Epidemic Model Based On Individual Heterogeneity And Its Application To The COVID-19 Pandemic

In time-varying networks,nodes or connected edges change continuously with time.Compared with static networks,time-varying networks can more accurately describe the dynamic process of real networks.Therefore,time-varying networks have received more and more attention from researchers.Especially in the area of propagation dynamics,scholars have made some important progress.However,there are still some unresolved issues that need further in-depth research and exploration.For example,how does the correlation between node attributes affect the topology of timevarying networks and the dynamical processes on them? How do changes in individual migration pathways affect the transmission processes of epidemics over time-varying population networks? How to prevent and control the spread of COVID-19 in large cities with diverse individual travel patterns? In light of this,this thesis provides insight into the impact of individual behavioral heterogeneity on transmission dynamics as follows:Firstly,the effect of individual active-attractive potential correlation in time-varying networks on communication dynamics is investigated.A temporal network model is proposed where individual activity potential is correlated with attractiveness potential,and the strength of this correlation is studied for its effects on disease spread and rumor diffusion.Theoretical analysis and numerical simulations reveal distinct impacts of temporal dynamics and correlations on the two types of spread dynamics.Specifically,temporal dynamics greatly suppress disease spread but promote rumor diffusion to some extent.The stronger the correlation between activity and attractiveness potentials in the temporal network,the faster the disease outbreak but the more suppressed the rumor diffusion.Furthermore,the effects of nodes on disease spread and rumor diffusion are studied from the perspective of maximizing the propagation influence.The results show that central nodes in temporal networks greatly promote disease spread but significantly suppress rumor diffusion,whereas nodes with low activity and attractiveness potentials can promote rumor diffusion to some extent.This work highlights the importance of individual heterogeneity and correlation in different spread dynamics and can aid in controlling the spread process in real temporal networks.Secondly,the spreading dynamics in time-varying population networks with active potential-attracting potential correlation are explored.A time-varying population network model is proposed,and the impact of the correlation between individual time-varying migration paths and the correlation between activity and attractiveness on propagation dynamics is investigated.The study derived the global migration threshold for disease outbreaks and found that increasing the correlation between activity and attractiveness can lower the outbreak threshold and suppress outbreak size.The study also explored the impact of non-pharmaceutical interventions(self-isolation and self-protection)on epidemic transmission in different correlation networks.Nonpharmaceutical interventions are more effective in reducing disease transmission on negatively correlated networks than on positively correlated networks.These findings provide useful insights into the impact of the correlation between node attributes on the dynamics of time-varying population networks and valuable recommendations for epidemic prevention and control.Furthermore,the effect of individual travel on epidemic transmission in the city is studied.Using actual transportation data and individual travel modes,the study constructs a multi-layer commuting network including commuting methods,commuting infections,and fixed commuting to explore the impact of population movement on epidemic spread within cities.The model fits well with COVID-19 data from Wuhan in 2020 and Shanghai in 2022,and confirms the importance of commuting infections.The study finds that an increase in the number of people commuting across districts in a city leads to a larger outbreak,and that commuting on the subway layer contributes the most to the spread of the epidemic.Interestingly,an increase in fixed commuting reduces the final size of high transmission rate diseases but increases the final size of low transmission rate diseases.The study also suggests that reducing the proportion of subway traffic and implementing a negative correlation traffic allocation strategy can effectively control the outbreak of epidemics.This work provides practical methods and strategies for predicting and preventing epidemics by establishing a multi-layer transportation network within cities.Finally,the effect of non-pharmacological interventions on the transmission of COVID-19 was quantified using a traffic hierarchical ensemble population network transmission model.Based on Shanghai’s commuting data,the model accurately predicted the development of the epidemic,particularly the infection status of each district.Results showed that NPIs implemented at the societal level,such as increasing the efficiency of nucleic acid testing,can significantly reduce the number of infections.Large population administrative regions require strict control measures,and the effectiveness of lockdown measures decreases with increasing infection rates.Additionally,the study found that after lifting restrictions,different regions had varied epidemic developments,with densely populated and economically developed areas reaching infection peaks first.This research provides practical methods and strategies for predicting and controlling epidemics,deepening the understanding of the impact of urban social and district controls on the spread of infectious diseases.In summary,this thesis not only studied the impact of correlations between node attributes in time-varying networks on the dynamics of propagation,but also explored the influence of individual time-varying transition paths on the spread of epidemics in a collective population network.The thesis used complex network analysis methods to study the intrinsic mechanisms of COVID-19 transmission in large cities with different individual travel patterns.Through in-depth research on network propagation dynamics based on individual behavioral heterogeneity,this paper thesis deepen scholars’ understanding of network propagation dynamics and provides important theoretical references for predicting and controlling the spread of epidemics.