Research On Dynamics Of Disease Spreading And Isolation Measures In Complex Networks Based On SEIR Model

The frequent occurrence of the epidemic spreading poses a tremendous threat to the safety of human life and causes serious damage to social order.Scholars are actively looking for strategies that can effectively curb the spread of the epidemic.The previous research mainly focused on how to propose effective immunization strategies to suppress the spread of infectious disease with existing vaccines,and for the suppression of infectious diseases for which no effective vaccine has been developed,the potential infection and the susceptible are isolated.During the spread of some typical epidemics,because it is difficult to identify the exposed with infectious ability,large-scale and non-discriminatory isolation can only be taken.Although this can effectively prevent and control the spread of disease,it brings a ponderous burden to the economic society and the resource system for preventing and controlling the spread of diseases.In order to more accurately identify those with the widest reach and the most influence and isolate them to reduce these costs,this thesis uses the Susceptible-Exposed-Infected-Recovered(SEIR)model with infectious exposed individuals,and proposes dynamic isolation measures to optimize the contagion prevention and restraint strategy.The main contents of the thesis are as follows:(1)This thesis uses SEIR model with infectious exposed individuals.Different from the standard SEIR model,the model in this thesis considers the situation where the incubation period is contagious.In addition,two new states are added to the model due to this thesis has proposed dynamic isolation measures:Quarantined susceptible(Q_S)and Quarantined exposed(Q_E).During the quarantine period,the Q_Ewill be removed from the transmission network because it will be transformed into the infected and then become immune to the disease after being treated.After the quarantine period is over,the Q_Sis returned to the transmission network to achieve the goal of paying a small cost and suppressing large-scale spread.(2)Based on the research model of this thesis,two categories total three kinds of dynamic isolation measures are proposed:one is that,if the global information of the network is available,the degree of isolation candidates are used as its weight or the sum of the effective degrees,the sum of the effective betweennesses of the first order neighbors of the isolation candidates,and the sum of effective path lengths to all infected(I state)nodes are taking into account as its weight;the second is for the condition if there is only a certain range of information near the infected node,The sum of same chemotaxis of the number of connected edges between the candidate isolated node and other non-isolated nodes and the number of infected nodes around the candidate isolation node is used as the weight of the isolation candidates.When the isolation capacity is set to small,priority is given to isolating high weight nodes to build a barrier against the spread of diseases.It's showed based on the simulation results that the implementation of dynamic isolation measures can reduce the proportion of final infected nodes in the network.In addition,the key nodes are found out with the isolation measures on account of the global information better than them are found out with the isolation measures on account of the local information of the infected.More than 90%of nodes of the network scale are protected from infection by isolating these key nodes.(3)Through the simulation analysis of the propagation process on ER(Erd(?)s-Rényi)random network and scale-free network,this thesis finds that the two periods in the process of disease spreading be divided with the peak of the number of infected cases as the boundary.That is,the early period of the free spread of the disease and the middle and late periods of the removal and isolation safeguards work together to suppress the spread of the epidemic.Secondly,it was found that there are two characteristics of the node isolation process under the effect of isolation measures.That is,there will be an isolation peak in the early period of spreading and give first place to protect the susceptible,and the exposed are mainly isolated in the later period of spreading;Finally,the disease spreading threshold in scale-free network is smaller than the ER random network,because the heterogeneous performance of the scale-free network will accelerate the spread of the epidemic,and also makes the prevention and restraint of the disease spread more expensive.(4)According to the simulation experiment results of the dynamic isolation measures applied to the model,effective isolation measure application strategies are proposed.In the strategy of identifying candidate priority isolation nodes,it is found that the order of the three major indicators in node weights for identifying key nodes is:degree,betweenness,and shortest path length.The key nodes can be more accurately identified with combining the three indicators than sorting based on a single indicator.In the strategy of exploring the most appropriate scope of quarantine measures,it is found that if the global information of the network is available,it is most suitable to take the nodes in the range of the infected nodes as the center and the path length of?+1(?=Infectious disease incubation period)as the isolation candidates.If there is only local information of the infected nodes,it is most appropriate to take the set of?+1 order neighbor of the infected nodes as the isolation candidates.These findings can help people better understand the spread of emerging infectious disease on human contact networks,and provide some useful ideas and insights for suppressing the spread of disease at the same time.