Isolation prevention and control method and device for large-scale crowd

Abstract

The application provides an isolation prevention and control method and device for a large-scale crowd, comprising: constructing a physical contact network according to individuals in a prevention and control area and contact behaviors among the individuals at a previous time; dividing the physical contact network into a plurality of sub-networks; estimating an infection probability of each individual in the prevention and control area at a current time according to the physical contact network; and determining individuals in a region corresponding to each sub-network that need to be isolated at the current time based on the infection probability. The application formulates a refined isolation prevention and control strategy for a region corresponding to each sub-network on the basis of accurately estimating the infection probability of each individual in the prevention and control area, overcomes the problem of excessively high time cost of an isolation prevention and control algorithm caused by too large a scale of individuals in the prevention and control area, and improves the isolation prevention and control effect of the region corresponding to each sub-network.

Description

Technical Field

[0001] This invention relates to the field of infectious disease safety prevention and control technology, and in particular to a method and device for isolation and control of large-scale populations. Background Technology

[0002] Large-scale outbreaks of infectious diseases can cause enormous casualties and property damage to human society. Isolating individuals carrying pathogens is an effective measure to curb the spread of infectious diseases. Currently, most regions with very large populations adopt a proportional isolation approach for control, but this method is ineffective. Therefore, the development of precise isolation and control strategies (isolation and control strategies that can determine whether each individual should be isolated at any given time) is of great significance.

[0003] Existing precise isolation and control strategies, due to time complexity issues, can mostly only handle precise isolation and control of small populations. The only method capable of handling precise isolation and control of very large populations, Netshield+, has a time complexity proportional to the square of the number of isolated individuals in the environment. Therefore, in large populations (e.g., millions of people), it can generally only handle the isolation of a few dozen individuals, and isolating a few dozen individuals in a large population is often ineffective in controlling the spread of disease.

[0004] Therefore, existing precise isolation and control strategies are not suitable for environments with large populations. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and apparatus for isolation and control of large-scale populations. Based on accurately predicting the infection probability of each individual in the control area, it formulates refined isolation and control strategies for each sub-network corresponding to the region. This overcomes the problem of excessively high time costs for isolation and control algorithms caused by the large number of individuals in the control area and improves the isolation and control effectiveness of each sub-network corresponding to the region.

[0006] In a first aspect, the present invention provides a method for isolation and control of large-scale populations, the method comprising:

[0007] Construct a physical contact network based on the individuals and their contact behaviors within the control area at the previous moment;

[0008] The physical contact network is divided into multiple sub-networks;

[0009] Based on the physical contact network, the probability of infection for each individual in the prevention and control area at the current moment is estimated;

[0010] Based on the infection probability, determine the individuals that need to be isolated in the current time for each region corresponding to the sub-network.

[0011] According to the isolation and control method for large-scale populations provided by the present invention, dividing the physical contact network into multiple sub-networks includes:

[0012] Based on the principle that the nodes within a subnetwork are tightly connected and the nodes between subnetworks are sparsely connected, the physical contact network is divided into multiple subnetworks.

[0013] or

[0014] The physical contact network is divided into multiple sub-networks.

[0015] According to the isolation and control method for large-scale populations provided by the present invention, the step of estimating the infection probability of each individual in the control area at the current moment based on the physical contact network includes:

[0016] If there are no new confirmed cases in the control area at the current moment, the infection probability of each individual in the control area at the current moment is estimated by using the infection probability of each individual at the previous moment and the physical contact network.

[0017] According to the isolation and control method for large-scale populations provided by the present invention, the step of estimating the infection probability of each individual in the control area at the current moment based on the physical contact network includes:

[0018] If there are newly confirmed cases in the prevention and control area at the current time, the infection probability correction time relative to the prevention and control area is determined based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient at the previous time and every time before that time.

[0019] The infection probability of each individual in the prevention and control area is corrected sequentially at each time after the infection probability correction time until the corrected value of the infection probability of each individual in the prevention and control area at the previous time is obtained.

[0020] Using the corrected value of the infection probability of each individual in the control area at the previous moment and the physical contact network, the infection probability of each individual in the control area at the current moment is estimated.

[0021] According to the isolation and control method for large-scale populations provided by the present invention, determining the infection probability correction time relative to the control area based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient at the previous time and every time before that time includes:

[0022] Based on the probability distribution of the incubation period of infectious diseases, calculate the probability that each newly diagnosed patient was in the latent state at the previous moment and at every moment before that.

[0023] The probability of infection of each newly diagnosed patient at the previous time and at each time before that time is subtracted from the probability of each newly diagnosed patient being in the latent state at the previous time and at each time before that time, and the time when the difference is greater than 0 is taken as the correction time corresponding to each newly diagnosed patient.

[0024] The minimum time among all the corrected times corresponding to confirmed patients is taken as the infection probability corrected time.

[0025] According to the isolation and control method for large-scale populations provided by the present invention, the step of sequentially correcting the infection probability of each individual in the control area at each time after the infection probability correction time until the corrected value of the infection probability of each individual in the control area at the previous time includes:

[0026] For any time between the infection probability correction time and the previous time, based on the correction value of the infection probability of each individual at the previous time, the infection probability of each individual at any time is re-estimated.

[0027] If each of the individuals is not a newly diagnosed patient, then the infection probability of each individual at any time is corrected to the maximum value of the infection probability of each individual at any time and the re-estimated infection probability of each individual at any time.

[0028] If each of the individuals is a newly diagnosed patient, then the infection probability of each individual at any given time is revised to the maximum value among the infection probability of each individual at any given time, the re-estimated infection probability of each individual at any given time, and the probability that each individual is in a latent state at any given time.

