融合SEIR与LSTM模型的传染病预测研究

针对现有的传染病预测模型未充分考虑时间序列的复杂度且预测性能不稳定等问题,提出一种基于传染病动力学模型SEIR与长短时记忆网络(LSTM)的传染病组合预测模型.首先,通过计算Pearson相关系数分析气候因素与传染病新增人数之间的相关性;其次,通过FE(Fuzzy Entropy)算法提取序列的局部特征且保证序列的平稳性,降低时间序列的复杂度,提升时间序列的可预测性;最后,根据传染病特点,构建SEIR模型分析不同人群传播情况,并结合LSTM模型实现大幅度提升传染病预测精度.仿真结果表明,相较于传统的模型算法,本文提出的混合模型能保证预测的平稳性并实现更高的预测精度,同时,本文使用该混合模型在不同的干预策略下进行预测,表明了提早采取防控措施对遏制传染病传播的重要性.

GLOBAL DYNAMICS OF AN SEIR EPIDEMIC MODEL WITH IMMIGRATION OF DIFFERENT COMPARTMENTS

The SEIR epidemic model studied here includes constant inflows of new susceptibles, exposeds, infectives, and recovereds. This model also incorporates a population size dependent contact rate and a disease-related death. As the infected fraction cannot be eliminated from the population, this kind of model has only the unique endemic equilibrium that is globally asymptotically stable. Under the special case where the new members of immigration are all susceptible, the model considered here shows a threshold phenomenon and a sharp threshold has been obtained. In order to prove the global asymptotical stability of the endemic equilibrium, the authors introduce the change of variable, which can reduce our four-dimensional system to a three-dimensional asymptotical autonomous system with limit equation.

Global Asymptotical Stability of an SEIR Model with Random Perturbation

Stochasticity is introduced into a susceptible-exposed but not infectious-infectious-removed （SEIR） model describing epidemics＇ transmission, via the technique of parameter perturbation which is standard in stochastic population modeling. The existence and uniqueness of the model have been proved in this paper. And E detailed analysis on global asymptotic stability is also carried out.

Global Analysis of an SEIR Epidemic Model with Infectious Force under Intervention Strategies

In this paper, we investigate the global stability of an SEIR (Susceptible-Exposed-Infected-Remove) epidemic model with infectious force under intervention strategies. To address this issue, we prove that the basic reproduction number R0 plays an essential role in determining whether the disease extincts or persists. If , there is a unique disease-free equilibrium point of the model which is globally asymptotically stable and the disease dies out, and if , there exists a unique endemic equilibrium point which is globally asymptotically stable and the disease persists.

Global Stability of SEIRS Model in Epidemiology

In this article, an infectious model with saturation effect is considered. By using compound matrix theory and a series of theorems associated to qualitative theory of differential equations which are introduced by Muldowney and Micheal Li, we study globally stable problem of the model.

SEIR Model and Simulation for Vector Borne Diseases

An epidemic model is a simplified means of describing the transmission of infectious diseases through individuals. The modeling of infectious diseases is a tool which has been used to study the mechanisms by which diseases spread, to predict the future course of an outbreak and to evaluate strategies to control an epidemic. Epidemic models are of many types. Here, SEIR model is discussed. We first discuss the basics of SEIR model. Then it is applied for vector borne diseases. Steady state conditions are derived. A threshold parameter R0 is defined and is shown that the disease will spread only if its value exceeds 1. We have applied the basic model to one specific diseases-malaria and did the sensitivity analysis too using the data for India. We found sensitivity analysis very important as it told us the most sensitive parameter to be taken care of. This makes the work more of practical use. Numerical simulation is done for vector and host which shows the population dynamics in different compartments.

Global Analysis of an SEIR Epidemic Model with Nonlinear Incidence Rates

In this paper,an SEIR model with nonlinear incidence rates are studied.The basic reproduction number R_0 characterizes the disease transmission dynamics: if R_0≤ 1,the disease-free equilibrium is globally asymptotically stable and the disease always dies out,if R_0> 1 then there is a unique endemic equilibrium which is globally asymptotically stable and the disease persists.

Global Analysis of an SEIR Epidemic Model with a Ratio-Dependent Nonlinear Incidence Rate

In this paper, a SEIR model with ratio-dependent transmission rate in the form ?is studied and the basic reproduction number which determines the disease’s extinction or continued existence is obtained. By constructing the proper Lyapunov function, we prove that if R0 ≤ 1, the disease-free equilibrium point of the model is globally asymptotically stable and the disease always dies out;if R0 > 1, the endemic equilibrium point is globally asymptotically stable and the disease persists.

