泰勒公式外文翻译教程文件
高等数学3(6)泰勒公式课件
(k)( x0 )( x k!
x0 )k
o[( x
x0 )n ]
称为f ( x)按( x x0 )的幂展开的
带有皮亚诺型余项 的n阶泰勒公式.
(4) 展开式是唯一的
23
泰勒公式
f (x)
f ( x00 )
f ( x00 )( x x00 )
f
( x00 2!
)
(
x
x00
)2
f
(n) ( x00 n!
分析 即证 Rn ( x) f ( x) Pn ( x).
Rn( x)
f (n1) ( )
(n 1)!
(
x
x0
)n1
(在x0与x之间).
也即证
Rn ( x) ( x x0 )n1
f f( x(n)1) (Pn)( x) ( (xnx10)!)n1
f (n1) ( )
(n 1)!
16
泰勒公式
泰勒(Taylor)(英)1685-1731
第六节 泰勒(Taylor)公式
泰勒公式的建立 几个初等函数的麦克劳林公式 近似计算与误差估计 其它应用 小结 思考题 作业
第三章 微分中值定理与导数的应用
1
泰勒公式
一、泰勒公式的建立
多项式函数
简单的, 熟悉的函数来近似代替复杂函数.
用多项式近似表示函数—
则 21
泰勒公式
Rn( x)
f (n1)( ) (
(n 1)!
x
x0 )n1
M| (n 1)!
x
x0
|n1
从而
则
lim
x x0
Rn (
Rn (x)
幂级数中英文资料外文翻译文献
幂级数中英文资料外文翻译文献Power Series Expansion and Its ApplicationsIn the previous section, we discuss the convergence of power series, in its convergence region, the power series always converges to a function. For the simple power series, but also with itemized derivative, or quadrature methods, find this and function. This section will discuss another issue, for an arbitrary function ()x f , can be expanded in a power series, and launched into.Whether the power series ()x f as and function? The following discussion will address this issue. 1 Maclaurin (Maclaurin) formulaPolynomial power series can be seen as an extension of reality, so consider the function ()x f can expand into power series, you can from the function ()x f and polynomials start to solve this problem. To this end, to give here without proof the following formula.Taylor (Taylor) formula, if the function ()x f at 0x x = in a neighborhood that until the derivative of order 1+n , then in the neighborhood of the following formula :20000()()()()()()n n f x f x x x x x x x r x =+-+-++-+… (9-5-1)Among10()()n n r x x x +=-That ()n r x for the Lagrangian remainder. That (9-5-1)-type formula for the Taylor.If so 00x =, get2()(0)()n n f x f x x x r x=+++++…, (9-5-2) At this point,(1)(1)111()()()(1)!(1)!n n n n n f f x r x x x n n ξθ+++++==++ (01θ<<).That (9-5-2) type formula for the Maclaurin.Formula shows that any function ()f x as long as until the 1n +derivative, n can be equal to a polynomial and a remainder.We call the following power series()2(0)(0)()(0)(0)2!!n nf f f x f f x x x n '''=+++++…… (9-5-3) For the Maclaurin series.So, is it to ()f x for the Sum functions? If the order Maclaurin series (9-5-3) the first 1n + items and for 1()n S x +, which()21(0)(0)()(0)(0)2!!n nn f f S x f f x x x n +'''=++++…Then, the series (9-5-3) converges to the function ()f x the conditions1lim ()()n n s x f x +→∞=.Noting Maclaurin formula (9-5-2) and the Maclaurin series (9-5-3) the relationship between theknown1()()()n n f x S x r x +=+Thus, when()0n r x =There,1()()n f x S x +=Vice versa. That if1lim ()()n n s x f x +→∞=,Units must()0n r x =.This indicates that the Maclaurin series (9-5-3) to ()f x and function as the Maclaurin formula (9-5-2) of the remainder term ()0n r x → (when n →∞).In this way, we get a function ()f x the power series expansion:()()0(0)(0)()(0)(0)!!n n n nn f f f x x f f x x n n ∞='==++++∑……. (9-5-4) It is the function ()f x the power series expression, if, the function of the power series expansion is unique. In fact, assuming the function f (x ) can be expressed as power series20120()n n n n n f x a x a a x a x a x ∞===+++++∑……, (9-5-5)Well, according to the convergence of power series can be itemized within the nature of derivation,and then make 0x = (power series apparently converges in the 0x = point), it is easy to get()2012(0)(0)(0),(0),,,,,2!!n nn f f a f a f x a x a x n '''====…….Substituting them into (9-5-5) type, income and ()f x the Maclaurin expansion of (9-5-4) identical. In summary, if the function f (x ) contains zero in a range of arbitrary order derivative, and in thisrange of Maclaurin formula in the remainder to zero as the limit (when n → ∞,), then , the function f (x ) can start forming as (9-5-4) type of power series.Power Series()20000000()()()()()()()()1!2!!n n f x f x f x f x f x x x x x x x n '''=+-+-++-……,Known as the Taylor series.Second, primary function of power series expansionMaclaurin formula using the function ()f x expanded in power series method, called the direct expansion method.Example 1Test the function ()xf x e =expanded in power series of x . Solution because()()n x f x e =,(1,2,3,)n =…Therefore()(0)(0)(0)(0)1n f f f f '''====…,So we get the power series21112!!n x x x n +++++……, (9-5-6) Obviously, (9-5-6)type convergence interval (,)-∞+∞, As (9-5-6)whether type ()x f x e = is Sum function, that is, whether it converges to ()x f x e = , but also examine remainder ()n r x . Because1e ()(1)!xn n r x x n θ+=+ (01θ<<),且x x x θθ≤≤,Therefore11e e ()(1)!(1)!xx n n n r x x x n n θ++=<++,Noting the value of any set x ,xe is a fixed constant, while the series (9-5-6) is absolutely convergent, sothe general when the item when n →∞, 10(1)!n xn +→+ , so when n → ∞,there10(1)!n xx e n +→+, From thislim ()0n n r x →∞=This indicates that the series (9-5-6) does converge to ()xf x e =, therefore21112!!x n e x x x n =+++++…… (x -∞<<+∞). Such use of Maclaurin formula are expanded in power series method, although the procedure is clear,but operators are often too Cumbersome, so it is generally more convenient to use the following power series expansion method.Prior to this, we have been a functionx-11, xe and sin x power series expansion, the use of these known expansion by power series of operations, we can achieve many functions of power series expansion. This demand function of power series expansion method is called indirect expansion .Example 2Find the function ()cos f x x =,0x =,Department in the power series expansion. Solution because(sin )cos x x '=,And3521111sin (1)3!5!(21)!n n x x x x x n +=-+-+-++……,(x-∞<<+∞)Therefore, the power series can be itemized according to the rules of derivation can be342111cos 1(1)2!4!(2)!n nx x x x n =-+-+-+……,(x -∞<<+∞) Third, the function power series expansion of the application exampleThe application of power series expansion is extensive, for example, can use it to set some numerical or other approximate calculation of integral value.Example 3 Using the expansion to estimate arctan x the value of π.Solution because πarctan14= Because of357arctan 357x x x x x =-+-+…, (11x -≤≤),So there1114arctan14(1)357π==-+-+…Available right end of the first n items of the series and as an approximation of π. However, the convergence is very slow progression to get enough items to get more accurate estimates of πvalue.幂级数的展开及其应用在上一节中,我们讨论了幂级数的收敛性,在其收敛域内,幂级数总是收敛于一个和函数.对于一些简单的幂级数,还可以借助逐项求导或求积分的方法,求出这个和函数.本节将要讨论另外一个问题,对于任意一个函数()f x ,能否将其展开成一个幂级数,以及展开成的幂级数是否以()f x 为和函数?下面的讨论将解决这一问题.一、 马克劳林(Maclaurin)公式幂级数实际上可以视为多项式的延伸,因此在考虑函数()f x 能否展开成幂级数时,可以从函数()f x 与多项式的关系入手来解决这个问题.为此,这里不加证明地给出如下的公式.泰勒(Taylor)公式 如果函数()f x 在0x x =的某一邻域内,有直到1n +阶的导数,则在这个邻域内有如下公式:()20000000()()()()()()()()()2!!n n n f x f x f x f x f x x x x x x x r x n '''=+-+-++-+…,(9-5-1)其中(1)10()()()(1)!n n n f r x x x n ξ++=-+.称()n r x 为拉格朗日型余项.称(9-5-1)式为泰勒公式.如果令00x =,就得到2()(0)()n n f x f x x x r x =+++++…, (9-5-2)此时,(1)(1)111()()()(1)!(1)!n n n n n f f x r x x x n n ξθ+++++==++, (01θ<<).称(9-5-2)式为马克劳林公式.公式说明,任一函数()f x 只要有直到1n +阶导数,就可等于某个n 次多项式与一个余项的和. 我们称下列幂级数()2(0)(0)()(0)(0)2!!n nf f f x f f x x x n '''=+++++…… (9-5-3)为马克劳林级数.那么,它是否以()f x 为和函数呢?若令马克劳林级数(9-5-3)的前1n +项和为1()n S x +,即()21(0)(0)()(0)(0)2!!n nn f f S x f f x x x n +'''=++++…,那么,级数(9-5-3)收敛于函数()f x 的条件为1lim ()()n n s x f x +→∞=.注意到马克劳林公式(9-5-2)与马克劳林级数(9-5-3)的关系,可知1()()()n n f x S x r x +=+.于是,当()0n r x =时,有1()()n f x S x +=.反之亦然.即若1lim ()()n n s x f x +→∞=则必有()0n r x =.这表明,马克劳林级数(9-5-3)以()f x 为和函数⇔马克劳林公式(9-5-2)中的余项()0n r x → (当n →∞时).这样,我们就得到了函数()f x 的幂级数展开式:()()20(0)(0)(0)()(0)(0)!2!!n n n nn f f f f x x f f x x x n n ∞='''==+++++∑……(9-5-4) 它就是函数()f x 的幂级数表达式,也就是说,函数的幂级数展开式是唯一的.事实上,假设函数()f x 可以表示为幂级数20120()n n n n n f x a x a a x a x a x ∞===+++++∑……, (9-5-5)那么,根据幂级数在收敛域内可逐项求导的性质,再令0x =(幂级数显然在0x =点收敛),就容易得到()2012(0)(0)(0),(0),,,,,2!!n nn f f a f a f x a x a x n '''====…….将它们代入(9-5-5)式,所得与()f x 的马克劳林展开式(9-5-4)完全相同.综上所述,如果函数()f x 在包含零的某区间内有任意阶导数,且在此区间内的马克劳林公式中的余项以零为极限(当n →∞时),那么,函数()f x 就可展开成形如(9-5-4)式的幂级数.幂级数()00000()()()()()()1!!n n f x f x f x f x x x x x n '=+-++-……,称为泰勒级数.二、 初等函数的幂级数展开式利用马克劳林公式将函数()f x 展开成幂级数的方法,称为直接展开法. 例1 试将函数()xf x e =展开成x 的幂级数. 解 因为()()n x f x e =, (1,2,3,)n =…所以()(0)(0)(0)(0)1n f f f f '''====…,于是我们得到幂级数21112!!n x x x n +++++……, (9-5-6) 显然,(9-5-6)式的收敛区间为(,)-∞+∞,至于(9-5-6)式是否以()xf x e =为和函数,即它是否收敛于()x f x e =,还要考察余项()n r x .因为1e ()(1)!xn n r x x n θ+=+ (01θ<<), 且x x x θθ≤≤,所以11e e ()(1)!(1)!xx n n n r x x x n n θ++=<++.注意到对任一确定的x 值,xe 是一个确定的常数,而级数(9-5-6)是绝对收敛的,因此其一般项当n →∞时,10(1)!n xn +→+,所以当n →∞时,有10(1)!n xx e n +→+, 由此可知lim ()0n n r x →∞=.这表明级数(9-5-6)确实收敛于()x f x e =,因此有21112!!x n e x x x n =+++++…… (x -∞<<+∞). 这种运用马克劳林公式将函数展开成幂级数的方法,虽然程序明确,但是运算往往过于繁琐,因此人们普遍采用下面的比较简便的幂级数展开法.在此之前,我们已经得到了函数x-11,xe 及sin x 的幂级数展开式,运用这几个已知的展开式,通过幂级数的运算,可以求得许多函数的幂级数展开式.这种求函数的幂级数展开式的方法称为间接展开法.例2 试求函数()cos f x x =在0x =处的幂级数展开式. 解 因为(sin )cos x x '=,而3521111sin (1)3!5!(21)!n n x x x x x n +=-+-+-++……,(x -∞<<+∞), 所以根据幂级数可逐项求导的法则,可得342111cos 1(1)2!4!(2)!n nx x x x n =-+-+-+……,(x -∞<<+∞). 三、 函数幂级数展开的应用举例幂级数展开式的应用很广泛,例如可利用它来对某些数值或定积分值等进行近似计算. 例3 利用arctan x 的展开式估计π的值. 解 由于πarctan14=, 又因357arctan 357x x x x x =-+-+…, (11x -≤≤),所以有1114arctan14(1)357π==-+-+….可用右端级数的前n 项之和作为π的近似值.但由于级数收敛的速度非常慢,要取足够多的项才能得到π的较精确的估计值.。
3,3泰勒公式-52页PPT精品文档
f ( x0) +
f ( x0 )( x - x0 ) +
f
( x0 2!
