Airborne infectious disease prediction method based on personnel flow and contact intensity

Abstract

The invention provides an airborne infectious disease prediction method based on personnel flow and contact intensity, and belongs to the field of infectious disease transmission prediction, and the method comprises the steps: S1, carrying out the preprocessing of personnel movement track data, and obtaining the preprocessed track data; s2, based on the preprocessed trajectory data, identifying a contact event between persons in a set time window and a distance threshold, and constructing a contact matrix; s3, establishing a RealSEIR model based on the airborne infectious disease basic data and the prediction demand of the target region; s4, based on the RealSEIR model, establishing a disease model differential equation set; establishing a disease model cabin transmission rule based on the disease model differential equation set and the contact matrix; and S5, based on the infectious disease prediction demand, the contact matrix and the disease model cabin transmission rule, calculating the number of people in each state, and predicting the epidemic trend of airborne infectious diseases. The method can accurately predict the infectious disease transmission trend.

Description

Technical Field

[0001] The present invention belongs to the field of public health and infectious disease transmission prediction, and in particular relates to a method for predicting airborne infectious diseases based on personnel flow and contact intensity. Background Art

[0002] Airborne infectious diseases are diseases that are spread among people through pathogens in the air (such as bacteria, viruses or fungi). These pathogens usually rely on droplets, aerosols or dust particles for transmission, and can be released into the air through daily activities such as breathing, talking, coughing or sneezing, causing infected people to pass the pathogens to susceptible people. Common airborne infectious diseases include influenza, measles, tuberculosis, etc. The characteristics of airborne transmission make this type of infectious disease spread extremely quickly in closed or crowded environments, and due to the widespread nature of air media, the control of the epidemic faces great challenges. Therefore, when the above-mentioned airborne infectious diseases break out, countries usually implement strict public health intervention measures, such as mandatory wearing of masks, isolation of infected people, restriction of movement of people and closure of public places.

[0003] However, current airborne infectious disease prediction methods have obvious limitations: they generally assume that contacts between people are evenly mixed, that is, everyone has the same probability of contact with others. This assumption ignores the unevenness of contact caused by factors such as geography and social networks in real life. In addition, they generally find it difficult to accurately simulate the effects of dynamic intervention measures, such as testing, isolation, and vaccination. Moreover, they generally fail to consider the spatial mobility of the population, and the actual population mobility plays a vital role in the spread of infectious diseases, which makes the model insufficient in predicting the spread of complex infectious diseases. Therefore, how to accurately simulate the actual contact patterns of people, population mobility, and the impact of dynamic intervention measures to accurately predict the spread trend of airborne infectious diseases has become an urgent problem to be solved. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method for predicting airborne infectious diseases based on personnel flow and contact intensity, comprising the following steps:

[0005] Step S1: Preliminary processing is performed on the movement trajectory data of the personnel to filter out the data within the target area and starting from the starting date and having a length of the prediction period; the filtered data is subjected to sparse processing based on the set time interval to obtain the preprocessed trajectory data;

[0006] Step S2: Based on the preprocessed trajectory data, identifying contact events between people within a set time window and distance threshold; based on the contact events, constructing a contact matrix using a hash table;

[0007] Step S3: Based on the basic data of airborne infectious diseases in the target area and combined with the infectious disease prediction needs, a RealSEIR model is established, including: basic cabin structure, cabin transmission link and cabin conversion parameters;

[0008] Step S4: establishing a disease model differential equation group based on the RealSEIR model; establishing a disease model cabin propagation rule based on the disease model differential equation group and the contact matrix;

[0009] Step S5: Based on the predicted infectious disease demand, the contact matrix and the cabin transmission rules of the disease model, the number of people in each state in the target area is calculated, the epidemic trend of airborne infectious diseases in the target area is predicted, and a medical resource adjustment strategy for the target area is formulated.