[0029] According to the isolation and control method for large-scale populations provided by the present invention, determining the individuals who need to be isolated at the current moment in each sub-network area based on the infection probability includes:

[0030] Based on the infection probability, a disease transmission loss and an isolation loss are constructed for each sub-network; wherein, the disease transmission loss and isolation loss of each sub-network are both related to the isolation vector of each sub-network;

[0031] The goal is to minimize the weighted sum of the disease transmission loss and the isolation loss, and then solve for the isolation vector of each sub-network.

[0032] Based on the solution results of the isolation vector of each sub-network, determine the individuals that need to be isolated in the region corresponding to each sub-network at the current time;

[0033] The isolation vector of each sub-network represents whether each individual in each sub-network should be isolated at the current moment.

[0034] According to the isolation and control method for large-scale populations provided by the present invention, the step of constructing the disease transmission loss of each sub-network based on the infection probability includes:

[0035] The weighted sum of the probabilities of all individuals in each sub-network being healthy and then infected by already infected individuals within the network at the current moment is defined as the first disease transmission loss of each sub-network.

[0036] The weighted sum of the probabilities that all infected individuals in each sub-network will infect healthy individuals in the outside world at the current moment is defined as the second disease transmission loss of each sub-network.

[0037] The weighted sum of the probabilities of all individuals in each sub-network being healthy and then infected by an externally infected individual at the current moment is defined as the third disease transmission loss of each sub-network.

[0038] Based on the infection probability and the isolation vector of each unassigned sub-network, the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss are constructed.

[0039] The sum of the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss is taken as the disease transmission loss of each sub-network.

[0040] The isolation loss for constructing each of the sub-networks includes:

[0041] The total number of isolated individuals in each sub-network at the current moment is defined as the isolation loss of each sub-network, and the isolation loss is constructed based on the isolation vector of each sub-network that has not been assigned a value.

[0042] According to the isolation and control method for large-scale populations provided by the present invention, the first disease transmission loss The expression is as follows:

[0043]

[0044]

[0045] Second disease transmission loss The expression is as follows:

[0046]

[0047] The third disease transmission loss The expression is as follows:

[0048]

[0049] in, For nodes The weight corresponding to the individual, For nodes The probability that an individual at time t changes from being healthy to being infected by an already infected individual within the same group. Let be the set of all individuals in the region corresponding to the k-th subnetwork at time t. For nodes The probability of an individual being infected at time t. node The probability of an individual being infected at time t. For nodes The propagation rate of the corresponding individual at time t. For nodes The propagation rate of the corresponding individual at time t. The adjacency matrix of the physical contact network is the first... Line number Column elements, and The isolation vectors of the k-th sub-network are respectively The element and the first There are 1 elements, where a value of 1 indicates isolation and a value of 0 indicates no isolation. and The nodes before time t are respectively and nodes The value is 0 if the individual has been diagnosed with COVID-19, and 1 otherwise. Secondly, this invention provides an isolation and control device for large-scale populations, the device comprising:

[0050] The module is used to construct a physical contact network based on the individuals in the control area at the previous moment and their contact behavior;

[0051] A partitioning module is used to divide the physical contact network into multiple sub-networks;

[0052] The prediction module is used to predict the probability of infection of each individual in the prevention and control area at the current moment based on the physical contact network.

[0053] The determination module is used to determine, based on the infection probability, the individuals that need to be isolated in the current time for each region corresponding to the sub-network.

[0054] This invention provides a method and apparatus for isolating and controlling large-scale populations, comprising: constructing a physical contact network based on individuals in the control area at the previous moment and their contact behavior; dividing the physical contact network into multiple sub-networks; estimating the infection probability of each individual in the control area at the current moment based on the physical contact network; and determining, based on the infection probability, the individuals that need to be isolated in the area corresponding to each sub-network at the current moment. This invention, based on accurately estimating the infection probability of each individual in the control area, formulates a refined isolation and control strategy for the area corresponding to each sub-network, overcoming the problem of excessively high time costs for isolation and control algorithms caused by the large number of individuals in the control area, and improving the isolation and control effectiveness of the area corresponding to each sub-network. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0056] Figure 1 This is a flowchart illustrating the isolation and control method for large-scale populations provided by the present invention.

[0057] Figure 2 This is a comparison chart of the runtime of the present invention and the NetShield+ method;

[0058] Figure 3 This is a comparison chart showing the effectiveness of the present invention and existing methods in controlling disease transmission under the same average number of isolated individuals;

[0059] Figure 4 This is a schematic diagram of the structure of the isolation and control device for large-scale populations provided by the present invention;

[0060] Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention;

[0061] Figure label:

[0062] 510: Processor; 520: Communication interface; 530: Memory; 540: Communication bus. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0064] Transmission rate: The probability that a healthy individual will become infected when they come into contact with an infected person.

[0065] Recovery rate: The probability that a given infected person recovers and becomes a recovered person;

[0066] For infectious diseases with an incubation period, individuals in the environment may be in four states: susceptible (S), latent (E), infected (I), and recovered (R); among which:

[0067] Susceptible state S: Individuals in this state are healthy and not infectious, but may enter the latent state E at any time due to contact with an infected person.

[0068] Latent state E: Individuals in susceptible state S enter this state after infection. This state is asymptomatic and therefore difficult to distinguish, but it is infectious. Individuals in this state will enter infectious state I after a period of time.

[0069] Infectious state I: Entered after a period of time from latent state E. This state is symptomatic and contagious. After a period of time, it enters the recovery state R.

[0070] Recovery state R: Entered after a period of time from infected state I, this state will not be reinfected.

[0071] Infected individuals: Individuals in latent state E or infectious state I.

[0072] The following is combined Figures 1-5 This invention describes a method and apparatus for isolating and controlling large populations.