一类具有标准发生率的SEIR传染病模型的稳定性分析

根据肺结核(TB)的传播机理,建立了一类具有标准发生率的TB传染病SEIR模型,并讨论了该模型的稳定性。通过常数变易法和反证法证明了模型的正向不变集;利用下一代矩阵法计算得到模型的基本再生数R_(0);通过构造Lyapunov函数法证明了当R_(0)≤1时无病平衡点D_(0)是全局渐近稳定的;利用Hurwitz判据证明了当R_(0)>1时地方病平衡点D_(*)是局部渐近稳定的,且借助Li-Mulowney几何方法给出了地方病平衡点D*全局渐近稳定的条件;最后,通过数值模拟验证了所得结论的有效性.

数据与SEIR模型双驱动下COVID-19疫苗的空间分配策略

为提出一个适合疫情中后期新冠疫苗空间分配的新框架,通过收集2022年11月13日至28日国内31个省会城市的新冠疫情数据,求解传染病模型的逆问题得到初始潜伏人数,借助K-means算法对城市风险等级进行分类,进而根据风险分级后的城市建立疫苗分配模型,并通过遗传算法求解得到最优分配策略。数据结果表明,31个省会城市可分为5个疫情风险等级区域,风险等级越高的地区,疫苗在待接种人群的覆盖比例越高;在降低累计感染人数方面,所提出的分配策略优于地区人数比例分配策略和风险等级优先全覆盖分配策略。

Persistence in a Stochastic SEIR Model with Constant Immigration in Incubation Period

A kind of stochastic susceptible-exposed but not infectiousinfectious-removed (SEIR) model with constant immigration in incubation period is presented, based on a deterministic SEIR model, via the technique of parameter perturbation which is standard in stochastic population modeling. The influence of the environmental noise as a standard Gaussian white noise on the epidemics' transmission is studied. Furthermore,the condition for the epidemics' persistence is obtained by formulating the corresponding function and using It 's formula. And the asymptotic behavior of the model near the endemic disease equilibrium is also studied. In this way, the decision support is provided in the application of this kind of stochastic SEIR model on the epidemics' prevention and control.

融合主题的社交媒体舆情SEIR传播模型构建与仿真研究

社交媒体改变了当下舆情事件应对治理方式,本文旨在融合主题因素探究在舆情传播过程中各因素对传播演化的影响,揭示社交媒体平台上舆情传播的特点,在舆情治理领域为相关职能部门提供参考借鉴。为抓住社交媒体舆情传播演化的关键,本文融合主题因素,基于传播演化SEIR模型,引入主题热度和用户兴趣度构建改进型SEIR模型并进行仿真研究,探查各个参数变量对传播的影响。仿真实验刻画了主题热度和用户兴趣度等因素对舆情传播的影响,体现了主题热度和用户兴趣度在监测、预警舆情传播演化中的重要价值,进而为政府和相关部门治理舆情提供参考和借鉴。

基于真实世界数据的修正SEIR模型应用于疫情防控研究

背景奥密克戎在世界各地广泛传播,深圳作为连接国内外交通的重要枢纽,自2022年2月以来持续受其影响,感染者数量迅速增加。目的构建修正的易感-暴露-感染-康复(SEIR)模型,为深圳市疫情防控工作提供具有应用价值的政策参考和建议,以缓解防控压力。方法在传统SEIR传染病动力学模型基础上,针对奥密克戎传播速度快、隐匿性高、人群普遍易感等流行病学特征,引入具有政策性特征的组别,即密接者、次密接者、入深隔离者和携带者组别,构建修正SEIR模型,拟合2022-02-18—28的深圳疫情数据确定修正模型的相关参数。结果该模型的预测数据与2022-03-01—04的实际数据基本一致,为预测疫情后续发展提供了可靠依据;进一步预测了2022-03-05—19的疫情发展趋势,从疫情防控的人工干预程度、介入时间及床位数、隔离房间数等医疗卫生资源需求等方面为深圳后续的疫情防控措施提供了指导。结论修正SEIR模型在疫情发展预测、防控措施制定和调整及医疗资源配置等方面具有重要实用价值。

基于改进SEIR模型的突发传染病传播规律研究

突发传染病严重威胁着人民的健康和社会的稳定,已经成为社会各界广泛关注的焦点问题,为了更好地预防和控制传染病的发生和传播。综合考虑突发传染病的传播特性,在传统SEIR模型的基础上,引入接触者类型,考虑接触者的延迟隔离时间,提出一种改进的SEIR模型(SCH_(c)EIUHRD模型)。结果验证了模型的有效性,明确了接触者延迟隔离对突发传染病传播的影响机理,表明接触者的及时隔离对疫情控制的必要性。研究能够为传染病防控措施的制定提供理论依据。

具有饱和恢复率的SEIR时滞模型的行波解

研究一类具有饱和恢复率的SEIR时滞模型的行波解.首先,考虑一类二维系统初值问题的适定性;然后,通过构造一对有界的向量值上、下解得到一个闭凸集;最后,利用Schauder不动点定理证明:当基本再生数R^(0)>1,波速c>c^(*)时模型存在非平凡行波解.