[0010] Beneficial effects:

[0011] 1. The present invention provides a method for predicting airborne infectious diseases based on personnel flow and contact intensity, which shows obvious advantages over the existing technology. Through the fine processing of personnel movement trajectory data and accurate identification of contact events, the contact matrix constructed with the help of hash tables accurately captures the contact patterns and spatial behavior dynamics between individuals, overcoming the limitation of assuming uniform distribution of contacts in traditional models; by using an efficient hash table data structure, the construction of the contact matrix is ​​limited to actual contact events, which greatly reduces the storage of invalid data and improves the efficiency of the system in large-scale data processing.

[0012] 2. This invention innovatively proposes the RealSEIR (Test & Isolation) model, which successfully realizes the dynamic simulation of isolation measures by introducing detection and isolation mechanisms; by introducing parameters such as detection rate and detection sensitivity, the model can flexibly respond to changes in public health intervention measures, thereby more accurately predicting the development trend of the epidemic; the model is based on a set of differential equations and propagation rules, and can effectively predict key data such as the scale of spread of infectious diseases, infection peaks and end times, providing a scientific basis for the formulation of epidemic prevention and control strategies. These technological innovations not only enhance the model's ability to predict actual epidemics, but also provide solid technical support for public health decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a flow chart of a method for predicting airborne infectious diseases based on personnel flow and contact intensity of the present invention;

[0014] Figure 2 Schematic diagram of the RealSEIR (Test & Isolation) model. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0016] Embodiment 1

[0017] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting airborne infectious diseases based on personnel flow and contact intensity, comprising the following steps:

[0018] Step S1: Preliminary processing is performed on the movement trajectory data of the personnel to filter out the data within the target area and starting from the starting date and having a length of the prediction period; the filtered data is subjected to sparse processing based on the set time interval to obtain the preprocessed trajectory data;

[0019] Step S2: Based on the preprocessed trajectory data, identifying contact events between people within a set time window and distance threshold; based on the contact events, constructing a contact matrix using a hash table;

[0020] Step S3: Based on the basic data of airborne infectious diseases in the target area and combined with the infectious disease prediction needs, a RealSEIR model is established, including: basic cabin structure, cabin transmission link and cabin conversion parameters;

[0021] Step S4: Based on the RealSEIR model, a disease model differential equation group is established; based on the disease model differential equation group and the contact matrix, a disease model cabin propagation rule is established;

[0022] Step S5: Based on the infectious disease prediction demand, contact matrix and disease model cabin transmission rules, calculate the number of people in each state in the target area, predict the epidemic trend of airborne infectious diseases in the target area, and formulate a medical resource adjustment strategy for the target area.

[0023] In one embodiment, the above step S1: preliminarily processes the movement trajectory data of the personnel, filters out the data within the target area and starting from the start date and having a length of the prediction period; performs sparse processing on the filtered data based on a set time interval to obtain preprocessed trajectory data, specifically including:

[0024] Step S11: obtaining the movement trajectory data of the personnel, including: personnel ID, latitude, longitude, and time; deleting the trajectory data containing null values;

[0025] After obtaining the personnel movement trajectory data, in order to prevent the null values ​​from affecting the stability and accuracy of the algorithm, it is necessary to filter the data corresponding to the above four fields to filter out the data containing null values;

[0026] Step S12: acquiring trajectory data within the target area according to the latitude and longitude range of the target area;

[0027] This step can extract trajectory data only within the target area, thereby ensuring the geographical relevance of subsequent analysis and prediction and avoiding interference from irrelevant data.

[0028] Step S13: according to the start date start_date and the prediction period T, obtaining the trajectory data within the target time period in the trajectory data within the target area;

[0029] This step can extract trajectory data that is only within the target time period, thereby ensuring the temporal consistency of subsequent analysis and prediction and avoiding interference from irrelevant data.

[0030] In the embodiment of the present invention, start_date is selected as a certain day (eg, Monday) within the time covered by the trajectory data, T is set to 7 days, and the trajectory data within a complete week in the target area is filtered out.

[0031] Step S14: sampling the trajectory data within the target time period according to the set time interval time_interval to obtain pre-processed trajectory data.