[0073] Firstly, this invention provides a method for isolation and control of large-scale populations, such as... Figure 1 As shown, the method includes:

[0074] S11. Construct a physical contact network based on the individuals in the control area at the previous moment and their contact behavior;

[0075] S12. Divide the physical contact network into multiple sub-networks;

[0076] S13. Based on the physical contact network, estimate the infection probability of each individual in the prevention and control area at the current moment;

[0077] S14. Based on the infection probability, determine the individuals that need to be isolated in the current time for each region corresponding to the sub-network.

[0078] This invention proposes an isolation and control method based on graph segmentation, which approximates the isolation and control of the control area as the isolation and control of the area corresponding to each sub-network, in order to solve the problem that it is difficult to effectively control the spread of diseases in areas with a very large population due to time complexity issues.

[0079] The isolation and control method for large-scale populations provided by this invention formulates refined isolation and control strategies for each sub-network based on the accurate prediction of the infection probability of each individual in the control area. This overcomes the problem of excessively high time cost of isolation and control algorithms caused by the large number of individuals in the control area and improves the isolation and control effect of each sub-network.

[0080] Specifically, S11 includes:

[0081] Each individual in the control area is regarded as a node, and the physical contact between two individuals in the control area at the previous moment is regarded as the edge between the corresponding nodes of the two individuals, thus generating the physical contact network.

[0082] The physical contact network It reflects the physical contact between individuals at the previous time step (t-1).

[0083] For any individual in the control area at the previous time step, we represent them using a node on the physical contact network. If any two individuals corresponding to any two nodes on the physical contact network are in physical contact at time t-1, then an edge is drawn between these two nodes; otherwise, there is no edge connecting them. This allows us to construct the physical contact network from the previous time step.

[0084] It should be noted that physical contact here refers to being in the same enclosed space or in close proximity.

[0085] This invention uses physical contact networks This represents the contact relationships between individuals in the control area at the previous moment, laying the foundation for estimating the infection probability of individuals in the control area at the current moment.

[0086] Specifically, S12 includes:

[0087] Based on the principle that the nodes within a subnetwork are tightly connected and the nodes between subnetworks are sparsely connected, the physical contact network is divided into multiple subnetworks.

[0088] or

[0089] The physical contact network is divided into multiple sub-networks.

[0090] In this invention, we consider the case of extremely large scale, where the number of nodes in the contact network between individuals ranges from tens of thousands to hundreds of thousands or even higher. To reduce computational complexity in subsequent calculations, this invention first divides the extremely large-scale contact network into many sub-networks.

[0091] Alternatively, existing network partitioning methods can be used.

[0092] For example, partitioning methods that make the internal connections of subnetworks tighter and the connections between subnetworks sparser, partitioning methods that randomly divide the physical contact network, or partitioning methods that try to ensure that the subnetworks are of uniform size.

[0093] This invention divides the physical contact network into multiple sub-networks, laying the foundation for refined isolation and control of the corresponding areas of each sub-network.

[0094] Specifically, since the latent state E is asymptomatic but infectious, it is crucial to accurately estimate whether an individual is in this state. Therefore, this invention provides an infection probability estimation algorithm that considers two scenarios: the first scenario is that there are no newly confirmed cases in the control area at the current time, and the second scenario is that there are newly confirmed cases in the control area at the current time; patients who show symptoms are considered as confirmed cases.

[0095] In the first case, S13 includes:

[0096] S13.1: Using the infection probability of each individual at the previous moment and the physical contact network, estimate the infection probability of each individual in the prevention and control area at the current moment.

[0097] Furthermore, S13.1 includes:

[0098] Using the infection probability of each individual at the previous moment, the adjacency matrix of the physical contact network, and the pre-stored infection probability calculation formula, the infection probability of each individual in the prevention and control area at the current moment is estimated.

[0099] The formula for calculating the infection probability is:

[0100]

[0101] Where, γ i Let β be the recovery rate of the i-th individual in the control area. i Let i be the transmission rate of the i-th individual in the control area. V is the element in the i-th row and j-th column of the adjacency matrix of the physical contact network. t-1 Let be the set of all individuals in the control area at time t-1. Let be the infection probability of the i-th individual in the control area at time t. Let be the infection probability of the i-th individual in the control area at time t-1. Let be the probability of infection of the j-th individual in the control area at time t-1. This indicates whether the i-th individual in the control area before time t-1 has been diagnosed. If diagnosed, the value is 0; otherwise, the value is 1.

[0102] It is understandable that for the i-th individual in the physical contact network, the probability of infection at the current time (time t) is... for:

[0103]

[0104] Given that the risk of infection for any healthy individual after close contact with a confirmed case is very small, therefore... Approximately

[0105] Ultimately, we can obtain:

[0106]

[0107] Here, γ i and β i It was issued by an infectious disease research institution in response to the transmission characteristics of infectious diseases.

[0108] For ease of calculation, the above formula can also be rewritten in vector and matrix form:

[0109] p t =F(p t-1 )

[0110] =(1-γ+γ⊙s t-1 )⊙p t-1 +[1-p t-1 ]⊙β⊙s t-1 ⊙(A t-1 ·p t-1 )

[0111] Here, p t p t-1 , γ, s t-1 The i-th element in β and β are respectively γ i , and β i A t-1Let be the adjacency matrix of the physical contact network.

[0112] In the second scenario, when a new confirmed case appears at time t, it signifies the discovery of new information, allowing for the correction of the previously estimated individual infection probability and the estimation of the current individual infection probability. Therefore, in this case, S13 includes:

[0113] S13-A: Based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient at the previous time and every time before that time, determine the infection probability correction time relative to the control area;

[0114] S13-B: Sequentially correct the infection probability of each individual in the prevention and control area at each time after the infection probability correction time, until the corrected value of the infection probability of each individual in the prevention and control area at the previous time is obtained;

[0115] S13-C: Using the corrected value of the infection probability of each individual in the prevention and control area at the previous moment and the physical contact network, estimate the infection probability of each individual in the prevention and control area at the current moment.

