequence encoding;

[0078] Step 8.3: Exchange the segment composed of the a*b in the matrix sequence encodings of q1 and q2 to form two new individuals q1' and q2';

[0079] Step 8.4: Calculate the objective functions of the new individuals q1' and q2'. If the objective functions of q1' and q2' can dominate the scores of some individuals in the population P, then replace the worst one of these individuals with q1' and q2' respectively. Otherwise, repeat the collaborative deduction operation until new individuals q' better than some individuals in the population P are found, and replace the worst one of these individuals with q' or reach the upper limit of the execution times;

[0080] Step 8.5: Repeat Step 8 β times.

[0081] Figure 5 This is the collaborative deduction schematic diagram in the group deduction mechanism of the method of the present invention. By combining the excellent features of two different individuals, new individuals with higher fitness are generated. The collaborative deduction operation can increase the diversity in the population, making it easier for the algorithm to jump out of the local optimal solution during the search process.

[0082] Step 9: Provide new individuals based on the continuous improvement mechanism of group dynamic deduction; calculate the objective function of each individual, sort the individuals according to the results, eliminate the last individual, introduce new individuals, and repeat the execution for a set number of times;

[0083] The introduction of new individuals includes individual self-deduction operation or / and individual collaborative deduction operation or / and group dynamic deduction; the group dynamic deduction operation is specifically as follows: randomly select T individuals, select segments for these T individuals, and the sequences where the segments selected by each individual are located do not overlap. After obtaining T segments, combine them to form new individuals;

[0084] Step 9.1: Let r = i*θ / (α+β), θ is an artificially adjusted parameter. If r is an integer and the number of individuals in the group is greater than the initial number P, then execute steps 9.2 to 9.10, that is, remove the worst individual and add new individuals;

[0085] Step 9.2: Calculate the objective function for each individual in the population P, and sort each individual from high to low according to the quality of the objective function;

[0086] Step 9.3: Remove r individuals from the end of the sorted population P set;

[0087] Step 9.4: Generate r / 2 new individuals using individual self-deduction operation and individual collaborative deduction, and add them to the population P;

[0088] Step 9.5: Repeat step 9.2;

[0089] Step 9.6: Perform group dynamics analysis on population P;

[0090] Step 9.7: Starting from the sorted population P set, select T individuals in order;

[0091] Step 9.8: For the T individuals in the group dynamics deduction, select a subsequence in the matrix sequence encoding of each individual in order, that is, the first individual takes the 1 / Tth subsequence, the second individual takes the 2 / Tth subsequence, and so on, the λth individual takes the λ / Tth subsequence, and the T subsequences are fused to form a new individual t;

[0092] Step 9.9: Repeat step 9.8 r / 2 times;

[0093] Step 9.10: Add the newly generated r / 2 individuals to the population P;

[0094] Step 10: i++, restart from step 7;

[0095] Step 11: If the number of iterations i = τ, τ represents the maximum deductive algebra, then stop the calculation, otherwise go to step 7;

[0096] Step 12: Use the final resource allocation plan to schedule medical prevention and control resources.

[0097] Figure 6 , Figure 7The line graphs of IGD and HV indicators of the algorithm of the method of the present invention and other algorithms are shown respectively, and the line graphs of IGD and HV indicators of the solution sets of five algorithms, SDA-DTEM (abbreviation of the method of the present invention), NSGA-Ⅱ, MOEA / D, FP-GA, and SPEA-II, are shown as the number of iterations changes. IGD and HV are two indicators that evaluate the performance of the algorithm by comparing the solution set coverage. Among them, IGD evaluates the solution set by comparing the distance between the solution set and the Pareto front. The average distance from each Pareto front reference solution to the closest solution in the solution set is calculated, so it can not only evaluate the convergence of the solution set, but also the diversity of the solution set. HV is the hypervolume of the union of the solution set and the reference point in the target space. The reference point is generally selected at one end away from the Pareto front, that is, the point composed of the extreme value of each target. The wider the range covered by the solution set, the better the convergence and diversity. During the algorithm iteration process, the non-dominated solution sets obtained by all algorithms are extracted every 200 iterations for indicator analysis. In general, for the IGD index, the performance of SDA-DTEM and NSGA-Ⅱ algorithms is smoother and more stable than other algorithms, but before 500 iterations, the SDA-DTEM algorithm is better than the NSGA-Ⅱ algorithm. For the HV index, it is obvious that the method provided by the present invention is more stable.