[0032] The specific steps of sampling processing are as follows: First, arrange the trajectory data of each person in chronological order. Then, within each time_interval, select a trajectory point closest to the start time of the time_interval and retain it as a representative point. For other trajectory points within the time interval, they are discarded. Finally, the preprocessed trajectory data is obtained.

[0033] This step can effectively reduce the amount of data while retaining key information about the movement trajectory of people. The sparsely processed trajectory data not only reduces the complexity of storage and calculation, but also ensures the accuracy of subsequent contact event identification and prediction of the spread trend of airborne infectious diseases.

[0034] In the embodiment of the present invention, time_interval is set to 1 minute, that is, at each minute time node, the trajectory point of the person closest to the current time node within 1 minute is retained, and other trajectory points of the person within 1 minute are discarded.

[0035] In one embodiment, the above step S2: based on the preprocessed trajectory data, identifying contact events between people within a set time window and distance threshold; based on the contact events, using a hash table to construct a contact matrix, specifically includes:

[0036] Step S21: Identify and record contact events between people according to the set time window time_window and distance threshold distance_threshold; where a contact event is: within the time_window, if the geographical distance between person i and person j is less than or equal to distance_threshold, it is determined that the two have contacted, and a contact event contact_event is generated; contact_event includes {User i , User j , Date, Distance, Center_Latitude, Center_Longitude}, where Date is the day when the contact occurs in the prediction period T ( ); Distance is the distance between two users i and User j Center_Longitude and Center_Latitude are the latitude and longitude of the center point of the contact event;

[0037] In the embodiment of the present invention, time_window is set to 5 minutes and distance_threshold is set to 50 meters. The 7 days are subdivided into 2016 time periods of 5 minutes in length, and in each time period, all personnel are analyzed one by one, and all other personnel j whose distance from the current person i is less than or equal to distance_threshold are identified, and then these contact events (contact_event) are recorded. Finally, all contact events are generated.

[0038] Step S22: define the frequency of contact events as the contact intensity between persons i and j , using a hash table, based on Construct a contact matrix H whose keys are combinations of pairs of time and person numbers , the value is the contact intensity between personnel i and personnel j .

[0039] Since the hash table only stores pairs of people who actually have contact, it avoids storing a lot of invalid data and greatly reduces memory usage. In addition, the hash table can be dynamically expanded without the need to pre-allocate a huge memory space, and has high memory efficiency, ensuring that the system can still run efficiently in infectious disease prediction tasks involving large numbers of people.

[0040] In the embodiment of the present invention, if on the tth day in the prediction period, the contact intensity between persons i and j is , the IDs of the personnel are and , then H The value of .

[0041] In one embodiment, the above step S3: based on the basic data of airborne infectious diseases in the target area and combined with the infectious disease prediction needs, a RealSEIR model is established, including: basic cabin structure, cabin transmission link and cabin conversion parameters, specifically including:

[0042] Step S31: Obtain basic data of airborne infectious diseases in the target area, including: population data, case data and hospital bed data in the target area;

[0043] In the embodiment of the present invention, the case data comes from the statistical data of the Health Commission of the target area, and the population data and hospital bed data come from the China City Statistical Yearbook published by the National Bureau of Statistics.

[0044] Step S32: Based on the basic data of airborne infectious diseases and in combination with the needs of infectious disease prediction, a basic cabin structure of the model is established; wherein the basic cabin structure includes multiple crowd cabins, and the crowd cabins include: a susceptible crowd cabin, an exposed crowd cabin, an exposed isolation crowd cabin, an infected crowd cabin, an infected isolation crowd cabin, a recovered crowd cabin, and a dead crowd cabin;

[0045] Among them, the characteristics of the people in each cabin are as follows:

[0046] Susceptible population cabin (Susceptible): healthy people who are susceptible to airborne infectious diseases;

[0047] Exposed: People who are infected with pathogens but not yet contagious, in the incubation period;

[0048] Quarantined Exposed: People who have been infected with pathogens but are not yet contagious and have been isolated. They are in the incubation period and are isolated from the general population through isolation measures;

[0049] Infected: People who are infected and contagious;

[0050] Quarantined: Infected and contagious people who have been isolated have limited infection scope and significantly reduced transmission risk.