)
(x
-
x0 )2
+
+
f
(n) ( x0 n!
)
(x
-
x0 )n
+
Rn
(
x)
其中 Rn( x) =
f (n+1)( ) ( x
(n + 1)!
-
x0 )n+1(
在
x0与
x之间).
上页 下页
泰勒公式
证明: 由 假 设 , R n ( x )在 ( a ,b )内 具 有 直 到 ( n + 1 ) 阶
L+f(nn )(!x0)(x-x0)n=kn=0 f(kk)(!x0)(x-x0)k
称f为 (x)按 (x-x0)的幂展 n次近开 似多的 项式 .
f(x)=kn =0f(kk )(!x0)(x-x0)k+R n(x)
称f为 (x)按 (x-x0)的幂展 n阶泰 开 勒公的 式.
R n(x)=fn (n + + 1)1 (!)(x-x 0)n + 1(在 x 0 与 x 之 )间
x
以直代曲
f ( x ) = f ( x 0 ) + f ( x 0 ) x - ( x 0 ) + o ( x - x 0 )
例 如 ,当 x很 小 时 ,ex1+x,ln1+ (x)x
(如下图)
上页 下页
泰勒公式
以直代曲
y = ex
y = ex
y=x
y=1+x
泰勒公式及应用翻译(原文)
On Taylor’s formula for the resolvent of a complex matrixMatthew X. Hea, Paolo E. Ricci b,_Article history:Received 25 June 2007Received in revised form 14 March 2008Accepted 25 March 2008Keywords:Powers of a matrixMatrix invariantsResolvent1. IntroductionAs a consequence of the Hilbert identity in [1], the resolvent )(A R λ= 1)(--E A λof a nonsingular square matrix A (E denoting the identity matrix) is shown to be an analytic function of the parameter λ in any domain D with empty intersection with the spectrum ∑A of A .Therefore, by using Taylor expansion in a neighborhood of any fixed D ∈0λ, we can find in [1] a representation formula for )(A R λ using all powers of )(0A R λ.In this article, by using some preceding results recalled, e.g., in [2], we write down a representation formula using only afinite number of powers of )(0A R λ. This seems to be natural since only the first powers of )(0A R λ are linearly independent.The main tool in this framework is given by the multivariable polynomials ),...,,(21,r n k v v v F (,...1,0,1-=n ;r m k ≤=,...,2,1) (see [2–6]), depending on the invariants ),...,,(21r v v v of )(A R λ); heremdenotes the degree of theminimalpolynomial.2. Powers of matrices a nd n k F ,functionsWerecall in this section some results on representation formulas for powers of matrices (see e.g. [2–6] and the referencestherein). For simplicity we refer to the case when the matrix is nonderogatory so that r m =.Proposition 2.1. Let A be an )2(≥⨯r r r complex matrix, and denote by r u u u ,...,,21 the invariants of A , and by∑=--=-E =rj j r j j u A P 0)1()det()(λλλ.its characteristic polynomial (by convention 10-=u ); then for the powers of A with nonnegative integral exponents the following representation formula holds true:E +++=-----),,(),...,,(),...,(211,2211,2171,1r n r r r n r i n n u u uF A u u u F A u u F A . (2.1) The functions ),,(1,r n k u u F that appear as coefficients in (2.1) are defined by the recurrence relation),()1(),(),,(),,(,1,1,12,211,11,r r n k r r r n k r n k r n k u u F u u u F u u u F u u u F -----++-=,)1;,,1(-≥=n r k (2.2)and initial conditions:,),,(,12,17h k h k r u u F σ=-+- ),,1,(r h k =. (2.3)Furthermore, if A is nonsingular )0(≠r u , then formula (2.1) still holds for negative values of n, provided that we define the n k F , function fornegative values of n as follows:)1,,,(),,(7112,171,u u u u u F u u F r r r r n k r n k --+-+-=,)1;,,1(-= n r k . 3. Taylor expansion of the resolventWe consider the resolvent matrix )(A R λ defined as follows:1)()(--E =≡A A R R λλλ. (3.1)Note that sometimes there is a change of sign in Eq. (3.1), but this of course is not essential.It is well known that the resolvent is an analytic (rational) function of λ in every domain D of the complex plane excludingthe spectrum of A , and furthermore it is vanishing at infinity so the only singular points (poles) of )(A R λare the eigenvaluesof A .In [6] it is proved that the invariants r v v v ,,,21 of )(A R λ are linked with those of A by the equations∑=-⎪⎪⎭⎫ ⎝⎛---=lj j l j j l u j l j r v 0()1()(λλ,),,2,1(r l =. (3.2) As a consequence of Proposition 2.1, and Eq. (3.2), the integral powers of )(A R λ can be represented as follows.For every ∑∉A λand N n ∈,∑-=--=10211,)())(,),(),(()(r k k r n k r nA R v v v F A R λλλλλ , (3.3) where the )(λl v ),,2,1(r l =are given by Eq.(3.2).Denoting by )(A ρthe spectral radius of A , for every λ,μ such that ),,min()(μλρ A the Hilbert identity holds true(see [1]):)()()()()(A R A R A R A R μλμλλμ-=-. (3.4)Therefore for every ∑∉A λ, we have)()(2A R d A dR λλλ-=, (3.5) and in general)()1()(1A kR d A R d k k k k +-=λλλ,);,2,1(∑∉=A k λ (3.6) so, for every )(,0A R D λλ∈can be expanded in the Taylor seriesk k k k A kR A R ))(()1()(0010λλλλ--=∑∞=+, (3.7) which is absolutely and uniformly convergent in D.Defining)(,),(000110λλr r v v v v == , (3.8) ),,(010,,0r n k n k v v F F =, (3.9)where the )(λl v are defined by Eq. (3.2), we can prove the following theorem.The Taylor expansion (3.7) of the resolvent )(A R λ in a neighborhood ofany regular point 0λ can be written in the form)()()1()(01000,0A R F A R n r h k k k n r k λλλλ∑∑-=∞=-⎪⎭⎫ ⎝⎛--=. (3.10) Therefore we can derive as a consequence:For every ∑∉A 0λand r L ,2,1=the series expansions∑∞=--00,0)()1(k k k l k F λλ(3.11) are convergent.Proof. Recalling (3.3), we can writeE +++=--+k r r k r k kF R F R F R ,02,201,101000 λλλ,)N k ∈, )()(01,021,2011,10000λλλλλλ-⎥⎦⎤⎢⎣⎡E +++-=--r R R F R F R F R A R --⎥⎦⎤⎢⎣⎡E ++++--202,022,2012,10)(00λλλλr r r F R F R F +-⎥⎦⎤⎢⎣⎡E +++-+--k k r r k r k k F R F R F )()1(0,02,201,1000λλλλ Therefore, taking into account the initial conditions (2.3) we can write+⎥⎦⎤⎢⎣⎡--+⎥⎦⎤⎢⎣⎡--=∑∑∞=-∞=000,1000,0)()1()()1()(λλλλσλλR F F A R k k k r k k k k r k 100,100)()1(-∞=⎥⎦⎤⎢⎣⎡--+∑R k k k k R F λλλ, so (3.10) holds true. The convergence of series expansions (3.11) is a trivial consequence of the convergence of the initial expansion (3.7).4. Concluding remarksIt is worth noting that the resolvent )(A R λis a keynote element forrepresenting analytic functions of a matrix A . In fact,denoting by )(z f a function of the complex variable z , analytic in a domain containing the spectrum of A , and denoting by ),2,1(s k k =λthe distinct eigenvaluesof A with multiplicities k μ, the Lagrange –Sylvester formula (see [4]) is given by∑∑=-==s k j j k k j k jf A f 110)()()(μελ, where ),,2,1()0(s k k k ==εεis the projector associated with the eigenvalue k λ, and)1,,1,0;,,2,1(,)()0()(-==-=k k j k j k j s k A l μελε .Denoting by k γ a Jordan curve, the boundary of the domain k D , separating a fixed k λ from all other eigenvalues, recalling theRiesz formula, it follows that⎰=k d A R ik γλλπε)(21. When k λis only known approximately, this projector cannot be derived by using the residue theorem.In this case it is necessary to integrate )(A R λ along k γ (being possibly a Gershgorin circle), by using the knownrepresentation of the resolvent (see [3])k r k k r j j k r j j A P A R ∑∑-=--=---⎥⎦⎤⎢⎣⎡-=10101)1()(1)(λμλλ, (4.1) or by substituting )(A R λ with its Taylor expansion, and assuming asinitial point any K λλ≠0 inside k D .Which is the best formula depends on the relevant stability and computational cost. From the theoretical point of view,formulas (3.7), (3.10) and (4.1) seem to beequivalent from the stability point of view, since all require knowledge ofinvariants of the given matrix A . However, in our opinion, in the situation considered, Eq. (3.10) seems to be less expensivewith respect to (3.7), since it requires one to approximate r series of elementary functions instead of an infinite series ofmatrices. AcknowledgementsWe are grateful to the anonymous referees for comments that led usto improve this paper.References[1] I. Glazman, Y. Liubitch, Analyse linéaire dans les espaces de dimensionfinies: Manuel et problèmes, in: H. Damadian (Ed.), Traduit du russe par, Mir,Moscow, 1972.[2] M. Bruschi, P.E. Ricci, Sulle potenze di una matrice quadrata dellaquale sia noto il polinomio minimo, Pubbl. Ist. Mat. Appl. Fac. Ing.Univ. Stud. Roma,Quad. 13 (1979) 9–18.[3] V.N. Faddeeva, Computational Methods of Linear Algebra, Dover Pub.Inc., New York, 1959.[4] F.R. Gantmacher, The Theory of Matrices, Vols. 1, 2 (K.A. Hirsch,Trans.), Chelsea Publishing Co., New York, 1959.[5] M. Bruschi, P.E. Ricci, Sulle funzioni Fk,n e i polinomi di Lucas diseconda specie generalizzati, Pubbl. Ist. Mat. Appl. Fac. Ing. Univ. Stud.Roma, Quad. 14(1979) 49–58.[6] M. Bruschi, P.E. Ricci, An explicit formula for f (A) and the generatingfunction of the generalized Lucas polynomials, SIAM J. Math. Anal. 13 (1982)162–165.。
泰勒公式外文翻译教学内容