[0098] Table 1 shows the C-metric values ​​between the proposed method and other reference algorithms.

[0099]

[0100]

[0101] Table 1 shows the C-metric solution set coverage of the group dynamics deductive algorithm and the other five algorithms. C-metric is an indicator for comparing the advantages and disadvantages of two solution sets. It calculates the proportion of solutions in one solution set that are dominated by at least one solution in the other solution set, so it is also called solution set coverage. The simulation use cases are numbered ExMy-z, where x is the total number of epidemic prevention people, y is the number of epidemic prevention days, and z is the index of the simulation use cases of the same scale. The value of x can be any positive integer, the value of y can be 7, 14, 21, and each combination of x and y has ten use cases, that is, the value of z can be 1-10. For example, E50M7-1 represents the use case with a total number of epidemic prevention people of 50 and a number of epidemic prevention days of 7. In each use case, nearly half of the C-metric values ​​of the group dynamics deductive algorithm are significantly better than those of other algorithms, and many of them have C-metric values ​​equal to 1, indicating that each solution in the solution set obtained by other algorithms is dominated by at least one solution in the solution set obtained by the group dynamics deductive algorithm. Conversely, nearly half of the C-metric values ​​of other algorithms for the group dynamics deductive algorithm are significantly close to zero, and some are equal to zero. Obviously, the dominance of the group dynamics deductive algorithm over the other four algorithms is much higher than the dominance of the other four algorithms over the group dynamics deductive algorithm. The results show that the group dynamics deductive algorithm is superior to the other four algorithms in terms of convergence.

[0102] Table 2 shows the IGD and HV values ​​of the method of the present invention and other reference algorithms

[0103]

[0104]

[0105] Table 2 shows the IGD and HV values ​​of the proposed method and other reference algorithms in different use cases. The simulated use cases are the same as those in Table 1. The IGD and HV values ​​of the group dynamics deduction algorithm are the best in some use cases. Overall, the index value of NSGA-Ⅱ is second to none, and its performance in the IGD index is also excellent. Then, it occasionally shows the best performance in some use cases. Compared with other algorithms, FP-GA has the worst performance in IGD and HV indicators.

[0106] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism, characterized in that: It includes the following steps: Step 1: Define the dynamic spatio-temporal contact network of contacts. Set each individual as a node and determine the activity space of each node that changes over time; Step 2: Matrix sequence encoding; The nodes in the dynamic spatio-temporal contact network and the prevention and control materials are the two dimensions of the matrix, and the allocated quantity of materials is the element value in the matrix sequence encoding. The entire sequence is arranged in sequence according to the prevention and control period T; Step 3: Read the geographical information grid and prevention and control resource information; Step 4: Set the population quantity as P, and generate a population P containing |Ρ| individuals based on the initial solution. Each individual in the population corresponds to a matrix sequence encoding; Step 5: Calculate the objective functions of the minimum prevention and control resource usage, minimum prevention and control duration, and minimum number of infected people for each individual; Step 6: Let the iteration number i = 1 and start the iteration; Step 7: Implement the individual self-deduction operation: Randomly select a certain segment of the matrix sequence encoding corresponding to the individual for update, and the process is executed α times; Step 8: Implement the individual collaborative deduction operation: Randomly select two individuals and exchange a certain segment of the matrix sequence encoding corresponding to the two individuals, and the process is executed β times; Step 9: Provide new individuals based on the continuous improvement mechanism of group dynamic deduction; calculate the objective function of each individual, sort the individuals according to the results, eliminate the last individual, introduce new individuals, and repeat the set number of times; The introduction of new individuals includes individual self-deduction operation or / and individual collaborative deduction operation or / and group dynamic deduction; the specific group dynamic deduction operation is to randomly select T individuals, select segments for these T individuals, and the sequences where the segments selected by each individual are located do not overlap. After obtaining T segments, combine them to form a new individual; Step 10: i++, and start executing from Step 7 again; Step 11: If the iteration number i = τ, where τ represents the maximum number of deductions, then stop the calculation, otherwise go to Step 7; Step 12: Adopt the final resource allocation plan for medical prevention and control resource scheduling.