[0051] Recovered: People who have recovered from the infection are no longer contagious and have permanent immunity.

[0052] Deceased: People who died after infection and no longer participate in the disease transmission process;

[0053] In an embodiment of the present invention, based on the basic data of airborne infectious diseases A in the target area and combined with the prediction needs, the basic cabin structure selected for establishing the model includes: cabin for susceptible people, cabin for exposed people, cabin for exposed isolated people, cabin for infected people, cabin for infected isolated people, cabin for recovered people, and cabin for dead people.

[0054] Step S33: Based on the basic cabin structure, a cabin propagation link is established, including multiple personnel transfer paths between cabins:

[0055] The transfer path in the embodiment of the present invention includes:

[0056] The transfer path of the first cabin: susceptible-exposed-infected-recovered;

[0057] The transfer path of the second compartment: susceptible-exposed-infected-dead;

[0058] The transfer path of the third compartment: susceptible person-exposed person-infected person-infected isolation person-recovered person;

[0059] Transfer path for the fourth compartment: susceptible person - exposed person - infected person - infected person in isolation - deceased person;

[0060] The transfer path of the fifth compartment: susceptible person-exposed person-exposed isolation person-infected isolation person-recovered person;

[0061] The transfer path of the sixth compartment: susceptible person - exposed person - exposed person in isolation - infected person in isolation - deceased person;

[0062] In the embodiment of the present invention, based on the established basic cabin structure of the model and according to the possible disease transmission process, all the above-mentioned transmission links between the model cabins are selected and constructed.

[0063] Step S34: Based on the basic cabin structure and cabin transmission link, combined with the basic data of airborne infectious diseases and infectious disease prediction needs, set the cabin conversion parameters of the RealSEIR model, including: infection rate , Isolation infection rate , the rate at which exposed people become infected σ , the rate at which exposed quarantined persons become infected quarantined persons , the rate at which infected people recover , the recovery rate of infected quarantined people , infectious disease detection rate among latent population , infectious disease detection rate among infected people , Lurker Detection Sensitivity , Sensitivity of infection detection , mortality rate among infected persons , mortality rate of infected people in isolation , Isolation contact frequency adjustment factor .

[0064] In the embodiment of the present invention, based on the basic cabin structure and cabin transmission link of the established model, combined with the basic data and prediction requirements of airborne infectious diseases A in the target area, all the above cabin conversion parameters are selected. At this point, the RealSEIR (Test & Isolation) model is established.

[0065] Figure 2 A schematic diagram of the RealSEIR (Test & Isolation) model is shown.

[0066] In one embodiment, the above step S4: establishing a disease model differential equation group based on the RealSEIR model; establishing a disease model cabin propagation rule based on the disease model differential equation group and the contact matrix, specifically includes:

[0067] Step S41: Based on the RealSEIR model, a disease model differential equation group is established, including:

[0068] (1)

[0069] (2)

[0070] (3)

[0071] (4)

[0072] (5)

[0073] (6)

[0074] (7)

[0075] (8)

[0076] Among them, S(t) is the number of susceptible people at time t, E(t) is the number of exposed people at time t, Q E (t) is the number of people in isolation at time t, I(t) is the number of infected people at time t, Q I (t) is the number of infected people in isolation at time t, R(t) is the number of recovered people at time t, and F(t) is the number of deaths at time t; N(t) is the total number of people at time t;

[0077] In an embodiment of the present invention, based on the established RealSEIR (Test & Isolation) model, corresponding differential equations are established to characterize the model, and the basic cabin structure, cabin propagation link and cabin conversion parameters of the comprehensive model are used to select all equations in the above-mentioned disease model differential equation group.