泰勒公式外文翻译Taylor's Formula and the Study of Extrema1. Taylor's Formula for MappingsTheorem 1. If a mapping Y U f →: from a neighborhood ()x U U = of a point x in a normed space X into a normed space Y has derivatives up to order n -1 inclusive in U and has an n-th order derivative ()()x f n at the point x, then()()()()()⎪⎭⎫ ⎝⎛++++=+n n n h o h x f n h x f x f h x f !1,Λ (1) as 0→h .Equality (1) is one of the varieties of Taylor's formula, written here for rather generalclasses of mappings.Proof. We prove Taylor's formula by induction.For 1=n it is true by definition of ()x f ,.Assume formula (1) is true for some N n ∈-1.Then by the mean-value theorem, formula (12) of Sect. 10.5, and the induction hypothesis, we obtain.()()()()()()()()()()()()()⎪⎭⎫ ⎝⎛=⎪⎭⎫ ⎝⎛=⎪⎪⎭⎫ ⎝⎛-+++-+≤⎪⎭⎫ ⎝⎛+++-+--<<n n n n n n h o h h o h h x f n h x f x f h x f h x f n h f x f h x f 11,,,,10!11sup !1x θθθθθΛΛ,as 0→h .We shall not take the time here to discuss other versions of Taylor's formula, which are sometimes quite useful. They were discussed earlier in detail for numerical functions. At this point we leave it to the reader to derive them (see, for example, Problem 1 below).2. Methods of Studying Interior ExtremaUsing Taylor's formula, we shall exhibit necessary conditions and also sufficient conditions for an interior local extremum of real-valued functions defined on an open subset of a normed space. As we shall see, these conditions are analogous to the differential conditions already known to us for an extremum of a real-valued function of a real variable.Theorem 2. Let R U f →: be a real-valued function defined on an open set U in a normed space X and having continuous derivatives up to order 11≥-k inclusive in a neighborhood of a point U x ∈ and a derivative ()()x f k of order k at the point x itself.If ()()()0,,01,==-x f x f k Λ and ()()0≠x f k , then for x to be an extremum of the function f it is: necessary that k be even and that the form()()k k h x f be semidefinite,andsufficient that the values of the form ()()k k h x f on the unit sphere 1=h be bounded away from zero; moreover, x is a local minimum if the inequalities()()0>≥δk k h x f ,hold on that sphere, and a local maximum if()()0<≤δk k h x f ,Proof. For the proof we consider the Taylor expansion (1) of f in a neighborhood of x. The assumptions enable us to write()()()()()k k k h h h x f k x f h x f α+=-+!1where ()h α is a real-valued function, and ()0→h α as 0→h .We first prove the necessary conditions.Since ()()0≠x f k , there exists a vector 00≠h on which()()00≠k k h x f . Then for values of the real parameter t sufficiently close to zero,()()()()()()k k k th th th x f k x f th x f 0000!1α+=-+()()()kk k k t h th h x f k ⎪⎭⎫ ⎝⎛+=000!1αand the expression in the outer parentheses has the same sign as()()k k h x f 0. For x to be an extremum it is necessary for the left-hand side (and hence also the right-hand side) of this last equality to be of constant sign when t changes sign. But this is possible only if k is even.This reasoning shows that if x is an extremum, then the sign of the difference ()()x f th x f -+0 is the same as that of ()()k k h x f 0 for sufficiently small t; hence in that case there cannot be two vectors 0h , 1h at which the form ()()x f k assumes values with opposite signs.We now turn to the proof of the sufficiency conditions. For definiteness we consider the case when ()()0>≥δk k h x f for 1=h . Then()()()()()k k k h h h x f k x f h x f α+=-+!1()()()k k k h h h h x f k ⎪⎪⎪⎭⎫ ⎝⎛+⎪⎪⎭⎫ ⎝⎛=α!1 ()k h h k ⎪⎭⎫ ⎝⎛+≥αδ!1and, since ()0→h α as 0→h , the last term in this inequality is positive for all vectors0≠h sufficiently close to zero. Thus, for all such vectors h,()()0>-+x f h x f ,that is, x is a strict local minimum.The sufficient condition for a strict local maximum is verified similiarly.Remark 1. If the space X is finite-dimensional, the unit sphere ()1;x S with center at X x ∈, being a closed bounded subset of X, is compact. Then the continuous function()()()()k k i i i i k k h h x f h x f ⋅⋅∂=ΛΛ11 (a k-form) has both a maximal and a minimal value on ()1;x S . Ifthese values are of opposite sign, then f does not have an extremum at x. If they are both of the same sign, then, as was shown in Theorem 2, there is an extremum. In the latter case, a sufficient condition for an extremum can obviously be stated as the equivalent requirement that the form ()()k k h x f be either positive- or negative-definite.It was this form of the condition that we encountered in studying realvalued functions on n R . Remark 2. As we have seen in the example of functions R R fn →:, the semi-definiteness of the form ()()k k h x f exhibited in the necessary conditions for an extremum is not a sufficient criterion for an extremum.Remark 3. In practice, when studying extrema of differentiable functions one normally uses only the first or second differentials. If the uniqueness and type of extremum are obvious from the meaning of the problem being studied, one can restrict attention to the firstdifferential when seeking an extremum, simply finding the point x where ()0,=x f3. Some ExamplesExample 1. Let ()()R R C L ;31∈ and ()[]()R b a C f ;,1∈. In other words, ()()321321,,,,u u u L u u u α is a continuously differentiable real-valued function defined in 3R and ()x f x α a smooth real-valued function defined on the closed interval []R b a ⊂,.Consider the function()[]()R R b a C F →;,:1 (2)defined by the relation()[]()()f F R b a C f α;,1∈()()()R dx x f x f x L ba ∈=⎰,,, (3) Thus, (2) is a real-valued functional defined on the set of functions ()[]()Rb a C ;,1.The basic variational principles connected with motion are known in physics andmechanics. According to these principles, the actual motions are distinguished among all the conceivable motions in that they proceed along trajectories along which certain functionals have an extremum. Questions connected with the extrema of functionals are central in optimal control theory. Thus, finding and studying the extrema of functionals is a problemof intrinsic importance, and the theory associated with it is the subject of a large area ofanalysis - the calculus of variations. We have already done a few things to make the transition from the analysis of the extrema of numerical functions to the problem of finding andstudying extrema of functionals seem natural to the reader. However, we shall not go deeply into the special problems of variational calculus, but rather use the example of the functional(3) to illustrate only the general ideas of differentiation and study of local extrema considered above.We shall show that the functional (3) is a differentiate mapping and find its differential. We remark that the function (3) can be regarded as the composition of the mappings()()()()()x f x f x L x f F ,1,,= (4)defined by the formula()[]()[]()R b a C R b a C F ;,;,:11→ (5)followed by the mapping[]()()()R dx x g g F R b a C g ba ∈=∈⎰2;,α (6) By properties of the integral, the mapping 2F is obviously linear and continuous, so that its differentiability is clear.We shall show that the mapping 1F is also differentiable, and that()()()()()()()()()()x h x f x f x L x h x f x f x L x h f F ,,3,2,1.,,,∂+∂= (7) for ()[]()R b a C h ;,1∈.Indeed, by the corollary to the mean-value theorem, we can write in the present case()()()i i i u u u L u u u L u u u L ∆∂--∆+∆+∆+∑=32131321332211,,,,,,()()()()()()∆⋅∂-∆+∂∂-∆+∂∂-∆+∂≤<<u L u L u L u L u L u L 3312211110sup θθθθ ()()ii i i u L u u L i ∆⋅∂-+∂≤=≤≤=3,2,110max max 33,2,1θθ (8)where ()321,,u u u u = and ()321,,∆∆∆=∆. If we now recall that the norm()1c f of the function f in ()[]()R b a C ;,1 is ⎭⎬⎫⎩⎨⎧c c f f ,,max (where c f is the maximum absolute value of the function on the closed interval []b a ,), then,setting x u =1, ()x f u =2, ()x f u ,3=, 01=∆, ()x h =∆2, and ()x h ,3=∆, we obtain from inequality (8),taking account of the uniform continuity of the functions ()3,2,1,,,321=∂i u u u L i , on bounded subsets of 3R , that()()()()()()()()()()()()()()()()x h x f x f x L x h x f x f x L x f x f x L x h x f x h x f x L b x ,,3,2,,,0,,,,,,,,max ∂-∂--++≤≤ ()()1c h o = as ()01→c h But this means that Eq. (7) holds.By the chain rule for differentiating a composite function, we now conclude that thefunctional (3) is indeed differentiable, and()()()()()()()()()()⎰∂+∂=badx x h x f x f x L x h x f x f x L h f F ,,3,2,,,,, (9) We often consider the restriction of the functional (3) to the affine space consisting of the functions ()[]()R b a C f ;,1∈ that assume fixed values ()A a f =, ()B b f = at the endpoints of the closed interval []b a ,. In this case, the functions h in the tangent space ()1f TC , must have the value zero at the endpoints of the closed interval []b a ,. Taking this fact into account, we mayintegrate by parts in (9) and bring it into the form()()()()()()()()⎰⎪⎭⎫ ⎝⎛∂-∂=b a dx x h x f x f x L dx d x f x f x L h f F ,3,2,,,,, (10) of course under the assumption that L and f belong to the corresponding class ()2C .In particular, if f is an extremum (extremal) of such a functional, then by Theorem 2 we have ()0,=h f F for every function ()[]()R b a C h ;,1∈ such that ()()0==b h a h . From this and relation (10) one can easily conclude (see Problem 3 below) that the function f must satisfy the equation()()()()()()0,,,,,3,2=∂-∂x f x f x L dx d x f x f x L (11)This is a frequently-encountered form of the equation known in the calculus of variations as the Euler-Lagrange equation.Let us now consider some specific examples.Example 2. The shortest-path problemAmong all the curves in a plane joining two fixed points, find the curve that has minimal length.The answer in this case is obvious, and it rather serves as a check on the formalcomputations we will be doing later.We shall assume that a fixed Cartesian coordinate system has been chosen in the plane, in which the two points are, for example, ()0,0 and ()0,1 . We confine ourselves to just the curves that are the graphs of functions ()[]()R C f ;1,01∈ assuming the value zero at both ends of the closed interval []1,0 . The length of such a curve()()()⎰+=102,1dx x f f F (12)depends on the function f and is a functional of the type considered in Example 1. In this case the function L has the form()()233211,,u u u u L +=and therefore the