[0116] Preferably, S13-A includes:

[0117] S13-A-1: Based on the probability distribution of the incubation period of infectious diseases, calculate the probability that each newly diagnosed patient was in the latent state at the previous time and at every time before that time.

[0118] S13-A-2: Subtract the infection probability of each newly confirmed patient at the previous time and at each time before from the probability of each newly confirmed patient being in the latent state at the previous time and at each time before, and take the time when the difference is greater than 0 as the correction time corresponding to each newly confirmed patient.

[0119] S13-A-3: The minimum time among all the corrected times corresponding to confirmed patients shall be used as the infection probability corrected time.

[0120] S13-A-1 includes:

[0121] For a newly diagnosed patient k at time t, the probability that he / she is in a latent state at time t-τ is:

[0122]

[0123] Where g(t) is the probability distribution of the incubation period of an infectious disease, provided by an infectious disease research institution; k∈Φ t , Φ tLet be the set of newly diagnosed patients at time t, where 0 ≤ τ ≤ t-1.

[0124] The S13-A-2 is expressed by the formula:

[0125]

[0126] t k =(t-1)-t 0,k

[0127] Among them, t k For the corrected time point corresponding to newly confirmed patient k, t 0,k The time frame for the newly confirmed patient k was adjusted and shifted forward. The probability of infection of newly confirmed patient k at time t-1-τ.

[0128] The S13-A-3 is expressed by the formula:

[0129] t w =mint k

[0130] k∈Φ t

[0131] Among them, t w This is the timeframe for correcting the infection probability.

[0132] Preferably, S13-B includes:

[0133] S13-B-1: For any time between the infection probability correction time and the previous time, based on the correction value of the infection probability of each individual at the previous time, re-estimate the infection probability of each individual at any time.

[0134] S13-B-2: If each of the individuals is not a newly diagnosed patient, then the infection probability of each individual at any time is corrected to the maximum value of the infection probability of each individual at any time and the re-estimated infection probability of each individual at any time.

[0135] If each of the individuals is a newly diagnosed patient, then the infection probability of each individual at any given time is revised to the maximum value among the infection probability of each individual at any given time, the re-estimated infection probability of each individual at any given time, and the probability that each individual is in a latent state at any given time.

[0136] The S13-B-1 includes:

[0137] Adjust time t based on infection probabilityw Re-estimate the number of individuals in the control area [t] w The probability of infection at each time step in [t-1)], assuming f is [t w For any time in [t-1], the re-prediction result of the infection probability of the i-th individual in the control area at time f can be:

[0138]

[0139] here, and These are the correction values ​​for the infection probability of the i-th individual and the j-th individual in the control area at time f-1, respectively.

[0140] The S13-B-2 is expressed by the formula:

[0141] If i∈Φ t hour,

[0142] like hour,

[0143] in, This is a correction value for the infection probability of the i-th individual in the control area at time f.

[0144] The S13-C includes:

[0145] Similar to the first case, the infection probability of the i-th individual in the corrected control area at time t-1 is calculated. Substitute into the formula Estimate the probability of infection of the i-th individual in the control area at time t.

[0146] This invention proposes a method for real-time estimation and updating of the probability of infection of all individuals in a control area (i.e., the probability of being in the latent state E or the infectious state I). This method estimates the probability of infection of individuals at time t+1 based on the individual infection probability at time t, which can help formulate the optimal isolation strategy.

[0147] Specifically, S14 includes:

[0148] S14.1: Construct the disease transmission loss and the isolation loss of each sub-network based on the infection probability; wherein, the disease transmission loss and the isolation loss of each sub-network are both related to the isolation vector of each sub-network;

[0149] S14.2: Solve for the isolation vector of each sub-network with the objective of minimizing the weighted sum of the disease transmission loss and the isolation loss;

[0150] S14.3: Based on the solution results of the isolation vector of each sub-network, determine the individuals that need to be isolated in the region corresponding to each sub-network at the current time;

[0151] The isolation vector of each sub-network represents whether each individual in each sub-network should be isolated at the current moment.

[0152] This invention defines two types of loss: disease transmission loss and isolation loss. The total loss is a weighted sum of the two losses. Minimizing the total loss aims to control disease transmission and reduce the number of people in isolation.

[0153] Preferably, in S14.1, constructing the disease transmission loss for each sub-network based on the infection probability includes:

[0154] S14.1.1: The weighted sum of the probabilities of all individuals in each sub-network going from healthy to being infected by already infected individuals in the sub-network at the current moment is defined as the first disease transmission loss of each sub-network;

[0155] S14.1.2: The weighted sum of the probabilities that all infected individuals in each sub-network will infect healthy individuals in the outside world at the current moment is defined as the second disease transmission loss of each sub-network;

[0156] S14.1.3: The weighted sum of the probabilities of all individuals in each sub-network going from healthy to being infected by an externally infected individual at the current moment is defined as the third disease transmission loss of each sub-network;

[0157] S14.1.4: Based on the infection probability and the isolation vector of each of the unassigned sub-networks, construct the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss;

[0158] S14.1.5: The sum of the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss shall be taken as the disease transmission loss of each sub-network;

[0159] The isolation loss for each sub-network constructed in S14.1 includes:

[0160] S14.1-I: Define the total number of isolated individuals in each sub-network at the current time as the isolation loss of each sub-network, and construct the isolation loss based on the isolation vector of each sub-network that has not been assigned a value.