[0078] Step S42: Based on the disease model differential equations and the contact matrix, establish the disease model cabin propagation rules, including:

[0079] 1) Susceptible person S LurkerE

[0080] When in contact with Infected Person I: (9)

[0081] Q I On contact: (10)

[0082] in, It's a staff and infected persons / infected quarantined persons The contact intensity on the tth day in the prediction period; if the personnel The status is susceptible S, and any infected person / infected isolation Contact strength > 0, then The probability that personnel Spread to exposed person E;

[0083] 2) Lurker E Infected I

[0084] (11)

[0085] If the staff The state is lurker E, then with a fixed probability σ Spread to infected person I;

[0086] 3) Lurker E Isolation Lurker Q E

[0087] (12)

[0088] If the staff The state is lurker E, then the detection rate and detection sensitivity The product probability is propagated to the exposed isolate Q E ;

[0089] 4) Infected I Infected Isolator Q I

[0090] (13)

[0091] If the staff The status is infected person I, then the detection rate and detection sensitivity The product probability of spreading to the infected isolated person Q I ;

[0092] 5) Infected I Recovered

[0093] (14)

[0094] If the staff If the state is infected person I, then with a fixed probability Spread to recovered person R;

[0095] 6) Infected I Death F

[0096] (15)

[0097] If the staff The state is infected person I, then with a fixed probability Spread to the deceased person F;

[0098] 7) Isolate Lurker Q E Infected Isolator Q I

[0099] (16)

[0100] If the staff The status is isolated lurker Q E , then with a fixed probability Spread to infected quarantined person Q I ;

[0101] 8) Infected Isolator Q I Recovered

[0102] (17)

[0103] If the staff The status is infected quarantined person Q I , then with a fixed probability Spread to recovered person R;

[0104] 9) Infected Isolator Q I Death F

[0105] (18)

[0106] If the staff The status is infected quarantined person Q I , then with a fixed probability Spread to the deceased F.

[0107] In an embodiment of the present invention, all the above-mentioned cabin propagation rules are selected based on the disease model differential equations and the contact matrix.

[0108] In one embodiment, the above step S5: based on the infectious disease prediction demand, the contact matrix and the disease model cabin transmission rules, the number of people in each state in the target area is calculated, the epidemic trend of airborne infectious diseases in the target area is predicted, and a medical resource adjustment strategy for the target area is formulated, which specifically includes:

[0109] Step S51: Based on the infectious disease prediction requirements, set the total number of days for the prediction duration , the proportion of infected people in the initial state , the proportion of infected people to the total number of survivors at the end of the infection ; Based on the contact matrix, the total number of personnel is obtained ;

[0110] In the embodiment of the present invention, the total number of days of the forecast duration is set based on the forecast demand: , the proportion of infected people in the initial state , the proportion of infected people to the total number of survivors at the end of the infection (as a criterion for determining the end of the infection) ; Based on the contact matrix, the total number of personnel is obtained .

[0111] Step S52: Initialize personnel categories: S, E, Q E , I, Q I , R, F, generate Personnel, in proportion Randomly select a portion of people and initialize their status as infected persons I, and the remaining unselected people are initialized as susceptible persons S;

[0112] In the embodiment of the present invention, the generated Personnel in proportion Randomly select a portion of people and initialize their status as infected persons I, and the remaining unselected people are initialized as susceptible persons S;

[0113] Step S53: Substitute the cabin conversion parameters into the cabin propagation rules of the disease model, and execute the STEPS round cycle in sequence. In the nth round cycle, the parameters in the hash table need to be substituted. Finally, the number of people in different categories in the target area within STEPS days is calculated, including: the number of susceptible people, the number of exposed people, the number of exposed and isolated people, the number of infected people, the number of infected and isolated people, the number of recovered people, and the number of deaths;

[0114] In the embodiment of the present invention, the cabin conversion parameters are set as follows:

[0115] Table 1 Cabin conversion parameters

[0116]

[0117] Substitute the above cabin conversion parameters into the cabin propagation rules of the disease model, and execute 250 rounds of loops in sequence. In the nth round of loop, the parameters in the hash table need to be substituted. For example, if the forecast period is T=7 days, and the current cycle is the 10th cycle n=10, then the calculation process of t is as follows: This means that in the 10th cycle, the data of the 4th day in the corresponding prediction period in the hash table needs to be used. Similarly, according to this calculation method, the data in the prediction period is used cyclically throughout the prediction period to represent the periodically changing contact intensity during the entire prediction period.