necessary condition (11) for an extremal here reduces to the equation ()()()012,,=⎪⎪⎪⎭⎫ ⎝⎛+x f x f dx d from which it follows that ()()()常数≡+x f x f 2,,1 (13)on the closed interval []1,0Since the function 21u u+ is not constant on any interval, Eq. (13) is possible only if()≡x f ,const on []b a ,. Thus a smooth extremal of this problem must be a linear function whose graph passes through the points ()0,0 and ()0,1. It follows that ()0≡x f , and we arrive at the closed interval of the line joining the two given points.Example 3. The brachistochrone problemThe classical brachistochrone problem, posed by Johann Bernoulli I in 1696, was to find the shape of a track along which a point mass would pass from a prescribed point 0P to another fixed point 1P at a lower level under the action of gravity in the shortest time.We neglect friction, of course. In addition, we shall assume that the trivial case in which both points lie on the same vertical line is excluded.In the vertical plane passing through the points 0P and 1P we introduce a rectangularcoordinate system such that 0P is at the origin, the x-axis is directed vertically downward, and the point 1P has positive coordinates ()11,y x .We shall find the shape of the track among the graphs of smooth functions defined on the closed interval []1,0x and satisfying the condition ()00=f ,()11y x f =. At the moment we shall not take time to discuss this by no meansuncontroversial assumption (see Problem 4 below).If the particle began its descent from the point 0P with zero velocity, the law of variation of its velocity in these coordinates can be written asgx v 2= (14)Recalling that the differential of the arc length is computed by the formula()()()()dx x f dy dx ds 2,221+=+= (15) we find the time of descent()()()⎰+=102,121x dx x x f g f F (16) along the trajectory defined by the graph of the function ()x f y = on the closed interval []1,0x . For the functional (16) ()()1233211,,u u u u u L +=,and therefore the condition (11) for an extremum reduces in this case to the equation ()()()012,,=⎪⎪⎪⎪⎪⎭⎫⎝⎛⎪⎭⎫ ⎝⎛+x f x x f dx d ,from which it follows that ()()()x c x f x f =+2,,1 (17)where c is a nonzero constant, since the points are not both on the same vertical line. Taking account of (15), we can rewrite (17) in the formx c ds dy = (18)However, from the geometric point of viewϕcos =ds dx,ϕsin =ds dy(19)where ϕ is the angle between the tangent to the trajectory and the positive x-axis.By comparing Eq. (18) with the second equation in (19), we find ϕ22sin 1c x = (20) But it follows from (19) and (20) that dx dy d dy =ϕ,2222sin 2sin c c d d tg d dx tg d dx ϕϕϕϕϕϕϕ=⎪⎪⎭⎫⎝⎛==,from which we find()b c y +-=ϕϕ2sin 2212 (21)Setting a c =221and t =ϕ2, we write relations (20) and (21) as()()b t t a y t a x +-=-=sin cos 1 (22)Since 0≠a , it follows that 0=x only for πk t 2=,Z k ∈. It follows from the form of the function(22) that we may assume without loss of generality that the parameter value 0=t corresponds to the point ()0,00=P . In this case Eq. (21) implies 0=b , and we arrive at the simpler form()()t t a y t a x sin cos 1-=-= (23)for the parametric definition of this curve.Thus the brachistochrone is a cycloid having a cusp at the initial point 0P where the tangent is vertical. The constant a, which is a scaling coefficient, must be chosen so that the curve (23) also passes through the point 1P . Such a choice, as one can see by sketching the curve (23), is by no means always unique, and this shows that the necessary condition (11) for an extremum is in general not sufficient. However, from physical considerations it is clear which of thepossible values of the parameter a should be preferred (and this, of course, can be confirmed by direct computation).泰勒公式和极值的研究1.映射的泰勒公式定理1 如果从赋范空间X 的点x 的邻域()x U U =到赋范空间Y 的映射Y U f →:在U 中有直到n-1阶(包括n-1在内)的导数,而在点x 处有n 阶导数。
教程(2)参考译文
第一单元参考译文—英语教程(2)一位诺贝尔奖得主虚度的岁月希尔维亚·纳萨尔1994年诺贝尔经济学奖于10月11日颁布前的几个星期,两位数学家——哈罗德·W·库恩和小约翰·福布斯·纳什——来到离此地不远的草甸湖一家疗养院拜访了他们年迈的老师,年近九旬、卧床不起的艾伯特·W·塔克。
纳什先生已经很多年没跟自己的导师交谈过。
两人长达一小时的谈话,谈的是数论。
库恩先生没有参加。
等纳什先生走出房问后,库恩先生回进去告诉塔克先生一个惊人的秘密:纳什先生本人尚不知晓,瑞典皇家科学院打算为纳什先生在1949年师从老人时取得的研究成果而授予他诺贝尔奖,该成果后来对经济学含有革命性的意义。
这次授奖堪称奇迹。
这不仅仅因为作为战后的一位数学天才,纳什先生终于获得了应有的赞誉。
也不是因为他是由于几乎半个世纪前才21岁时写的薄薄一篇27页的博士论文而获此殊荣。
真正的奇迹是,66岁的纳什先生——高个子,鬓发灰白,眼神忧伤,有着说话不多的人那种轻轻的粗嗓音——竟活着健康地接受这一奖项。
因为约翰·纳什早在30多年前就身忠类偏狂型精神分裂症。
纳什先生身患顽症在数学家和经济学家之间是个公开的秘密。
1958年7月《财富》杂志刚刚评选纳什先生为“新数学”美国杰出青年新秀,这一疾病就摧毁了他的个人生活和学术生涯。
1958年迄今他没有发表过一篇学术论文。
1959年迄今他没有担任过任何学术职务。
不少人听到讹传,说他接受了脑白质切除术。
其他人,主要是那些不是普林斯顿的,干脆以为他已经死了。
他没有死,但他的生活,一度曾如此光辉灿烂,前程似锦,却变得地狱般可怕。
一次次被送进精神病院。
并不见效的治疗。
可怕的妄想症。
一度漫游欧洲。
在母亲与妹妹居住的弗吉尼亚州洛诺克散步。
最后又回到自己曾是那里冉冉升起之明星的普林斯顿。
在普林斯顿,他成了美楼幽灵,一个缄默无言的身影,在数学楼里的黑板上涂写奇奇怪怪的公式,急切地探寻着数字中的隐秘信息。
Toría del Conocimiento
TEORIA DEL CONOCIMIENTOCapítulo 11.1 INTRODUCCIÓN1-1.1 CUESTIÓN DE NOMBRESAl tipo de investigaciones que vamos a realizar se le han dado varios nombres:gnoseología , epistemología , crítica , lógica maior y teoria del conocimiento.La expresión "teoria del conocimiento" la introdujo por primera vez Ernest Reinhold en 1832 , y fue recogida por Edward Zeller en 1862 , utilizándose desde entonces de forma corriente. En Inglaterra , Alemania , Francia e Italia se utiliza "epistemología" ; en España , por predominio de la escolástica , se utilizaba "crítica" y "gnoseología".El término "epistemología" alude más bien a una teoría de la ciencia , y el de "crítica" a una dimensión fundamental de la teoría del conocimiento , pero no exclusiva ; "gnoseología" sería lo mismo que "teoría del conocimiento" en griego. Nosotros utilizaremos la expresión "teoría del conocimiento" por considerar que indica mejor nuestro estudio ; vamos a explicar esta expresión.1-1.2 TEORIA DEL CONOCIMIENTOTeoría , en griego , significa "contemplación". A lo largo de la historia se ha entendido este términode dos formas distintas:a) En Grecia (Aristóteles) y el medievo como "contemplación" en oposición y complementariedad con "praxis" y "poiesis". Esta concepción presupone:- que la actitud teorética es pasiva , y la teoria es un reflejo fiel de lo teorizado.- que lo teorizado se presenta en forma de organización acabada (cosmos).- que esta organización acabada es acorde con la capacidad teorizante.b) En el mundo moderno y sobre todo en el pensamiento actual se entiende por teoría una construcción intelectual que resulta del trabajo filosófico y/o científico. De ahí que aquellos presupuestos cambien de signo:- la actitud teorética es activa y la teoría es el resultado de una actividad compleja del hombre.- lo teoretizado , el mundo , se nos presenta en un cierto desorden , problemático , y hayque "ponerlo en orden".- este orden alcanzado , la teoría , puede no tener un carácter terminado , cumplido , puede ser siempre sólo una "conjetura".N.B. teorias , sistemas y modelosNi la reflexión filosófica ni científica se suelen interesar por los individuos aislados , sino por los sistemas. Un sistema es una entidad compleja formada por diversos individuos y por una serie de funciones y relaciones entre esos individuos (p.ej.: el sistema solar o el ecosistema de l´Albufera).El científico , o el filósofo aspira a elaborar una teoría del sistema , es decir , un conjunto de enunciados (ecuaciones , fórmulas , esquemas...) que permitan describir adecuadamente el funcionamiento del sistema. Si el sistema funciona tal y como lo dice la teoría , decimos que el sistema es un modelo de la teoría ; p.ej.: el sistema solar es un modelo de la teoría de Kepler. Notemos que "modelo" aquí se emplea como "aquello que cumple una teoría" y no como "aquello que sirve de paradigma a una teoría".Pues bien , en la Teoría del conocimiento , el sistema es el formado por los diversos problemas que el complejo hecho cognoscitivo ofrece , y sus interacciones , ya que estos problemas nunca aparecen aislados. Las teorias para explicarlos son múltiples: Realismo , Empirismo , Idealismo... Pero los modelos que los cumplen ven el hecho cognoscitivo escorado hacia uno u otro problema ; por eso la tendencia actual es ver el conocimiento bajo una teoría más abarcadora , la teoría de la comprensión (Hermenéutica). ((Confrontar para todo esto Conceptos y teoría en la ciencia , de Jesús Mosterín ,pag.149))Una última nota sobre la teoría: las teorias actúan muy distintamente , si son teorias de las Ciencias de la Naturaleza , o si por el contrario , se trata de la teoría de las Ciencias del Espíritu. En la teoría de las Ciencias de la Naturaleza , quien manda es la naturaleza y nos atenemos a una cierta objetividad. En la teoría de las Ciencias del Espíritu , la investigación está determinada por la própia posición del investigador y se investigará desde esta determinación.Conocimiento. La Teoría del Conocimiento será una reflexión crítica sobre el conocimiento , pero "conocimiento" se puede entender de tres formas diferentes:1ª.- La facultad de conocer , que la estudiará la Antropología.2ª.- La actividad de conocer , que la estudiará la psicología.3ª.