您好,欢迎访问三七文档
当前位置:首页 > 办公文档 > 会议纪要 > 泰勒公式在常微分数值解法中的应用
江苏理工学院数理学院2016届毕业论文泰勒公式在常微分方程数值解法中的应用学院名称:数理学院专业:数学与应用数学班级:12数学S学号:12121131姓名:冯莹指导教师姓名:周群艳指导教师职称:副教授二〇一六年四月本科毕业论文江苏理工学院数理学院2016届毕业论文I泰勒公式在常微分方程数值解法中的应用摘要:泰勒公式是高等数学中的重要内容,对于泰勒公式及其应用的研究较多,而对于泰勒公式在常微分方程中的应用也有一部分,在深入到常微分方程数值解法的研究目前还不多。泰勒公式能构造一些数值计算方法,这些方法也能在实际生活中应用,凸显出了泰勒公式的基础性与重要性。本文主要从介绍泰勒公式的概念,欧拉(Ouler)公式和龙格-库塔(Runge-kutta)法用泰勒公式如何构造,局部截断误差精度以及在常微分方程数值解法的应用。关键词:泰勒公式;常微分方程;数值解法江苏理工学院数理学院2016届毕业论文IITheapplicationofTaylorformulainclassicalnumericalsolutionofordinarydifferentialequationAbstract:Taylor'sformulaisanimportantcontentinhighermathematics,lotsofresearchesonTaylorformulaanditsapplication,andtheTaylorformulainoftendifferentialequationintheapplicationarealsoapart.Indeepintooftenstudyofnumericalmethodsforsolutionofdifferentialequationsisnotmuch.Taylorformulacanbeconstructedsomenumericalmethods,thesemethodscanalsobeusedinreallife,highlightingthebasicandimportanceoftheTaylorformula.ThispapermainlyfromtheintroductionoftheconceptofTaylorformula,EulerOulerformulaandRungeKutta(RungeKutta)methodwiththeTaylorformulahowtoconstruct,thelocaltruncationerrorandtheapplicationofnumericalsolutionofdifferentialequationsoften.Keywords:Taylorformula;Ordinarydifferential;Numericalsolution江苏理工学院数理学院2016届毕业论文III目录摘要·····························································································ⅠAbstract························································································Ⅱ第1章引言···················································································1第2章泰勒公式简介····································································3§2.1公式的推导··········································································3§2.2公式的推广··········································································8第3章泰勒公式推导常微分方程数值解法····································11§3.1欧拉法···············································································11§3.1.1推导过程和几何意义······················································13§3.1.2截断误差和精度分析······················································13§3.2改进的欧拉法······································································11§3.2.1推导过程····································································16§3.2.2截断误差和精度分析······················································17§3.3龙格-库塔法········································································18第4章常微分方程数值解法在减肥模型中的应用·························26第5章总结··················································································36参考文献······················································································37致谢·····························································································38江苏理工学院数理学院2016届毕业论文第1页共41页第一章引言一阶常微分方程的数值解分为初值和边值问题,本论文讨论的是初值解问题,一般形式为0)(),(yayyxfdxdy(1.1)其中f为x,y的已知函数,0y为给定的初值。