[0161] The weighted sum of the probabilities of all individuals in the control area going from healthy to infected at the current moment is defined as the disease transmission loss in the control area. Therefore, the disease transmission loss in the control area... The calculation formula is:

[0162]

[0163]

[0164] Where, δ i Let i be the weight of the i-th individual in the prevention and control area. Let be the probability that the i-th individual in the control area changes from healthy to infected at time t. and These are the i-th and j-th elements in the isolation vector, respectively. A value of 1 indicates isolation, and a value of 0 indicates no isolation.

[0165] For ease of calculation, the above formula can also be rewritten as:

[0166]

[0167] Here, v is a weight vector representing the importance of different individuals. For example, older people in the environment can be given higher weights because they are at higher risk of infection. The i-th element in δ is δ_i. i ,x t This is the isolation vector.

[0168] Since the individual scale in the prevention and control area is too large, it will lead to excessive time cost of the algorithm. Therefore, the present invention will next calculate the disease transmission loss of each sub-network. This loss can be divided into three parts: the first disease transmission loss, which represents the disease transmission loss within the sub-graph; the second disease transmission loss, which represents the loss caused by the sub-network spreading to the outside; and the third disease transmission loss, which represents the loss caused by the outside spreading to the inside of the sub-network.

[0169] Therefore, the first disease transmission loss The expression is as follows:

[0170]

[0171]

[0172] in, For nodes The weight corresponding to the individual, For nodes The probability that an individual at time t changes from being healthy to being infected by an already infected individual within the same group. Let be the set of all individuals in the region corresponding to the k-th subnetwork at time t. For nodes The probability of an individual being infected at time t. node The probability of an individual being infected at time t. For nodes The propagation rate of the corresponding individual at time t. and The isolation vectors of the k-th sub-network are respectively The element and the first There are 1 elements, where a value of 1 indicates isolation and a value of 0 indicates no isolation. The node before time t Whether the individual has been diagnosed before; if so, the value is 0, otherwise the value is 1. The adjacency matrix of the physical contact network is the first... Line number Column elements.

[0173] Its matrix form is:

[0174]

[0175] in, β k and δ k x t A t p t s t The corresponding part of the k-th subnetwork in β and δ.

[0176] Second disease transmission loss The expression is as follows:

[0177]

[0178] in, For nodes The propagation rate of the corresponding individual at time t. The node before time t The value indicates whether the individual has been diagnosed with the disease; if so, the value is 0, otherwise the value is 1.

[0179] The isolation state outside a subgraph is generally unknown, so this invention assumes the worst-case scenario, i.e., no isolation measures are taken outside the subgraph. Therefore, during calculation, the worst-case scenario can be assumed. Approximately 0, thus obtaining

[0180] The approximate matrix form is as follows:

[0181]

[0182] in, for Middle node The corresponding value, For the nodes in the adjacency matrix of the physical contact network The element values ​​between the node and the node outside the k-th subnetwork. and β -k p t s t The external part of the k-th subnetwork in β.

[0183] The third disease transmission loss The expression is as follows:

[0184]

[0185] For the same reasons as the loss from the spread of the second disease, the calculation can be performed using... Approximately 0, thus obtaining

[0186] The approximate matrix form is as follows:

[0187]

[0188]

[0189] in, for Middle node The corresponding value.

[0190] The total loss from disease transmission is the sum of three parts, namely

[0191] In addition, the total isolation loss is the total number of all isolated individuals in the control area, calculated using the following formula:

[0192]

[0193] For ease of calculation, the above formula can also be rewritten as:

[0194]

[0195] Then the isolation loss of the kth subnetwork is

[0196] The total loss in S14.2 is expressed by the formula:

[0197]

[0198] Here, k1 and k2 are the weights of the two types of loss, respectively. By setting different weights, a trade-off can be made between different losses.

[0199] To find the optimal isolation method, i.e. the optimal To make the total loss To minimize the total loss, this invention will reduce the overall loss. The minimum optimization problem is transformed into a standard 0-1 quadratic programming problem. Since 0-1 quadratic programming is an NP-hard problem and cannot be solved in polynomial time, it is transformed into a semidefinite programming problem through convex relaxation and then solved using the CVX toolbox.

[0200] S14.3 includes:

[0201] The solution Individuals corresponding to nodes with a value of 1 are isolated.

[0202] This invention establishes an isolation optimization model for corresponding regions of a sub-network to control disease transmission and reduce the number of people requiring isolation. The time complexity of this method is significantly lower than other existing methods, and its performance is far superior.

[0203] Furthermore, this invention improves the algorithm's performance by using a convex relaxation method for solving the problem.

[0204] The superiority of this invention can be verified in the following ways:

[0205] (1) Compare the disease transmission control effects of the method of this invention with those of the Netshield+ method. Specifically, simulations were performed on a network with 100,000 nodes, and the results are as follows: Figure 2 The above is a comparison chart of runtime.

[0206] As can be seen, the method of the present invention has a low running time regardless of the isolation ratio, even with 100,000 nodes. In contrast, the NetShield+ method only has a short running time when the isolation ratio is very low, and the running time increases sharply as the isolation ratio increases.

[0207] (2) Given several existing disease transmission control methods, Netshield, Netshield+, Acquaintance, GreedyDrop, SVID, Mod-Centrality, Comm-Centrality, HighDegree, and LowDegree, compare the disease transmission control effectiveness of the method of this invention with that of existing disease transmission control methods, assuming the same average number of isolated nodes. These existing disease transmission control methods are briefly described below:

[0208] Netshield and Netshield+: Determine which individuals should be isolated by measuring the impact of isolating individuals on the network shield value;

[0209] Acquaintance: Isolating individuals by finding acquaintances;

[0210] GreedDrop: Identifying isolated individuals using graph theory;

[0211] SVID: Isolates individuals based on their Shapley value;

[0212] Mod-Centrality and Comm-Centrality: Community-based approaches that determine whether to isolate individuals based on their impact on the community structure; Mod-Centrality can be abbreviated as Mod, and Comm-Centrality can be abbreviated as Comm.