- El resultado de conocer , el saber que es patrimonio de las diversas ciencias (y otros saberes).Este último sentido constituye el objeto material de la Teoría del Conocimiento. Su objeto formal , estará constituido por aquellos problemas que la reflexión sobre el conocimiento , como tal , y prescindiendo de sus contenidos fácticos , nos va a plantear: alcance y límites del conocimiento , su posibilidad , el problema de la verdad , los modos o tipos de conocimiento , el fundamento del conocimiento etc. Es decir , los llamados " problemas del conocimiento "."La reflexión sobre la naturaleza de nuestro conocimiento da origen a algunosproblemas filosóficos complejos. Estos constituyen la temática de laEpistemología o Teoría del Conocimiento. Casi todos ellos fueron discutidos porlos antiguos griegos y aún hoy existe acaso acuerdo en cuanto a cómo han deresolverse o , por el contrario disolverse"."Cuestiones principales"1º) Distinción entre conocimiento y opinión verdadera. La evidencia.¿Qué es tener evidencia?2º) ¿Hay evidencia inmediata?3º) Nuestro conocimiento , en un momento dado , no es evidente ,¿Cómo justificarlo , cuál es su fundamento?4º) ¿Cuál es el alcace de nuestro conocimiento? En un conocimientoconcreto , ¿Cuál es el criterio para tenerlo como tal?5º) ¿Cuál es el status cognoscitivo de las "verdades de la razón"?6º) ¿Es lo mismo "saber" que "creer"? : el problema de la verdad.((Cfr. Roddrick M. Chisholm , Teoría del conocimiento , Ed.Tecnos ,Madrid , 1982))Noción del problema:"El verbo griego del que se deriva el sustantivo"problema" quiere decir lanzar oarrojar delante. Problema significa , ante todo , algo saliente , p.ej., unpromontorio ; de un modo más correcto , un obstáculo , algo con que meencuentro delante ; y por extensión metafórica , lo que llamamos usualmenteproblema intelectual. Pero repárese en que para que algo me sea obstáculo nobasta con que esté ahí delante ; también tengo delante la pared y ésta no me sirvede obstáculo , sino de abrigo - otro sentido que tiene la voz, -; paraque se convierta en obstáculo, no es suficiente su presencia ante mí: hace faltaque yo necesite pasar al otro lado , precísamente a travésde ella ; entonces es efectivo obstáculo , en la forma concreta de lo que losgriegos llaman "aporía", es decir , falta de poro o agujero por el que salir de unasituación.Podemos decir , por tanto , que uno de los elementos reales que integran unproblema es la situación en que como tal se constituye ; y una formulación de losproblemas que omita esa situación , es en el sentido más riguroso , unaformulación incompleta , abstracta y , ensuma , ficticia". (J.Marias , Introduc.Filosof. pag.24)Aquí cabe resaltar dos aspectos:1º.- Todo problema , también éste , hay que verlo en situación , no sólo a nivel del individuo , que es quien conoce , sino a nivel de los grupos humanos que han ido en su historia decantando sus conocimientos (ciencia , arte , valores ).2º.- El conocimiento no es "problema" porque por su própia índole nos aboca a la verdad ; si al acontecimiento le fuera indiferente la verdad o falsedad , no sería ningún problema. Tenemos que "alcanzar" la verdad.1-1.3 DIVERSOS MOMENTOS HCOS.DE INFLEXIÓN EN LA TEORÍA DEL CONOCIMIENTOLa Teoría del Conocimiento en Occidente ha tenido tres momentos en su história caracterizados por una distinta orientación:PRIMER MOMENTO Si dejamos aparte el modo de pensar mítico y nos centramos en el modo de pensar racional , el primer momento hay que situarlo en Grecia. Los griegos introdujeron lostérminos que utilizamos , "conocer" , "saber" , pero trataron los problemas gnoseológicos subordinados a cuestiones "ontológicas": el conocer era una actividad del alma y , como tal , pertenece al estudio de la misma. Lo mismo los medievales. En estas épocas se concibió el conocer como mera reproducción de las cosas del mundo ya ordenadas y dispuestas para ser conocidas ; pero no ha habido una problemática del conocer como tal , sino una explicación psicológica de la formación de conceptos y una explicaciónontológica de su adecuación a las realidades.SEGUNDO MOMENTO Mundo moderno. Con el Renacimiento comienza verdaderamente la Teoría del Conocimiento ; con anterioridad lo que se ha hecho ha sido una metafísica u ontología del conocimiento.¿Por qué se da esta inflexión , en la que el conocimiento aparece como un problema nuevo , central y , para algunos (p.ej.:Cassier) , exclusivo de la filosofía?Porque en el Renacimiento aparece la ciencia físico-matemática , que aleja la visión de la realidad y su interpretación ingenua e inmediata que nos da el sentido común. Por eso el nuevo objeto de la filosofía en los siglos XVII y XVIII será fundamentar o justificar la ciencia , objetivo que comenzará de modo sistemático en Descartes y culminará en la obra de Kant. Además el "Humanismo" marca una oposición entre hombre y naturaleza , entre sujeto y objeto: será papel de la filosofía reconciliar estaoposición (Hegel).TERCER MOMENTO Época actual. Podemos ver sus causas:a) En el fracaso de los sistemas totalizadores del idealismo alemán (Hegel). Después de Kant , se construyeron grandes sistemas racionales que abarcan la totalidad del universo y de la historia , actitud que se puede resumir con la frase de Hegel: "Todo lo racional es real y todo lo real es racional". Pero la terquedad con que lo real se empeña en desbordar los límites sistemáticos de lo racional obliga a lafilosofía a reconocer que el conocimiento ya no se basa en sí mismo ; no es la realidad última fundante de cualquier otra.b) En el agotamiento del modelo de la razón instrumental o ilustrada. La ciencia , que en la edad moderna sirvió como paradigma de todo saber , es superada desde su interior por sus propias crisis y por tanto debe haber un horizonte que le de sentido. Varias corrientes de pensamiento convergen en este punto : que el conocimiento no es la realidad radical. Estas corrientes son: la "filosofía de la vida" , el "pragmatismo" , la "teoría Diltheiana de la comprensión" , la crítica de la "ideología" , la "filosofía del lenguaje"...La posición a la que llegan todas esas corrientes , muy distintas en otros aspectos , la podemos caracterizar con estas palabras de Bollnow:"La teoría del Conocimiento clásica se caracterizaba por la búsqueda deun punto a partir del cual se pudiera construir un sistema deconocimiento y comprensión seguro de la realidad , previa expulsión detodo lo dudoso. Sin embargo , este punto arquimédico no existe ; elhombre vive en un mundo comprendido y no tiene sentido buscar unpunto inicial que le permita reconstruir la verdad desde la base".( Introd. a la filosof. del comprensión previa y laexperiencia de lo nuevo. Pag.27).En esta línea se situa la obra de H.G. Gadamer Verdad y método , en la cual se insiste en la imposibilidad de partir de cero en el intento de fundamentación , y en la necesidad de asumir los prejuicios en nuestro intento de comprensión. También podíamos situar en esta línea , aunque desde un ángulo muy distinto al de Gadamer , la tesis sostenida por T.S.Kuhn de que el desarrollo de la ciencia se hace atravésde "revoluciones" , consistentes en cambios de "paradigmas".En su obra La filosofía y el espejo de la naturaleza (1979) R.Rorty va un punto más allá.No se trata sólo de desenmascarar el pretendido punto arquimédico "en la teoría del Conocimiento" , ni tampoco de asumir y "comprender" los presupuestos en que toda la problemática del conocimiento y toda teoría se ven inmersos a través de su historia.Ahora hay que rechazar el valor cognoscitivo de la própia hermenéutica ; ésta es sólo una "forma dearreglárselas"."La hermenéutica no es "otra forma de conocer"-la "comprensión" en oposición ala "explicación" (predicativa). Es mejor considerarla como otra forma dearreglárselas. Sería mejor para la claridad filosófica que entregáramos la idea"cognición" a la ciencia predictiva , y dejáramos de preocuparnos por los"métodos cognitivos alternativos". La palabra conocimiento no parecería dignade que se luchara por ella si no fuera por la tradición kantiana de que ser filósofoes tener una "Teoría del Conocimiento" , y la tradición platónica de que la acciónque no está basada en el conocimiento de la verdad de las proposiciones es"irracional" (Op. cit. Pags. 321 , 322 )Parece que en este pasaje Rorty apunta en una dirección similar a la de Feyerabend en su propuesta de metodología anarquista: "Todo sirve" (Tratado contra el método , Pag.12). Pero , me parece que va másallá en la dirección en que el conocimiento y la comprensión se situen en la línea marcada por Wittgesteinen su segunda época: el lenguaje (y el conocimiento) expresan "formas de vida" , ninguna de las cuales es privilegiada respecto a las demás.Un argumento similar lo encontramos , como veremos más tarde , en Ortega y Gasset. Másallá de su sabida distinción entre "ideas" y "creencias" , está su distinción entre "conocer" y "pensar". "Conocer" es una forma histórica que ha adquirido el "pensar". Y esto , el "pensar" sea cual sea la formahistórica que haya tenido o que tenga en un futuro , es lo que ineludiblemente el hombre tiene que hacerpara pervivir.Por todo lo dicho debemos darnos cuenta que una teoría del conocimiento aboca ,necesáriamente , a una teoría de la comprensión , y ésta , a su vez , sólo es posible realizarla refiriéndola ,en última instáncia , a la realidad en que consista el hombre , que es , al fín y a la postre , quien conoce.1.2 FENOMENOLOGÍA DEL CONOCIMIENTOEl conocimiento