[0118] Finally, the number of people in different categories in the target area within 250 days was calculated, among which we selected the number of people in all different categories mentioned above.

[0119] Step S54: Based on the number of people of different categories in the target area within STEPS days, key data indicators reflecting the scale of infectious disease spread are calculated, including:

[0120] Basic reproduction number: , based on the traditional basic reproduction number, the contact intensity between people is introduced to more accurately measure the transmission ability of infectious diseases without intervention;

[0121] Peak infection number and time: predict the maximum number of infected people I and their corresponding time point within the duration, and evaluate the peak number of infections under isolation measures;

[0122] Peak number of infected people in quarantine and time: Q of infected people in quarantine within the predicted time I The maximum value of the number and its corresponding time point are used to evaluate the peak number of isolated infections under isolation measures;

[0123] Total peak infection number: peak infection number + peak quarantine infection number, reflecting the highest infection burden during the epidemic;

[0124] Average doubling time: , the average time required for the number of infected people to double, reflecting the rate of increase in the number of infected people during the entire forecast period;

[0125] Herd immunity establishment time: The number of recovered people R reaches the total number of surviving people, that is, NF, and the ratio is The time point indicates the time of formation of herd immunity; if the time of establishment of herd immunity is greater than the time of end of infection, it has no practical significance and its value is set to null;

[0126] Infection end time: Infected I and Infected Isolator Q I The total number of people reaches the proportion of the total surviving population The time point indicates that the epidemic is basically over;

[0127] In the embodiment of the present invention, based on the number of different categories of people in the target area within 250 days, key data indicators reflecting the scale of infectious disease spread are calculated, and all the above data indicators are selected. The calculation results are as follows:

[0128] Table 2 Calculation results of key data indicators reflecting the scale of infectious disease spread

[0129]

[0130] Step S55: Based on the number of people of different categories in the target area within STEPS days, combined with key data indicators reflecting the scale of infectious disease transmission, predict the epidemic trend of airborne infectious diseases in the target area;

[0131] Step S56: Based on the epidemic trend of airborne infectious diseases in the target area, formulate a medical resource adjustment strategy for the target area, including: increasing hospital beds and increasing the frequency of airborne infectious disease detection.

[0132] The above embodiments are provided only for the purpose of describing the present invention, and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims. Various equivalent substitutions and modifications made without departing from the spirit and principles of the present invention should all be included within the scope of the present invention.

Claims

1. A method for predicting airborne infectious diseases based on personnel flow and contact intensity, characterized in that: include: Step S1: Preliminary processing of the movement trajectory data of personnel is performed to filter out the data within the target area and starting from the starting date and with a length equal to the prediction period; The filtered data is subjected to sparse processing based on a set time interval to obtain preprocessed trajectory data; Step S2: Based on the preprocessed trajectory data, identifying contact events between people within a set time window and distance threshold; Based on the contact events, a contact matrix is ​​constructed using a hash table; Step S3: Based on the basic data of airborne infectious diseases in the target area and combined with the infectious disease prediction needs, a RealSEIR model is established, including: basic cabin structure, cabin transmission link and cabin conversion parameters; Step S4: establishing a disease model differential equation group based on the RealSEIR model; establishing a disease model cabin propagation rule based on the disease model differential equation group and the contact matrix; Step S5: Based on the predicted infectious disease demand, the contact matrix and the cabin transmission rules of the disease model, the number of people in each state in the target area is calculated, the epidemic trend of airborne infectious diseases in the target area is predicted, and a medical resource adjustment strategy for the target area is formulated.

2. The method for predicting airborne infectious diseases based on personnel flow and contact intensity according to claim 1, characterized in that: The step S1: preliminarily processes the movement trajectory data of the personnel, and filters out the data within the target area and starting from the starting date and having a length equal to the prediction period; The filtered data is sparsely processed based on the set time interval to obtain the preprocessed trajectory data, including: Step S11: obtaining the movement trajectory data of the personnel, including: personnel ID, latitude, longitude, and time; deleting the trajectory data containing null values; Step S12: acquiring trajectory data within the target area according to the longitude and latitude range of the target area; Step S13: according to the start date start_date and the prediction period T, obtaining the trajectory data within the target time period in the trajectory data within the target area; Step S14: sampling the trajectory data within the target time period according to the set time interval time_interval to obtain pre-processed trajectory data.