[0213] HighDegree and LowDegree: Isolate individuals with the highest degree (HighDegree) or the lowest degree (LowDegree).

[0214] Specifically, an experiment was conducted on a scale-free network with 1000 nodes and an average degree of 4. The transmission rate / recovery rate of the infectious disease was set to 1.3, and the incubation period was fixed at 15 days. During the simulation, to simulate real-world conditions, the disease was first allowed to spread in the environment. int After a certain time, control measures will be taken. int This is called the intervention time, which is obtained under the same average number of isolated nodes. Figure 3 The chart shows a comparison of the effectiveness of disease transmission control.

[0215] As can be seen, with the same average number of isolated nodes, the method of the present invention is far superior to existing methods in terms of disease control.

[0216] In summary, this invention solves the problem of controlling disease transmission through precise isolation in extremely large populations. Compared to previous methods, this method significantly reduces runtime and improves control effectiveness. Compared to the NetShield+ method, the time overhead of this invention is independent of the number of people isolated, allowing it to handle a wider range of real-world scenarios while greatly reducing time complexity.

[0217] Secondly, the isolation and control device for large-scale populations provided by the present invention will be described. The isolation and control device for large-scale populations described below and the isolation and control method for large-scale populations described above can be referred to in correspondence. Figure 4A structural schematic diagram of an isolation and control device for large-scale populations is provided, such as... Figure 4 As shown, the device includes:

[0218] Module 21 is used to construct a physical contact network based on the individuals in the control area at the previous moment and their contact behavior.

[0219] The partitioning module 22 is used to divide the physical contact network into multiple sub-networks;

[0220] The prediction module 23 is used to predict the infection probability of each individual in the prevention and control area at the current moment based on the physical contact network.

[0221] The determination module 24 is used to determine, based on the infection probability, the individuals that need to be isolated in the current time for each region corresponding to the sub-network.

[0222] The isolation and control device for large-scale populations provided by this invention formulates a refined isolation and control strategy for each sub-network based on the accurate prediction of the infection probability of each individual in the control area. This overcomes the problem of excessively high time cost of isolation and control algorithms caused by the large number of individuals in the control area and improves the isolation and control effect of each sub-network.

[0223] Based on the above embodiments, as an optional embodiment, the partitioning module is used for:

[0224] Based on the principle that the nodes within a subnetwork are tightly connected and the nodes between subnetworks are sparsely connected, the physical contact network is divided into multiple subnetworks.

[0225] or

[0226] The physical contact network is divided into multiple sub-networks.

[0227] Based on the above embodiments, as an optional embodiment, the prediction module includes: a first prediction unit, used for:

[0228] If there are no new confirmed cases in the control area at the current moment, the infection probability of each individual in the control area at the current moment is estimated by using the infection probability of each individual at the previous moment and the physical contact network.

[0229] Based on the above embodiments, as an optional embodiment, the prediction module includes: a second prediction unit, comprising:

[0230] The determination submodule is used to determine the infection probability correction time relative to the prevention and control area when there are newly confirmed patients in the current time. This is based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient in the previous time and every time before that time.

[0231] The correction submodule is used to sequentially correct the infection probability of each individual in the prevention and control area at each time after the infection probability correction time, until the correction value of the infection probability of each individual in the prevention and control area at the previous time is obtained.

[0232] The estimation submodule is used to estimate the infection probability of each individual in the prevention and control area at the current moment by using the correction value of the infection probability of each individual in the prevention and control area at the previous moment and the physical contact network.

[0233] Based on the above embodiments, as an optional embodiment, the determining submodule includes:

[0234] The computational subunit is used to calculate the probability that each newly diagnosed patient was in the latent state at the previous time and every time before that time, based on the probability distribution of the incubation period of infectious diseases.

[0235] The difference sub-unit is used to subtract the infection probability of each newly diagnosed patient at the previous time and at each time before the previous time from the probability of each newly diagnosed patient being in the latent state at the previous time and at each time before the previous time, and the time when the difference is greater than 0 is used as the correction time corresponding to each newly diagnosed patient.

[0236] A sub-unit is set up to use the minimum time among all the corrected times corresponding to confirmed patients as the infection probability corrected time.

[0237] Based on the above embodiments, as an optional embodiment, the correction submodule includes:

[0238] The re-prediction subunit is used to re-predict the infection probability of each individual at any time between the infection probability correction time and the previous time, based on the correction value of the infection probability of each individual at the previous time.

[0239] A subunit is configured to, if each individual is not a newly diagnosed patient, correct the infection probability of each individual at any given time to the maximum value between the infection probability of each individual at any given time and the re-estimated infection probability of each individual at any given time.

[0240] If each of the individuals is a newly diagnosed patient, then the infection probability of each individual at any given time is revised to the maximum value among the infection probability of each individual at any given time, the re-estimated infection probability of each individual at any given time, and the probability that each individual is in a latent state at any given time.

[0241] Based on the above embodiments, as an optional embodiment, the determining module includes:

[0242] A construction unit is configured to construct the disease transmission loss and the isolation loss of each sub-network based on the infection probability; wherein the disease transmission loss and the isolation loss of each sub-network are both related to the isolation vector of each sub-network.

[0243] The solving unit is used to solve for the isolation vector of each sub-network with the objective of minimizing the weighted sum of the disease transmission loss and the isolation loss;

[0244] The determining unit is used to determine the individuals that need to be isolated in the region corresponding to each sub-network at the current time, based on the solution result of the isolation vector of each sub-network.