es un hecho del cual no podemos dudar, de lo que podemos dudar es del valor del conocimiento. Ese hecho se nos da en forma bruta , innanalizada , y debemos llegar a ver en qué consiste formalmente el conocimiento , cual es su estructura , para poder ver en ella enraizados todos los problemas que suscita el proceso cognoscitivo para poder escoger la teoría más adecuada , de modo que las soluciones a dichos problemas sean modelos de la teoría. Para ello tenemos que realizar una descripción del conocimiento en cuanto tal ; pero antes será conveniente decir una palabra sobre el método fenomenológico que vamos a utilizar para hacer dicha descripción.1-2.1 EL MÉTODO FENOMENOLÓGICO DE HUSSERLE.Husserl concibió la fenomenología como una ciencia estricta que consistía en la descripción de las esencias ("cosas") tal como se nos presentan por ellas mismas sin que prejuzguemos nada al hacer esa descripción ; para ello utilizaba un método (que podíamos resumir en un eslogan "a las cosas mismas" de su época de Gotinga): el llamado método fenomenológico. Es cierto que él nunca sistematizó ese método, ni creo que fuera posible porque en él se implica toda una concepción de la actividad filosófica y de la realidad misma ; más bien lo fue elaborando conforme se le presentaban los problemas , "in actu exercito". Con todo , siguiendo a I.M.Bochenski en Los métodos actuales del pensamiento , Pag.41 y ss. , podemos distinguir dos momentos diferentes en el desarrollo husserliano del método fenomenológico:a) La reducción eidética , realizada principalmente en las Investigaciones lógicas (1901)b) La reducción fenomenológica estricta , que aparece a partir de Ideas I (1913).Como método apto para comienzo de una investigación filosófica parece adecuado el primero , la reducción eidética ; la reducción fenomenológica , por el contrario , está muy vinculada a la própiafilosofía de Husserl. Por lo tanto , describiremos sólo la reducción eidética.Reducción eidética: El método fenomenológico pretende una visión intelectual del objeto basada en una intuición de lo dado : consiste en dejar que las cosas se nos manifiesten por ellas mismas.Para llegar a esta visión necesitamos una triple eliminación o reducción ( ). (Advirtamos que en este contexto "reducir" no significa "negar" , sino poner entre "paréntesis" , "prescindir de").1ª.- de todo lo subjetivo: En la investigación debe orientarse el pensamiento exclusívamente hacia el objeto. El método fenomenológico , insistamos , es un método de intuición de lo dado , de la "cosa". A lo dado , Husserl lo llama "fenómeno" , del griego , lo que aparece , lo que está claro a nuestra vista , lo que está a la luz. Pero el mismo acto de intuir es una enunciación intelectual delfenómeno , un legein.La intuición de un fenómeno se opone a todo conocimiento discursivo , a toda abstracción y en general a todo psicologismo que tenga en cuenta la constitución fáctica del sujeto que conoce. Se trata de ir "a las cosas mismas" , y en este sentido , es una práctica que continúa la actitud teorética de la tradición griega.2ª.- de todo lo teórico: (hipótesis , demostraciones o cualquier forma de saber ya adqurido) , de manera que sólo entre en cuestión lo dado , el fenómeno.Con esta exclusión no se pretende rechazar todo lo teórico sin más. Pero sería admitido sólo después de su fundamentación fenomenológica. Es este el camino para llegar a la ciencia fundante , a la "lógica pura".3ª.- de toda tradición: de todo lo que se ha venido enseñando hasta ahora sobre el objeto. Másaún , debemos olvidarnos del nivel en que nos sitúe en este momento la ciencia ; ésta es un saber no fundamentado: de ahí su estado de crisis.En el objeto mismo hay que hacer otra doble reducción: a) de la existencia de la cosa y quedarse con lo que el objeto es (quidditas).b) De la quidditas hay que llegar a la esencia o eidos : estructura fundamental del objeto , sin referencia a nada accesorio. Advirtamos que esta esencia o eidos fenomenológico , no es algo oculto , sino precísamente un mostrarse "a sí en sí mismo" (Heidegger). Por tanto es un concepto más amplio que el aristotélico ; Aristóteles admite junto a la esencia otras propiedades o determinaciones en conexión necesaria con ella. La esencia fenomenológica incluye , en cambio , todo lo que está necesáriemente unido en el fenómeno , en el manifestarse.1-2.2 DESCRIPCIÓN FENOMENOLÓGICA DEL CONOCIMIENTOUna vez presentado el método fenomenológico tenemos que aplicarlo al objeto de nuestra investgación el conocimiento. El conocimiento tal como se nos presenta , sin supuestos , lo podemos describir del siguiente modo:1º.- El conocimiento es un proceso de carácter dialogal , de nosotros como sujeto con algo , lo conocido , que es el objeto. Este diálogo se mediatiza por la conciencia ; ésta actúa como frontera que , al mismo tiempo que une , separa o divide.2º.- El conocimiento es siempre un fenómeno consciente.3º.- Este proceso consciente supone siempre una dualidad , sujeto - objeto , no puede existir el proceso cognoscitivo sin esta dualidad. El sujeto y el objeto son irreductibles uno a otro en el mismo proceso cognoscitivo. En el lenguaje filosófico se expresa esto diciendo que sujeto y objeto songnoseológicamente trascendentes uno a otro. Esto quiere decir:a) No puede haber proceso cognoscitivo si falta uno de ellos.b) El sujeto se constituye formalmente como sujeto frente al objeto. El objeto se constituye formalmente como objeto frente al sujeto.c) Sujeto y objeto no tienen que ser necesáriamente dos realidades distintas (abriéndose al fenómenode la autoconciencia). Pero formalmente se tiene que constituir como tales oponiéndose.Para que la descripción fenomenológica del conocimiento quede completa debemos detenernos un momento en el análisis de la conciencia , escenario en que el proceso cognoscitivo se nos da. En esteanálisis la conciencia presenta dos caracteres fundamentales:A)Carácter estructurado y estructurante de la conciencia: Según Gurwistch en El campo de la conciencia (Pag.71), la conciencia se nos presenta como un "campo" estructurado en tres regiones o dimensiones:1º.- "Tema focal": aquello que enriquece actualmente nuestra conciencia , polarizando hacia sínuestra atención.2º.- "Campo temático": constituido por la totalidad de datos que nos son presentes materialmentey de algún modo objetívamente ; los considerados pertinentes al tema focal , sin que recaiga sobre ellosdiréctamente la atención.3º.- "Límite marginal": u horizonte del conjunto constituido por el tema focal y el campotemático. En sí está formado también por un elenco de datos , pero no se acusan como tales en la conciencia por su alejamiento del centro focal de la atención ; no son pertinentes al tema focal.Estas tres dimensiones que integran el "campo de conciencia", no son imposiciones absolutas que la conciencia recibe pasivamente , sino que forman una estructura organizada por la conciencia. La conciencia no es un puro escenario donde acontecen las experiencias (al estilo de Hume) , sino que estructura lo que en ella se realiza.B) Intencionalidad de la conciencia: Francisco Brentano , en su obra Psicología desde el punto de vista empírico , Cap.1º, Párrafo 5º al tratar de distinguir los fenómenos físicos de los fenómenos psíquicos considera que el carácter formal de estos es la "intencionalidad" y dice:"¿ Qué carácter positivo podemos , pues , indicar ? ¿ O acaso no hay ningunadefinición positiva , que valga conjúntamente para todos los fenómenos psíquicos ? Todofenómeno psíquico está caracterizado por lo que los escolásticos de la Edad Media hanllamado la inexistencia (no significa la no existencia , sino la existencia en) intencional (omental ) de un objeto , y que nosotros llamaríamos , si bien con expresiones noentéramente inequívocas , la referécia a un contenido , la dirección hacia un objeto (por elcual no hay que entender aquí una realidad ), o la objetividad inmanente."En Brentano , pues , la "intencionalidad" tiene dos aspectos: a) la "existencia en" o mental de los objetos psíquicos , y b) la "dirección hacia" un contenido u objeto de todo acto psíquico , frente alfenómeno físico.Husserl toma de Brentano la idea de "intencionalidad" , pero subraya el aspecto de "dirección hacia". Aunque en las Investigaciones Lógicas todavia se nota la influencia de Bretano en cuanto al doble aspécto del concepto de "intencionalidad"; en Ideas se ve más claro como acentúa el segundo aspecto:"Entendimos por intencionalidad la peculiaridad de las vivencias de " ser conciencia de algo ". Ante todo nos salió al encuentro esta maravillosa peculiaridad , a la que retrotraen todos los enigmas de la teoría de la razón y de la metafísica , en el cogito explícito: una percepción es percepción de algo , digamos de una cosa...En todo cogito actual , una "mirada" que irradia del yo puro , se dirige al "objeto" que es el respectivo correlato de la conciencia , a la cosa , la relación objetiva , etc. Y lleva a cabo la muy diversa conciencia de él" ( Ideas 1 , Cap. 84 , Pag.199 ).En obras posteriores Husserl destacó el carácter "constitutivo" de las intenciones : éstas son "cumplimiento" de los actos intencionales. En esta dirección avanzaríamos hacia un idealismo de carácter trascendental.Pero lo que nos interesa de Ideas 1 es que destaca el aspecto direccional de la conciencia ; que esta direccionalidad es múltiple y que constituye el pleno sentido de la conciencia :"La intencionalidad es lo que caracteriza la conciencia en su pleno sentido y lo queautoriza para designar a la vez la corriente entera de las vivencias como corriente deconciencia y como unidad de una conciencia". ( Ibid. Pag. 198)Limitaciones a estos análisis :1º.