您好,欢迎访问三七文档
当前位置:首页 > 商业/管理/HR > 质量控制/管理 > SIR模型原理优缺点中英混排(暂定版)
SIR模型是传染病模型中最经典的模型,其中S表示易感者。模型中把传染病流行范围内的人群分成三类:S类,易感者(Susceptible),指未得病者,但缺乏免疫能力,与感病者接触后容易受到感染;I类,感病者(Infective),指染上传染病的人,它可以传播给S类成员;R类,移出者(Removal),指被隔离,或因病愈而具有免疫力的人。SIRmodelisthemostclassicmodelinepidemicmodels.Thismodelclassifypeopleasthreegroupsfollows:GroupS(Susceptible):thesehealthypeoplehavenoimmunity.Theyareeasilyinfectedwhencontactingwithinfectedpeople.GroupI(Infectedpeople):Theviruseshavealreadyinfectedthem.TheycanspreadvirustoGroupS.GroupR(Removal):peoplewhoarecuredanddied.假设总人数N不变,易感者、感病者、移出者三者的比例分别为s(t)、i(t)、r(t),并设病人的日接触率(每个病人每天有效接触的平均人数)为常数λ,日治愈率(每天被治愈的病人占总病人数的比例)为常数μ,则传染期接触数σ=λ/μ,则有Nowweassumethatthetotalnumberofpeople(N)isfixed,thustheproportionofeachgroupsares(t),i(t)andr(t).Everyinfectedpeoplecontactswithλpeopleeveryday,μpeoplearecured.Sos(t)+i(t)+r(t)=1不妨设初始时刻的易感染者,染病者,恢复者的比例分别为𝑠0、𝑖0、𝑟0,即Attheverybeginning,theproportionofeachgroupsare𝑠0、𝑖0、𝑟0,so,𝑠(0)=𝑠0(𝑠00)𝑖(0)=𝑖0(𝑖00)r(0)=𝑟0(𝑟00)SIR基础模型用微分方程组表示如下:Usingdifferentialequations,wedescribeBasalSIRmodelasfollows:didtdsdtdrdtsiisii通常情况下,r(0)=𝑟0都很小,可近似看作𝑟0≈0,𝑖0+𝑠0≈1,以上方程可化简为Ingeneral,aresmall,soitcanbeconsideredas𝑟0≈0,𝑖0+𝑠0≈1.Then,theequationscanbesimplifiedto{𝑑𝑖𝑑𝑡=𝜆𝑠𝑖−𝜇𝑖𝑑𝑠𝑑𝑡=−𝜆𝑠𝑖𝑠(0)=𝑠0𝑖(0)=𝑖0但s(t)、i(t)的求解十分困难,可利用相轨线分析讨论解i(t)、s(t)的性质,其中箭头表示了随着时间t的增加s(t)和i(t)的变化趋向However,s(t)andi(t)aredifficulttosolve.Wecanusetrajectorytoanalyzeandobtainthecharactersofi(t)、s(t).Thearrowsstandforthetendenciesofi(t)、s(t)withtimegoingby.分析图像可以得到以下结论:Analyzingthefigure,wecometotheconclusionsdownside.为保证传染病不蔓延,需要满足𝑠01/𝜎。为了达到这个目的,一方面,可以提高阈值1/𝜎,需降低𝜎,即减小日接触率𝜆,可通过提高卫生水平的方式;增大日治愈率μ,可以通过提高医疗水平的方式。另一方面,也可以通过群体免疫来提高𝑟0,从而降低𝑠0,使病情不蔓延。When𝑠01/𝜎,thecontagionwillnotspread.Toachievethisconditiontherearetwoways.Ononehand,byimprovinghygienelevels,wecanlower𝜆andlessenμ,namelyraisethethresholdvalue1/𝜎.Ontheotherhand,bypromotingherdimmunity,wecanimprove𝑟0,therebyreduce𝑠0.Inthesemeasures,thestateoftheillnesswillnotrise.模型优缺点:Advantageanddisadvantage:基于微分方程组求解的SIR模型可以根据已有数据比较准确地拟合曲线,并利用相轨线分析得出使传染病不蔓延的措施,理论依据充分。ThesolutionsofSIRmodelbasedondifferentialequationcanfittotherealisticcurveapproximately.Meanwhile,byanalyzingwithtrajectory,weconcludewaystocontroltheillnessfromspreading.Theresultsshowthatthetheoreticalbasisispracticable.但是应注意到,模型对人群的分类不够细致,没有明确考虑隔离的因素。而现实中对疑似病人的隔离是控制疫情传播的有效手段。Butweshouldrealizethatthismodelclassifiespeopleinaverysimplewayandconsidersnothingaboutisolation.However,inreality,isolationmakesagreatdifferenceincontrollingtheillness.模型没有引入反馈机制,在预测过程中,单纯依据已有数据预测未来较长一段时间的数据,必然会使准确度降低。尤其是题目中药物的介入和卫生条件的改善在过去的数据中是无法体现出来的,采用已有数据无法体现出这些因素对疫情控制的影响,这是模型致命的漏洞。为此必须引入反馈机制达到自我调整的功能。ThereisnofeedbackmechanisminSIRmodel.Inpredictions,forecastingalongrundataonlyusingdatawealreadyhavesurelywillletdowntheaccuracy.Butweshouldknowthatmanyfactorswillchangeinthefuture.Datacollectedbeforecannotreflectthechangesespeciallythosementionedinthequestionsuchasimprovementsofmedicineandhygienelevel.Thereisadrawbackhere.Tooptimizethismodelandmakeamoreaccurateprediction,weaddfeedbackmechanisminthat.此外,微分方程组求解较为困难,且对初值比较敏感,这对模型的稳健性是一个很大的影响。Lastbutnottheleast,differentialequationsaresensitivetoinitialvalue.Thesedisadvantageswillgravelyreducethestability基于以上考虑,我们引入了反馈机制。但是这对原有的连续模型提出了一个挑战,我们无法做到实时反馈,事实上,我们只需要将连续的时间划分为等距的时间段,然后按照时间段反馈,这和每日统计疫情数据比较相似。于是,连续模型就改为离散模型。Consideringalltheseweaddfeedbackmechanismtooptimizeourmodel.However,ourmodelcannotfeedbackinstantly.Infact,weneedtodividetimeintoequalperiods,thenfeedbackaccordingtothetimeperiod.Thiswayworkssimilartothefactthatcollectingandrenewingthedataeveryday.So,differentfromISRmodel,ourmodelisadiscretemodel.
本文标题:SIR模型原理优缺点中英混排(暂定版)
链接地址:https://www.777doc.com/doc-2858683 .html