[0245] The isolation vector of each sub-network represents whether each individual in each sub-network should be isolated at the current moment.

[0246] Based on the above embodiments, as an optional embodiment, the building unit includes:

[0247] The first disease transmission loss definition submodule is used to define the weighted sum of the probabilities of all individuals in each subnetwork going from healthy to being infected by already infected individuals in the subnetwork at the current moment as the first disease transmission loss of each subnetwork.

[0248] The second disease transmission loss definition submodule is used to define the second disease transmission loss of each subnetwork as the weighted sum of the probabilities that all infected individuals in each subnetwork will infect healthy individuals in the outside at the current moment.

[0249] The third disease transmission loss definition submodule is used to define the third disease transmission loss of each subnetwork as the weighted sum of the probabilities of all individuals in each subnetwork going from healthy to being infected by externally infected individuals at the current moment.

[0250] A submodule is constructed to construct the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss based on the infection probability and the isolation vector of each unassigned subnetwork.

[0251] The sub-network disease transmission loss setting submodule is used to take the sum of the first disease transmission loss, the second disease transmission loss and the third disease transmission loss as the disease transmission loss of each sub-network.

[0252] The building unit further includes:

[0253] The isolation loss construction submodule is used to define the total number of isolated individuals in each subnetwork at the current time as the isolation loss of each subnetwork, and to construct the isolation loss based on the isolation vector of each subnetwork that has not been assigned a value.

[0254] Based on the above embodiments, as an optional embodiment, the first disease transmission loss The expression is as follows:

[0255]

[0256]

[0257] Second disease transmission loss The expression is as follows:

[0258]

[0259] The third disease transmission loss The expression is as follows:

[0260]

[0261] in, For nodes The weight corresponding to the individual, For nodes The probability that an individual at time t changes from being healthy to being infected by an already infected individual within the same group. Let be the set of all individuals in the region corresponding to the k-th subnetwork at time t. For nodes The probability of an individual being infected at time t. node The probability of an individual being infected at time t. For nodes The propagation rate of the corresponding individual at time t. For nodes The propagation rate of the corresponding individual at time t. The adjacency matrix of the physical contact network is the first... Line number Column elements, and The isolation vectors of the k-th sub-network are respectively The element and the first There are 1 elements, where a value of 1 indicates isolation and a value of 0 indicates no isolation. and The nodes before time t are respectively and nodes The value indicates whether the individual has been diagnosed with the disease; if so, the value is 0, otherwise the value is 1.

[0262] Thirdly, Figure 5 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include a processor 510, a communications interface 520, a memory 530, and a communication bus 540, wherein the processor 510, communications interface 520, and memory 530 communicate with each other via the communication bus 540. The processor 510 can call logical instructions in the memory 530 to execute a method for isolating and controlling large-scale populations. This method includes: constructing a physical contact network based on individuals in the control area at the previous moment and their contact behavior; dividing the physical contact network into multiple sub-networks; estimating the infection probability of each individual in the control area at the current moment based on the physical contact network; and determining, based on the infection probability, the individuals in each sub-network that need to be isolated at the current moment.

[0263] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0264] Fourthly, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the isolation and control methods for large-scale populations provided by the above methods. The method includes: constructing a physical contact network based on individuals in the control area at the previous moment and their contact behavior; dividing the physical contact network into multiple sub-networks; estimating the infection probability of each individual in the control area at the current moment based on the physical contact network; and determining, based on the infection probability, the individuals in the area corresponding to each sub-network who need to be isolated at the current moment.

[0265] Fifthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the isolation and control methods for large-scale populations provided by the methods described above. The method includes: constructing a physical contact network based on individuals in the control area at the previous moment and their contact behavior; dividing the physical contact network into multiple sub-networks; estimating the infection probability of each individual in the control area at the current moment based on the physical contact network; and determining, based on the infection probability, the individuals in the area corresponding to each sub-network who need to be isolated at the current moment.

[0266] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0267] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0268] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for isolation and control of large-scale populations, characterized in that, The method includes: Construct a physical contact network based on the individuals and their contact behaviors within the control area at the previous moment; The physical contact network is divided into multiple sub-networks; Based on the physical contact network, the probability of infection for each individual in the prevention and control area at the current moment is estimated; Based on the infection probability, a disease transmission loss and an isolation loss are constructed for each sub-network; wherein, the disease transmission loss and isolation loss of each sub-network are both related to the isolation vector of each sub-network; the isolation vector of each sub-network is solved by convex relaxation with the objective of minimizing the weighted sum of the disease transmission loss and the isolation loss; based on the solution of the isolation vector of each sub-network, the individuals that need to be isolated in the region corresponding to each sub-network at the current time are determined; wherein, the isolation vector of each sub-network represents whether each individual in each sub-network should be isolated at the current time; The method of constructing the disease transmission loss for each sub-network based on the infection probability includes: defining the first disease transmission loss for each sub-network as the weighted sum of the probabilities of all individuals in each sub-network at the current moment going from healthy to being infected by internally infected individuals; defining the second disease transmission loss for each sub-network as the weighted sum of the probabilities of all infected individuals in each sub-network at the current moment infecting external healthy individuals; defining the third disease transmission loss for each sub-network as the weighted sum of the probabilities of all individuals in each sub-network at the current moment going from healthy to being infected by externally infected individuals; constructing the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss based on the infection probability and the unassigned isolation vector of each sub-network; and using the sum of the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss as the disease transmission loss for each sub-network. The construction of the isolation loss for each sub-network includes: defining the total number of isolated individuals in each sub-network at the current moment as the isolation loss for each sub-network, and constructing the isolation loss based on the unassigned isolation vector of each sub-network.