3. The method for predicting airborne infectious diseases based on personnel flow and contact intensity according to claim 2, characterized in that: The step S2: based on the pre-processed trajectory data, identifying contact events between people within a set time window and distance threshold; Based on the contact events, a contact matrix is ​​constructed using a hash table, specifically including: Step S21: Identify and record contact events between people according to the set time window time_window and distance threshold distance_threshold; wherein the contact event is: within the time_window, if the geographical distance between person i and person j is less than or equal to distance_threshold, it is determined that the two have contacted, and a contact event contact_event is generated; contact_event includes {User i , User j , Date, Distance, Center_Latitude, Center_Longitude}, where Date is the day when the contact occurs in the prediction period T ( ); Distance is the distance between two users i and User j Center_Longitude and Center_Latitude are the latitude and longitude of the center point of the contact event; Step S22: define the frequency of contact events as the contact intensity between persons i and j , using a hash table, based on Construct a contact matrix H whose keys are combinations of pairs of time and person numbers , the value is the contact intensity between personnel i and personnel j .

4. The method for predicting airborne infectious diseases based on personnel flow and contact intensity according to claim 3, characterized in that: Step S3: Based on the basic data of airborne infectious diseases in the target area and combined with the infectious disease prediction needs, a RealSEIR model is established, including: basic cabin structure, cabin transmission link and cabin conversion parameters, specifically including: Step S31: Obtain basic data of airborne infectious diseases in the target area, including: population data, case data and hospital bed data in the target area; Step S32: Based on the basic data of airborne infectious diseases and in combination with the infectious disease prediction needs, a basic cabin structure of the model is established; wherein the basic cabin structure includes a plurality of crowd cabins, and the crowd cabins include: a susceptible crowd cabin, an exposed crowd cabin, an exposed isolation crowd cabin, an infected crowd cabin, an infected isolation crowd cabin, a recovered crowd cabin, and a dead crowd cabin; Step S33: Based on the basic cabin structure, a cabin communication link is established, including multiple personnel transfer paths between cabins: Step S34: Based on the basic cabin structure and the cabin transmission link, combined with the basic data of airborne infectious diseases and the infectious disease prediction requirements, the cabin conversion parameters of the RealSEIR model are set, including: infection rate , Isolation infection rate , the rate at which exposed people become infected σ , the rate at which exposed quarantined persons become infected quarantined persons , the rate at which infected people recover , the recovery rate of infected quarantined people , infectious disease detection rate among latent population , infectious disease detection rate among infected people , Lurker Detection Sensitivity , Sensitivity of infection detection , mortality rate among infected persons , mortality rate of infected people in isolation , Isolation contact frequency adjustment factor .