- Para Husserl parece que aprehender o captar un objeto , sea posesionarse de él sólo intelectualmente ; limita el significado de aprehender a ser el correlato mental de un enunciado sobre un objeto. Pero hay otras formas de aprehender : emotivas , práxicas , etc...2º.- ¿ Cuál es la naturaleza de lo aprehendido ? , No puede ser el objeto como tal objeto , porque la naturaleza de la conciencia es intencional y estructurante y el objeto es mucho más rico de lo que se presenta en la conciencia. El objeto se dividirá en dos ; pensemos que el objeto tiene dos dimensiones , el objeto como realidad , la cosa , y el objeto conocido que es el que está en mi conciencia cuando conozco ,y ambas pueden coincidir total o parcialmente en su estructura. El que el objeto conocido sea más pobreque el real , no quiere decir que lo que yo conozco sea diferente de lo real.3º.- La proporción de elementos sensibles , intelectuales , emotivos , culturales , sociales , etc... que intervienen en la representación del objeto por parte del sujeto , hacen prácticamente imposible la asepsiade la fenomenológica.1-3 PROBLEMAS DEL CONOCIMIENTO1-3.1 POSIBILIDAD DEL CONOCIMIENTO。
泰勒公式的推导及应用
泰勒公式的推导及应用泰勒公式是一种重要的数学工具,它可以将一个函数在某个点处展开成一个无限次可导函数的幂级数。
这个级数在某些情况下非常有用,可以用来近似数值计算和研究函数的性质。
本文将简要介绍泰勒公式的推导过程和一些应用。
一、泰勒公式的推导设$f(x)$在$x=a$处$n$阶可导,则$f(x)$在$x=a$处的$n$阶泰勒展开式为:$$f(x)=f(a)+\frac{f^{(1)}(a)}{1!}(x-a)+\frac{f^{(2)}(a)}{2!}(x-a)^2+\cdots+\frac{f^{(n)}(a)}{n!}(x-a)^n+R_n(x)$$其中$R_n(x)$为$f(x)$在$x=a$处的$n$阶拉格朗日余项,具体表达式为:$$R_n(x)=\frac{f^{(n+1)}(\xi)}{(n+1)!}(x-a)^{n+1}$$其中$\xi$介于$x$和$a$之间。
二、泰勒公式的应用1. 求函数的近似值泰勒公式可以用来近似计算函数的值,特别是在求解复杂问题时非常有用。
例如,如果我们需要计算$\sin0.1$的值,可以使用泰勒公式展开$\sin x$:$$\sin x=x-\frac{x^3}{3!}+\frac{x^5}{5!}-\cdots$$当$x=0.1$时,忽略高阶项,得到:$$\sin 0.1\approx 0.1-\frac{0.1^3}{3!}=0.0998*******$$这个值与真实值$0.0998*******$非常接近。
2. 求函数的导数泰勒公式可以用来求函数的导数,尤其是对于某些复杂的函数,可以通过泰勒公式求导简化计算过程。
例如,对于$f(x)=\sin x$,我们可以使用泰勒公式展开$\sin x$:$$\sin x=x-\frac{x^3}{3!}+\frac{x^5}{5!}-\cdots$$对该式两边求导,得到:$$\cos x=1-\frac{x^2}{2!}+\frac{x^4}{4!}-\cdots$$这个式子可以用来计算$\cos x$的值,也可以用来求导。
泰勒公式-文献翻译讲解
重庆理工大学文献翻译二级学院数学与统计学院班级 111010101 学生姓名学号 11101010110复杂矩阵用泰勒公式的解决方法一个复杂的R×R矩阵A中的解决方法Rλ(A)是与A的频谱ΣA空交集的任何域名的解析函数。
在任何给定的λ0∉ΣA的附近著名的泰勒展开Rλ(A)的修改考虑到Rλ0只有第一大国(A)是线性无关。
在这个框架的主要工具给出了多变量多项式查看MATHML源取决于不变V1,V2,...,VRRλ(A)的(m为最小多项式的程度)。
这些功能被用于以代表的Rλ0(A)的随后的权力的系数作为它们的第一米的线性组合。
一简介如在[1]中,预解Rλ(A)的≔(λI-A)一种非奇异正方形矩阵A(Ⅰ表示单位矩阵)-1所示的希尔伯特同一性的后果是参数λ的一个解析函数在与A的频谱ΣA因此,空交集的任何域D使用泰勒展开任何固定λ0∈D的附近,我们可以在[1]Rλ(A)的表示公式发现使用Rλ0的一切权力(A)。
在这篇文章中,通过使用一些前面的结果回忆说,例如,在[2]中,我们写下使用Rλ0(A)的权力,只有有限数量的表示公式。
这似乎是因为Rλ0的(A)是线性无关的只有第一个权力是自然的。
在此框架的主要工具是由多变量多项式给出查看MATHML来源(参见[2],[3],[4],[5]和[6]),根据不同的不变量V1,V2,...,vr中的Rλ(A);这里m表示极小多项式的程度。
二权力矩阵和F k,n功能我们还记得在本节一定的成效上表示公式矩阵的权力(见[2][3][4][5][6]和其参考文献)。
为简单起见,我们指的是情况下,当基质是非贬损使得M = R。
命题2.1。
设A是一个矩阵,由U1表示,U2,...,UR A的不变量,并通过∑=--=-=rj jr j j U A I p 0)1()det()(λλλ其特征多项式(按照惯例u0≔1);那么对于A 的非负整数指数的权力下表示公式也是如此:Iu u F A u u F A u u F r n r r r n r r n n )...,(...)...,()...,(A 11,211,2111,1-----+++=功能Fk,n(u1...ur)该显示为系数(2.1)由递推关系定义)....,()1(...)....,()....,()....,(1,112,211,11,r r n k r r r n k r n k r n k u u F u u u F u u u F u u u F -----++-=)1;......1(-≥=n rk 和初始条件:)....1,(,).....(,12,1r h k u u F h k r h k r ==-+-δ此外,如果A 是非奇异(ur ≠0),则式(2.1)仍然保持对于n 的负值,只要我们定义FK ,n 功能对于n 的负值如下:)1,......,()...(113,11,rr r r r n k r r n k u u u u u F u u F --+-+-=)1;...1(-<=n r k三 泰勒展开式的解决对策我们认为解决方法矩阵R λ(A )定义如下:.)(:)(1--=≡A I A R R λλλ注意,有时有标志的公式的变化。
泰勒公式 G翻译完成
Taylor ’s FormulaG. B. FollandThere ’s a lot more to be said about Taylor ’s formula than the brief discussion on pp.113–4 of Apostol. Let me begin with a few de finitions.De finitions. A function f de fined on an interval I is called k times di ff erentiable on I if the derivatives(),,,kf f f ''' exist and are finite on I , and f is said to be of class k C onI if these derivatives are all continuous on I . (Note that if f is k times di ff erentiable, thederivatives()1,,,k f f f -''' are necessarily continuous, by Theorem 5.3; the only question is thecontinuity of()k f .) If f is (at least)k times di ff erentiable on an open interval I and c I ∈its k th order Taylor polynomial about c is the polynomial ()()()(),0!j kjk cj f c P x x c j ==-∑(where,of course, the “zeroth derivative ” ()0f is f itself), and its k th order Taylor remainder is thedi fference()()(),,.k c k c R x f x P x =-Remark 1. The k th order Taylor polynomial (),k c p x is a polynomial of degree at most k ,but its degree may be less than k because()()k f c might be zero.Remark 2. We have()(),k c p c f c = , and by di ff erentiating the formula for(),k c p x repeatedly and then setting x c = we see that ()()()(),j j k c p c f c = for j k ≤. That is, ,k c p is the polynomial of degree k ≤ whose derivatives of order k ≤at c agree with those off .For future reference, here are a fe w frequently used examples of Taylor polynomials:()()()()()()()()()()()()(),002,00221,0021,10;!1cos ;2!1sin ;21!1(1)log ;j xk j kjj k kj jj k k j j j k j kx f x e P x j x f x x P x j x f x x P x j x f x x P x j≤≤≤≤+≤≤-≤≤==-==-==+--==∑∑∑∑Note that (for example) 2112x -is both the 2nd order and the 3rd order Taylor polynomial of cos x , because the cubic term in its Taylor expansion vanishes. (Also note that in higher mathematics the natural logarithm function is almost always called log rather than ln.)For 1k = we have ()()()()1,cP x f c f c x c '=+-; this is the linear function whose graphis the tangent line to the graph of f at x c =. Just as this tangent line is the straight line that bestapproximates the graph of f nearx c =, we shall see that (),k c P x is the polynomialof degree k ≤ that best approximates ()f x near x c = . To justify this assertion we needto see that the remainder(),k c R x is suitably small near x c =, and there are several waysof making this precise. The first one is simply this: the remainder(),k c R x tends to zero asx c → faster than any nonzero term in the polynomial (),k c P x , that is, faster than ()kx c - .Here is the result:Theorem 1. Suppose f is k times differentiable in an open interval I containing the pointc . Th en ()()()()(),,limlim0k c k c kkx cx cR x f x P x x c x c →→-==--Proof. Since f and its derivatives up to order k agree with,k c p and its derivatives up to orderk at x c =, the di ff erence ,k c R and its derivatives up to order k vanish at x c =. Moreover,()kx c -and its derivatives up to order 1k - also vanish at x c =, so we can apply l ’Hopital ’srule k times to obtain ()()()()()(),,0limlim0!11k k c k c kx cx cR x R x k x c k k x c →→===--- There is a convenient notation to describe the situation in Theorem 1: we say that()()(),kk c R x x c ο=-as x c → ,meaning that(),k c R x is of smaller order than ()kx c - as x c →. More generally, if g andh are two functions, we say that ()()()h x g x ο= as x c → (where c might be ±∞) if()()/0h x g x →as x c →. The symbol ()()g x ο is pronounced “little oh of g of x ”; itdoes not denote any particular function, but rather is a shorthand way of describing any function that is of smaller order than()g x as x c → . For example, Corollary 1 of l ’H ˆpital ’s ruleo (see thenotes on l ’H ˆpital ’s rule) says that for any 0a >, ()axx e ο=and ()log ax x ο=as x →∞,and ()log ax xο-= as x c →. Another example: saying that ()()1h x ο= asx c → ,simply means that ()lim 0x ch x →=.In order to simplify notation, in the following discussion we shall assume that 0c = and writek P instead of ,k c P . (The Taylor polynomial ,0k k P P = is often called the k th orderMaclaurin polynomial of f .) There is no loss of generality in doing this, as one can always reduce to the case 0c = by making the change of variable x x c =-and regarding all functions in question as functions of xrather thanx .泰勒公式 G.B.Foolland比之Apostol 在113-4讨论的简单讨论,泰勒公式有很多可以说的地方,现在让我开始一些简单的定义。
第七节泰勒公式课件
02
泰勒公式的推导过程
泰勒公式的多项式逼近
泰勒公式通过将复杂的函数展 开为多项式,以逼近原函数。
多项式的每一项系数由函数在 某点的导数值确定,从而反映 了函数在该点的局部性质。
通过选取不同的基点,可以得 到不同形式的泰勒多项式逼近 。
泰勒公式的余项表示
泰勒公式的余项用于描述多项式 逼近与原函数之间的误差。
余项有多种形式,常见的有余项 、拉格朗日余项和柯西余项等。
余项的阶数决定了多项式逼近的 精度,阶数越高,逼近效果越好
,但同时计算也越复杂。
泰勒公式的收敛性分析
泰勒公式的收敛性是指当多项式的项数趋于无穷时,多项式逼近的结果是否趋近于 原函数。
收敛性的分析涉及到函数的可导性、收敛域和余项的性质等因素。
物理建模
在物理建模中,泰勒公式 用于描述物理现象的数学 模型,如振动、波动等。
泰勒公式的历史背景
起源
泰勒公式最早由英国数学 家布鲁克·泰勒在18世纪提 出。
发展
随着数学的发展,泰勒公 式的应用范围不断扩大, 成为数学分析和物理研究 中的重要工具。
应用
泰勒公式的应用不仅限于 数学和物理领域,还扩展 到了经济学、工程学等其 他领域。
第七节泰勒公式课件
目录
• 泰勒公式简介 • 泰勒公式的推导过程 • 泰勒公式的应用举例 • 泰勒公式的扩展与推广 • 泰勒公式的注意事项与限制条件
01
泰勒公式简介
泰勒公式的定义
01
泰勒公式定义
泰勒公式是一个用无穷级数表示函数的方法,它将一个函数展开成无穷
项的加和,每一项都是函数在某一点的导数与该点的x值的乘积。
对于一些需要高精度计算的情况,如科学计算、工程计算等 ,需要选择高阶的泰勒多项式。
new第八节多元函数的Taylor公式.ppt
x f ) F x fF d z (f y x 。 d x F x fF y z
第八节 多元函数的Taylor公式与极值
• 多元函数的Taylor公式 • 多元函数的极值 • 函数的最大值与最小值
一 、 多元函数的 Taylor 公式 与一元函数类似 , 多元函数也有 Taylor 公式 , 下面
( 8 . 6 ) 当 余 项 取 形 式 ( 8 . 4 ) 时 称 为 L a g r a n g e 余 项 , 当 余 项 取
形 式 ( 8 . 5 ) 时 称 为 P e a n o 余 项 .