2. The isolation and control method for large-scale populations according to claim 1, characterized in that, The process of dividing the physical contact network into multiple sub-networks includes: Based on the principle that the nodes within a subnetwork are tightly connected and the nodes between subnetworks are sparsely connected, the physical contact network is divided into multiple subnetworks. or, The physical contact network is divided into multiple sub-networks.

3. The isolation and control method for large-scale populations according to claim 1, characterized in that, The step of estimating the infection probability of each individual in the prevention and control area at the current moment based on the physical contact network includes: If there are no new confirmed cases in the control area at the current moment, the infection probability of each individual in the control area at the current moment is estimated by using the infection probability of each individual at the previous moment and the physical contact network.

4. The isolation and control method for large-scale populations according to claim 3, characterized in that, The step of estimating the infection probability of each individual in the prevention and control area at the current moment based on the physical contact network includes: If there are newly confirmed cases in the prevention and control area at the current time, the infection probability correction time relative to the prevention and control area is determined based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient at the previous time and every time before that time. The infection probability of each individual in the prevention and control area is corrected sequentially at each time after the infection probability correction time until the corrected value of the infection probability of each individual in the prevention and control area at the previous time is obtained. Using the corrected value of the infection probability of each individual in the control area at the previous moment and the physical contact network, the infection probability of each individual in the control area at the current moment is estimated.

5. The isolation and control method for large-scale populations according to claim 4, characterized in that, The determination of the infection probability correction time relative to the control area, based on the probability distribution of the incubation period of the infectious disease and the infection probability of each newly confirmed patient at the previous time and every time before that time, includes: Based on the probability distribution of the incubation period of infectious diseases, calculate the probability that each newly diagnosed patient was in the latent state at the previous moment and at every moment before that. The probability of infection of each newly diagnosed patient at the previous time and at each time before that time is subtracted from the probability of each newly diagnosed patient being in the latent state at the previous time and at each time before that time, and the time when the difference is greater than 0 is taken as the correction time corresponding to each newly diagnosed patient. The minimum time among all the corrected times corresponding to confirmed patients is taken as the infection probability corrected time.

6. The isolation and control method for large-scale populations according to claim 4, characterized in that, The step of sequentially correcting the infection probability of each individual in the prevention and control area at each time point after the infection probability correction time until the corrected infection probability value of each individual in the prevention and control area at the previous time point includes: For any time between the infection probability correction time and the previous time, based on the correction value of the infection probability of each individual at the previous time, the infection probability of each individual at any time is re-estimated. If each of the individuals is not a newly diagnosed patient, then the infection probability of each individual at any time is corrected to the maximum value of the infection probability of each individual at any time and the re-estimated infection probability of each individual at any time. If each of the individuals is a newly diagnosed patient, then the infection probability of each individual at any given time is revised to the maximum value among the infection probability of each individual at any given time, the re-estimated infection probability of each individual at any given time, and the probability that each individual is in a latent state at any given time.

7. The isolation and control method for large-scale populations according to claim 1, characterized in that, The first disease transmission loss The expression is as follows: ； ； Second disease transmission loss The expression is as follows: ； The third disease transmission loss The expression is as follows: ； in, For nodes The weight corresponding to the individual, For nodes The corresponding individual in The probability of a person going from healthy to being infected by an already infected individual at any given moment. for Time of the first The set of all individuals in the region corresponding to each subnetwork. For nodes The corresponding individual in The probability of infection at any given time. node The corresponding individual in The probability of infection at any given time. For nodes The corresponding individual in The rate of dissemination at any given moment For nodes The corresponding individual in The rate of dissemination at any given moment The adjacency matrix of the physical contact network is the first... Line number Column elements, and The first In the isolation vector of the nth subnetwork The element and the first There are 1 elements, where a value of 1 indicates isolation and a value of 0 indicates no isolation. and They are respectively Nodes before the time and nodes The value indicates whether the individual has been diagnosed with the disease; if so, the value is 0, otherwise the value is 1.

8. A device for isolating and controlling large populations, characterized in that, The device includes: The module is used to construct a physical contact network based on the individuals in the control area at the previous moment and their contact behavior; A partitioning module is used to divide the physical contact network into multiple sub-networks; The prediction module is used to predict the probability of infection of each individual in the prevention and control area at the current moment based on the physical contact network. A determination module is used to construct the disease transmission loss and isolation loss of each sub-network based on the infection probability; wherein the disease transmission loss and isolation loss of each sub-network are related to the isolation vector of each sub-network; the isolation vector of each sub-network is solved by convex relaxation with the objective of minimizing the weighted sum of the disease transmission loss and the isolation loss; based on the solution of the isolation vector of each sub-network, the individuals that need to be isolated in the region corresponding to each sub-network at the current time are determined; wherein the isolation vector of each sub-network represents whether each individual in each sub-network should be isolated at the current time. The determining module is further configured to define the first disease transmission loss of each sub-network as the weighted sum of the probabilities of all individuals in each sub-network at the current moment going from healthy to being infected by internally infected individuals; define the second disease transmission loss of each sub-network as the weighted sum of the probabilities of all infected individuals in each sub-network at the current moment infecting external healthy individuals; define the third disease transmission loss of each sub-network as the weighted sum of the probabilities of all individuals in each sub-network at the current moment going from healthy to being infected by externally infected individuals; construct the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss based on the infection probabilities and the unassigned isolation vector of each sub-network; and use the sum of the first disease transmission loss, the second disease transmission loss, and the third disease transmission loss as the disease transmission loss of each sub-network. The determining module is further configured to define the total number of isolated individuals in each sub-network at the current moment as the isolation loss of each sub-network, and to construct the isolation loss based on the isolation vector of each sub-network that has not been assigned a value.

Images