2. According to claim 1, a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism is characterized in that: The geographical information grid includes the contact time and closeness between nodes, and the prevention and control resource information includes the total number of resources, the maximum number of resources that can be allocated to each node, and the total number of nodes.

3. According to claim 1, a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism is characterized in that: The specific process of generating the initial solution is as follows: First, input the node set, consumable material set, occupancy material set, and prevention and control period set, and use days as the unit for a single prevention and control period. Then, for the nodes in the current period, calculate the quantity of consumable materials and occupancy materials that need to be allocated. If all nodes have recovered, there is no need to allocate materials anymore; otherwise, it is necessary to allocate epidemic prevention materials of random types and random quantities for each node and increase the prevention and control duration; Finally, return the epidemic prevention resource allocation result as the initial solution.

4. According to claim 1, a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism is characterized in that: The specific steps of the individual self-deduction operation are as follows: Step 7.1: Randomly select an individual p from the population Ρ; Step 7.2: Randomly select the nth row and mth column in the matrix sequence encoding of the individual p; 0 < n < total number of prevention and control resources, 0 < m < total number of nodes; Step 7.3: Recalculate the quantity of consumable materials and occupancy materials that need to be allocated for the segment composed of m * n; Step 7.4: Replace the original m*n segment with the newly generated m*n segment to form a new individual p'. Step 7.5: Calculate the objective function of the new individual p'. If the objective function of p' can dominate the scores of some individuals in the population P, then replace the worst one of these individuals with p'. Otherwise, repeat the self-deduction operation until a new individual p' that is better than some individuals in the population P is found, and replace the worst one of these individuals with p' or reach the upper limit of the execution times. Step 7.6: Repeat Steps 7.1 - 7.5 for α times.

5. According to claim 4, a method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism is characterized in that: The specific steps of the individual collaborative deduction operation are as follows: Step 8.1: Randomly select two individuals q1 and q2 from Ρ. Step 8.2: Randomly select the number of rows a (0 < a < total number of prevention and control resources) and the number of columns b (0 < b < total number of nodes) in the matrix sequence encoding. Step 8.3: Exchange the segments composed of the a*b in the matrix sequence encodings of q1 and q2 to form two new individuals q1' and q2'. Step 8.4: Calculate the objective functions of the new individuals q1' and q2'. If the objective functions of q1' and q2' can dominate the scores of some individuals in the population P, then replace the worst one of these individuals with q1' and q2' respectively. Otherwise, repeat the collaborative deduction operation until a new individual q' that is better than some individuals in the population P is found, and replace the worst one of these individuals with q' or reach the upper limit of the execution times. Step 8.5: Repeat Steps 8.1 - 8.3 for β times.

6. A method for scheduling medical prevention and control resources based on a group dynamic deduction mechanism according to claim 5, characterized in that: The specific steps of Step 9 are as follows: Step 9.1: Let r = i*θ / (α + β), where θ is a parameter adjusted by humans. If r is an integer and the number of individuals in the population is greater than the initial number P, then execute Steps 9.2 to 9.10, that is, remove the worst individuals and supplement new individuals. Step 9.2: Calculate the objective function for each individual in the population Ρ and sort each individual from high to low according to the quality of the objective function. Step 9.3: Remove r individuals from the end of the sorted population Ρ set. Step 9.4: Generate r / 2 new individuals using the individual self-deduction operation and individual collaboration deduction and add them to the population Ρ. Step 9.5: Calculate the objective function and sort each individual from high to low according to the quality of the objective function; Step 9.6: Perform group dynamic deduction on the population Ρ. Step 9.7: Starting from the first one in the sorted population Ρ set, sequentially select T individuals in order. Step 9.8: For the T individuals in the group dynamic deduction, select a subsequence in the matrix sequence encoding of each individual in order, that is, the first individual takes the 1 / T subsequence, the second individual takes the 2 / T subsequence, and so on, the λth individual takes the λ / T subsequence, and fuse the T subsequences to form a new individual t. Step 9.9: Repeat Step 9.8 for r / 2 times. Step 9.10: Add the newly generated r / 2 individuals to the population Ρ.

Images