基于SEIR模型分析相关干预措施在新型冠状病毒肺炎疫情中的作用

目的:评价新型冠状病毒肺炎(COVID-19)疫情中相关干预措施的作用.方法:收集COVID-19疫情数据,分析COVID-19的传播特点,利用Python建立SEIR模型,加入潜伏期传染率、感染人群变化率等新参数,尝试分析防控手段的有效性.结果:给出多种相关干预措施下疫情发展趋势的预测.模型显示,基于严格限制出行的隔离措施能够减缓COVID-19发病曲线发展的趋势,使潜伏和感染人群的峰值降低45.71%和29.90%,潜伏人群数量较快到达群峰;模拟发现,本次疫情的拐点将出现在3月初,但对比COVID-19疫情实际数据,疫情拐点或将提前出现.结论:COVID-19疫情的防控措施是有效的,各地方政府应严格执行隔离制度,切断传播途径,巩固成果,全力遏制COVID-19传播.

基于SEIR的社交网络信息传播模型

结合传染病动力学的SEIR模型,本文建立了适用于社交网络的信息传播模型.该模型基于社交网络用户的行为特征,分析了社交网络的传播机理和网络参数对信息传播过程的影响,得出了动力学演化方程组,刻画了信息传播过程随时间的演化规律.仿真实验的结果表明,本文提出的模型能正确地反映社交网络上的信息传播的趋势,较为准确地描述信息传播过程.

SARS流行病的SEIR动力学模型及其参数辨识

建立了带有潜伏期及终身免疫的SARS流行病的SEIR动力学模型．根据部分国家或地区的SARS疫情数据,对模型中的参数进行了参数辨识,数值结果表明模型与实际疫情是比较吻合的,并得到了一个SARS流行病的带形控制区域．论证了系统的流不变性和弱不变性等数学性质．

基于SEIRS传染病模型的函数优化方法——SEIRS算法

为了解决复杂函数优化问题,采用SEIRS传染病模型提出了SEIRS算法.在该算法中,假设某个生态系统由若干人类个体组成,每个个体均由若干个特征来表征.该生态系统存在一种传染病在个体之间传染,该传染病攻击的是个体的部分特征.每个染病个体均经历易感、潜伏、发病和治愈等阶段,这些阶段的综合作用决定了个体的体质强弱;利用SEIRS传染病模型所描述的疾病传播机理构造出了相关算子,使个体之间能充分交换信息.结果表明:E-E,I-I和R-R算子能使体质强壮的个体向体质弱的个体传递强壮特征信息,使得后者能向好的方向发展;S-E,S-R,E-I(ω)和R-S(ω)算子能使处于不同状态的个体获得其他个体的平均特征信息,从而降低了该个体陷入局部最优解的概率;S-S算子能使个体的活跃度提高,从而扩大其搜索范围;E-R和I-R算子既具有S-S算子的特征又具有S-E,S-R,E-I(ω)和R-S(ω)算子的特征.体质强壮的个体能继续生长,而体质虚弱的个体则停止生长,从而确保本算法具有全局收敛性.测试结果表明:本算法具有搜索能力强的特点,对求解复杂函数优化问题具有很高的收敛速度.

基于SEIRS的建设工程质量风险传递模型及仿真研究

工程建设过程、从业人员专业水平和信息存在较为突出的碎片化现象,导致了工程质量风险的传递。为揭示平行承发包模式下建设工程质量风险的传递效应和放大效应,利用传染病模型中的动力学原理,综合考虑工程质量风险传递的影响因素,构建了SEIRS的建设工程质量风险传递模型。仿真研究结果表明,建设工程在质量风险传递过程中存在阈值;质量风险传递阈值h≥1时,在质量风险传递阻碍因素的作用下,风险逐步衰减直至消失;质量风险传递阈值h<1时,在质量风险传递促进因素的作用下,风险会稳定存在于建设工程中,甚至危害工程的整体质量。研究结论为政府实现由直接监管向间接监管、由事前监管向事中事后监管的转变等提供理论支撑。