在以下的讨论中,我们做出以下假设:设函数),(yxf在区域:Xxx0,y内连续,并且关于y满足利普希茨(Lipschitz)条件,即存在常数L(Lipschitz常数),使得yyLyxfyxf),(),(由常微分方程理论,在以上假设下,初值问题(1.1)必定且唯一存在数值解)(xy。但实际上,求解仍有许多困难,到目前为止,我们只能对少数几个特殊类型的方程求其精确解,实际问题中遇到的许多形如(1.1)的初值问题,它们的解常常不能用初等函数表示,需要求其数值解。什么是数值解法?它是一种离散化方法。利用这种方法,可以在一系列离散点Nxxx,...,21上求出未知函数)(xy之值)(1xy,)(2xy...)(Ny的近似值Nxxx,...,21,自变量x的离散值Nxxx,...,21是事先取定的,称为节点,通常取成等距的,即hxx01,hxx212,...NhxxN0其中h)0(为步长,必要时可以改变t它的大小,而Nyyy,...,21通常称为初值。在数学历史上,泰勒公式起源于牛顿插值的有限差分法。1715年泰勒出版了《增量法及其逆》,在这本书中载有现在微积分教程中以他的名字命名的一元函数的幂级数展开公式,当时他是通过对格雷戈里—牛顿插值公式求极限而得到的。一百多年后,柯西对无穷级数的收敛性给出了一个严格的证明。1755年,欧拉把泰勒级数用于他的“微分学”时才认识到其价值,后来拉格朗日用带余项江苏理工学院数理学院2016届毕业论文第2页共41页的级数作为其函数理论的基础,从而进一步确认了泰勒级数的重要地位。泰勒也以函数的泰勒展开而闻名于后世。泰勒公式是高等数学中的重要内容,是微分概念和微分中值定理的推广和延伸。泰勒公式把对解析函数)(xf的近似与逼近提高到更精密的程度,是进行误差分析,收敛速度分析,确定复杂极限式的重要工具。要求熟练掌握常用函数的泰勒展式及其余项。对于泰勒公式及其应用的研究较多,而对于泰勒公式在常微分方程中的应用也有一部分,在深入到常微分方程数值解法的研究目前还不多。泰勒公式能构造一些数值计算方法,这些方法也能在实际生活中应用。本问从介绍泰勒公式的简介,泰勒公式推导常微分数值解法以及数值解法在模型中的应用凸显出了泰勒公式的基础性与重要性。江苏理工学院数理学院2016届毕业论文第3页共41页第二章泰勒公式简介§2.1公式的推导无论是近似计算还是理论分析,人们都希望用一些简单函数来近似表示复杂的函数,多项式就是一类简单函数,他只包含加和乘两种基本运算,最适用于计算机进行处理。天赐,人们常用多项式来近似表示函数,方法有很多。本文之考虑其中一种,即:给定函数)(xf,要找一个在指定点x附近)(xf与近似的多项式nnxaxaxaaxP...)(2210设函数)(xf在点0x处可导,于是按定义有:))(())(()()(000'0xxoxxxfxfxf,))(())(()()(000'0xxoxxxfxfxf.这表明在点0x附近,可用一次多项式表达)(xf,而误差是高于一阶的无穷小量。从几何角度看,这就是用曲线过点0x的切线来近似曲线。从微分学来讲,这种近似特点为在点0x处近似多项式的函数值和一阶导数值与被近似函数)(xf的相应值相等,但无论是理论证明还是实际计算中,在许多情况下使用这种逼近精度是不够的,所以采取以下办法。江苏理工学院数理学院2016届毕业论文第4页共41页(1)提高近似多项式的精度为了提高这种近似表示的精确度,我们希望用一个关于)(0xx的n次多项式nnnxxaxxaxxaaxP)(...)()()(0202010来近似表示)(xf,要求它与)(xf之差是比nxx)(0高阶的无穷小。为使求得的近似多项式与)(xf在数值与性质方面吻合的更好,我们进一步要求)(xPn在0x处的函数值及其直到n阶的导数值都与)(xf的相应值分别相等即要求)()(0)(0)(xfxPkkn),...2,1,0(nk按此要求,)()(00xfxpn;)()(0'0'xfxpn;............;)()(0)(0)(xfxpnnn令nnnxxaxxaxxaaxp)(...())(0202010)(,则10021')(...)2)(nnnxxnaxxaaxp(20032'')()1(...)(32)(nnnxxannxxaaxp...nnnanxp!)()()()(000xfxpan)()(0'0'1xfxpan)()(!20''0''2xfxpan...)()(!0)(0)(xfxpannnnn于是nnnxxxfnxxxfxxxfxfxp))((!1...))((!21))(()()(00)(200''00'0我们称)(xPn为)(xf在0x点的n阶的泰勒多项式即泰勒公式。江苏理工学院数理学院2016届毕业论文第5页共41页(2)误差估计命题2.1(带拉格朗日余项的泰勒公式)若函数f在点0x的某邻域)(0xo中1n阶可微,则对每个)(0xox,0xx,在x与0x之间存在,使)())((!1...))((!21))(()()(00)(200''00'0xRxxxfnxxx
本文标题:泰勒公式在常微分数值解法中的应用
链接地址:https://www.777doc.com/doc-7770862 .html