5. The method for predicting airborne infectious diseases based on personnel flow and contact intensity according to claim 4, characterized in that: The step S4: establishing a disease model differential equation group based on the RealSEIR model; Based on the disease model differential equation group and the contact matrix, a disease model cabin propagation rule is established, which specifically includes: Step S41: Based on the RealSEIR model, a disease model differential equation group is established, including: (1) (2) (3) (4) (5) (6) (7) (8) Among them, S(t) is the number of susceptible people at time t, E(t) is the number of exposed people at time t, Q E (t) is the number of people in isolation at time t, I(t) is the number of infected people at time t, Q I (t) is the number of infected people in isolation at time t, R(t) is the number of recovered people at time t, and F(t) is the number of deaths at time t; N(t) is the total number of people at time t; Step S42: Based on the disease model differential equation group and the contact matrix, a disease model cabin propagation rule is established, including: 1) Susceptible person S LurkerE When in contact with Infected Person I: (9) Isolate from infected people Q I On contact: (10) in, It's a staff and infected persons / infected quarantined persons The contact intensity on the tth day in the prediction period; if the personnel The status is susceptible S, and any infected person / infected isolation Contact strength > 0, then The probability that personnel Spread to exposed person E; 2) Lurker E Infected person I (11) If the staff The state is lurker E, then with a fixed probability σ Spread to infected person I; 3) Lurker E Isolation Lurker Q E (12) If the staff The state is lurker E, then the detection rate and detection sensitivity The product probability is propagated to the exposed isolate Q E ; 4) Infected I Infected Isolator Q I (13) If the staff The status is infected person I, then the detection rate and detection sensitivity The product probability of spreading to the infected isolated person Q I ; 5) Infected I Recovered (14) If the staff The state is infected person I, then with a fixed probability Spread to recovered person R; 6) Infected I Death F (15) If the staff The state is infected person I, then with a fixed probability Spread to the deceased person F; 7) Isolate Lurker Q E Infected Isolator Q I (16) If the staff The status is isolated lurker Q E , then with a fixed probability Spread to infected quarantined person Q I ; 8) Infected Isolator Q I Recovered (17) If the staff The status is infected quarantine Q I , then with a fixed probability Spread to recovered person R; 9) Infected Isolator Q I Death F (18) If the staff The status is infected quarantine Q I , then with a fixed probability Spread to the deceased F.

6. The method for predicting airborne infectious diseases based on personnel flow and contact intensity according to claim 5, characterized in that: The step S5: based on the predicted infectious disease demand, the contact matrix and the cabin transmission rules of the disease model, the number of people in each state in the target area is calculated, the epidemic trend of airborne infectious diseases in the target area is predicted, and a medical resource adjustment strategy for the target area is formulated, which specifically includes: Step S51: Based on the infectious disease prediction requirements, set the total number of days for the prediction duration , the proportion of infected people in the initial state , the proportion of infected people to the total number of survivors at the end of the infection ; Based on the contact matrix, the total number of personnel is obtained ; Step S52: Initialize personnel categories: S, E, Q E , I, Q I , R, F, generate Personnel, in proportion Randomly select a portion of people and initialize their status as infected persons I, and the remaining unselected people are initialized as susceptible persons S; Step S53: Substitute the cabin conversion parameters into the cabin propagation rules of the disease model, and execute the STEPS round cycle in sequence, wherein the nth round cycle needs to substitute the , and finally the number of people of different categories in the target area within STEPS days is calculated; Step S54: Based on the number of people of different categories in the target area within STEPS days, key data indicators reflecting the scale of infectious disease spread are calculated, including: Basic reproduction number: , based on the traditional basic reproduction number, the intensity of contact between people is introduced to more accurately measure the transmission capacity of infectious diseases without intervention; Peak infection number and time: predict the maximum number of infected people I and their corresponding time point within the duration, and evaluate the peak number of infections under isolation measures; Peak number of infected people in quarantine and time: Q of infected people in quarantine within the predicted time I The maximum value of the number and its corresponding time point are used to evaluate the peak number of isolated infections under isolation measures; Total peak infection number: peak infection number + peak quarantine infection number, reflecting the highest infection burden during the epidemic; Average doubling time: , the average time required for the number of infected people to double, reflecting the rate of increase in the number of infected people during the entire forecast period; Herd immunity establishment time: The number of recovered people R reaches the total number of surviving people, that is, NF, and the ratio is The time point indicates the time of formation of herd immunity; if the time of establishment of herd immunity is greater than the time of end of infection, it has no practical significance and its value is set to null; Infection end time: Infected I and Infected Isolator Q I The total number of people reaches the proportion of the total surviving population The time point indicates that the epidemic is basically over; Step S55: Based on the number of people of different categories in the target area within STEPS days, combined with the key data indicators reflecting the scale of infectious disease transmission, predict the epidemic trend of airborne infectious diseases in the target area; Step S56: Based on the epidemic trend of airborne infectious diseases in the target area, formulate a medical resource adjustment strategy for the target area, including: increasing hospital beds and increasing the frequency of airborne infectious disease detection.

Images