例 1 写 出在 f ( x , yx ) l n ( 1 y ) ( 0 , 0 ) 的 T a y l o r 公 式 . f f 1 1 [ 解 ] f ( 0 , 0 ) 0 , ( 1 x y ), ( 1 x y ) , x y 2 2 f f 2 2 ( 1 x y ) , ( 1 x y ) , 2 x x y
由 (8 .1 )得
1 k j ln(1 x y) 0 Ck (1)k1(k 1)! x j yk j k1 k! j0 1 n1 j n n j n1 j C ( 1) n !(1 x y ) x y n1 (n 1)! j1 (1)k1 k j j k j (1)n ( x y)n1 Ck x y n k n 1 (1 x y ) k1 j0
n
n
(0 1)
k 1 n n 1 (1 ) (1 ) t k [ 解 法 2 ] 根 据 l n ( 1 t ) = t , n n 1 ( 1 t ) k 1 k 0 1 取 t xy ,得 n
D3_3泰勒.pdf
二、几个初等函数的麦克劳林公式 第三节一、泰勒公式的建立三、泰勒公式的应用— 应用用多项式近似表示函数理论分析近似计算泰勒 ( Taylor )公式第三章特点:)(01x p ′)(0x f =)(0x f ′=一、泰勒公式的建立)(x f xy)(x f y =o))(()(000x x x f x f −′+≈)(1x p 以直代曲0x )(1x p )(01x p 在微分应用中已知近似公式 :需要解决的问题如何提高精度 ?如何估计误差 ?xx 的一次多项式1. 求 n 次近似多项式要求:,)(x p n )(0!212x p a n′′=,)(0x f ′′=,L )(0)(!1x p a n n n n =)(0)(x fn =故=)(x p n )(0x f ))((00x x x f −′+L +!21!1n n n x x x f ))((00)(−+!1n 200))((x x x f −′′+!21令=)(x p n 则=′)(x p n =′′)(x p n L L L L n a n !=)()(x p n n )(00x p a n =,)(0x f =,)()(00x f x p n =)(01x p a n ′=,)(0x f ′=1a )(202x x a −+10)(−−++n n x x a n L 2!2a 20)()1(−−−++n n x x a n n L ,)()(00x f x p n ′=′)()(,0)(0)(x f x p n n n =L 0a nn x x a x x a x x a )()()(020201−++−+−+L)0(之间与在n x ξξ )( )(10+−=n n x x x R )(2)1( )(0)(x n R n n n n −+=ξξL 2. 余项估计)()()(x p x f x R n n −=令(称为余项) ,)(0x R n )(0x R n′=0)(0)(===x R n n L 10)()(+−n n x x x R n n x n R ))(1()(011−+′=ξξ ))(1( )(011n n x n R −+′=ξξ1022)()1()(−−+′′=n n x n n R ξξL=!)1()()1(+=+n R n n ξ则有)(0x R n −0−)(0x R n ′−0−)(0)(x R n n −0−x )01(之间与在x x ξ)102(之间与在ξξx)()()(x p x f x R n n −=10)()(+−n n x x x R !)1()()1(+=+n R n n ξ)0(之间与在x x ξ,0)()1(=+x p n n Q10)1()(!)1()()(++−+=n n n x x n fx R ξ)()()1()1(x fx R n n n ++=∴时的某邻域内当在M x f x n ≤+)()1(0)0(之间与在x x ξ1!)1()(+−+≤n n x x n M x R )())(()(00x x x x o x R nn →−=∴公式 ① 称为的 n 阶泰勒公式 .)(x f 公式 ② 称为n 阶泰勒公式的拉格朗日余项 .泰勒中值定理 :内具有的某开区间在包含若),()(0b a x x f 1+n 直到阶的导数 ,),(b a x ∈时, 有=)(x f )(0x f ))((00x x x f −′+200)(!2)(x x x f −′′+L+n n x x n x f )(!)(00)(−+)(x R n +①其中10)1()(!)1()()(++−+=n n n x x n fx R ξ②则当)0(之间与在x x ξ公式 ③ 称为n 阶泰勒公式的佩亚诺(Peano Peano)) 余项 .在不需要余项的精确表达式时 , 泰勒公式可写为L+=)(x f )(0x f ))((00x x x f −′+200)(!2)(x x x f −′′+n n x x n x f )(!)(00)(−+])[(0n x x o −+])[()(0nn x x o x R −=注意到③④* 可以证明:阶的导数有直到在点n x x f 0)(④ 式成立特例:(1) 当 n = 0 时, 泰勒公式变为=)(x f )(0x f ))((0x x f −′+ξ(2) 当 n = 1 时, 泰勒公式变为给出拉格朗日中值定理=)(x f )(0x f ))((00x x x f −′+20)(!2)(x x f −′′+ξ可见≈)(x f )(0x f ))((00x x x f −′+201)(!2)()(x x f x R −′′=ξ误差=)(x f )(0x f ))((00x x x f −′+L +10)1()(!)1()(++−++n n x x n f ξ200)(!2)(x x x f −′′+n n x x n x f )(!)(00)(−+fd )0(之间与在x x ξ)0(之间与在x x ξ)0(之间与在x x ξ)0(之间与在x x ξ称为麦克劳林( Maclaurin )公式 .,)10(,00<<==θθξx x 则有=)(x f )0(f x f )0(′+L +1)1(!)1()(++++n n x n x f θ2!2)0(x f ′′+nn xn f !)0()(+在泰勒公式中若取=)(x f )(0x f ))((00x x x f −′+L +10)1()(!)1()(++−++n n x x n f ξ200)(!2)(x x x f −′′+n n x x n x f)(!)(00)(−+)0(之间与在x x ξ≈)(x f )0(f x f )0(′+L +,)()1(M x fn ≤+则有误差估计式1!)1()(++≤n n x n M x R 2!2)0(x f ′′+nn x n f !)0()(+若在公式成立的区间上由此得近似公式二、几个初等函数的麦克劳林公式xe xf =)()1(,)()(xk e x f =Q),2,1(1)0()(L ==k fk xe ∴1=x +!33x +L +!n xn+)(x R n +!22x +其中=)(x R n !)1(+n )10(<<θ1+n x xe θ)sin(+x xx f sin )()2(==)()(x f k Q x sin ∴x =!33x −!55x +!)12(12−+−m x m )(2x R m +其中=)(2x R m )sin(212πθ++m x 2π⋅k 2sin )0()(πk f k =⎩⎨⎧=m k 2=,012−=m k ,)1(1−−m ),2,1(L =m L −1)1(−−m )10(<<θ12+m x!)12(+m )cos()1(x m θ−!)2(2m x m+xx f cos )()3(=类似可得x cos 1=!22x −!44x +)(12x R m ++其中=+)(12x R m !)22(+m )cos()1(1x m θ+−)10(<<θL +m )1(−22+m x)1()1()()4(−>+=x x x f α=)()(x f k Qα)1(x +∴1=x α+2x n x )(x R n +其中=)(x R n 11)1(!)1()()1(+−−++−−n n xx n n αθαααL )10(<<θkx k −++−−αααα)1)(1()1(L )1()1()0()(+−−=k fk αααL L+),2,1(L =k !2 +)1(−αα!n +)1()1(+−−n αααL)1()1ln()()5(−>+=x x x f 已知)1ln(x +x =22x −33x +nx n+)(x R n +其中=)(x R n 11)1(1)1(++++−n n n x xn θ)10(<<θL −1)1(−−n 类似可得=)()(x f k kk x k )1(!)1()1(1+−−−),2,1(L =k三、泰勒公式的应用1. 在近似计算中的应用误差1!)1()(++≤n n x n M x R M 为)()1(x fn +在包含 0 , x 的某区间上的上界.需解问题的类型:1) 已知 x 和误差限 , 要求确定项数 n ;2) 已知项数 n 和 x , 计算近似值并估计误差;3) 已知项数 n 和误差限 , 确定公式中 x 的适用范围.≈)(x f )0(f x f )0(′+L +2!2)0(x f ′′+nn xn f !)0()(+已知例1. 计算无理数 e 的近似值 , 使误差不超过.106−解:xe !)1(++n x e θ1+n x 令 x = 1 , 得e )10(!)1(!1!2111<<++++++=θθn en L )10(<<θ由于,30<<<e e θ欲使)1(n R !)1(3+<n 610−<由计算可知当 n = 9 时上式成立 ,因此e !91!2111++++≈L 718281.2=x e 1=x +!33x +L +!n x n +!22x +的麦克劳林公式为说明: 注意舍入误差对计算结果的影响.本例若每项四舍五入到小数点后 6 位,则各项舍入误差之和不超过,105.076−××总误差为6105.07−××610−+6105−×<这时得到的近似值不能保证误差不超过.106−因此计算时中间结果应比精度要求多取一位 .e !91!2111++++≈L例2. 用近似公式!21cos 2xx −≈计算 cos x 的近似值,使其精确到 0.005 , 试确定 x 的适用范围.解:近似公式的误差)cos(!4)(43x xx R θ=244x ≤令005.0244≤x解得588.0≤x 即当588.0≤x 时, 由给定的近似公式计算的结果能准确到 0.005 .2. 利用泰勒公式求极限例3. 求.43443lim 20xx x x −−++→解:由于x 4312+=43+x 21)1(43x +[] 2=)(14321x ⋅+⋅+!21)1(2121−243)(x )(2x o +用洛必塔法则不方便 !2x 用泰勒公式将分子展到项,11)1(!)1()()1(+−−++−−+n n x x n n αθαααL n x ! n +)1()1(+−−n αααL α)1(x +1=x α+2x L +!2 +)1(−αα)10(<<θx 34−21)1(243x −=2=20 lim xx →=∴原式)(2216921x o x +⋅−329−=x 43−)(2216941x o x +⋅−2=x 43+)(2216941x o x +⋅−11)1(!)1()()1(+−−++−−+n n x x n n αθαααL n x !n +)1()1(+−−n αααL α)1(x +1=x α+2x L +!2 +)1(−αα)10(<<θ3. 利用泰勒公式证明不等式例4. 证明).0(82112>−+>+x xx x 证:21)1(1x x +=+Q21x +=2)121(21!21x −⋅+325)1)(221)(121(21!31xx −+−−⋅+θ)10(<<θ3225)1(161821x x x x −++−+=θ)0(82112>−+>+∴x x x x内容小结1. 泰勒公式其中余项))((0nx x o −=当00=x 时为麦克劳林公式 .=)(x f )(0x f ))((00x x x f −′+200)(!2)(x x x f −′′+L+n n x x n x f )(!)(00)(−+)(x R n +10)1()(!)1()()(++−+=n n n x x n fx R ξ)0(之间与在x x ξ2. 常用函数的麦克劳林公式 ( P140 ~ P142 ),x e ,)1ln(x +,sin x ,cos x α)1(x +3. 泰勒公式的应用(1) 近似计算(3) 其他应用求极限 , 证明不等式 等.(2) 利用多项式逼近函数 , x sin 例如4224642024612!)12()1(9!917!715!513!311sin −−−−+++−+−=n n x xx x x x x n L )(2nx o +!33x x y −=!5!353x x x y +−=!7!5!3753x xx x y −+−=xy sin =x y =xsin 泰勒多项式逼近12!)12()1(9!917!715!513!311sin −−−−+++−+−=n n x xx x x x x n L )(2nx o +xsin 4224642246xy sin =!9!7!5!39753x x x xx y +−+−=!11!9!7!5!3119753x x x x x x y −+−+−=泰勒多项式逼近思考与练习计算.3cos 2lim 402xx ex x −+→)(!2114422x o x x e x +++=Q )(!4!21cos 542x o x x x ++−=)()!412!21(3cos 2442x o x x e x +⋅+=−+∴127)(lim 4441270=+=→xx o x x 解:原式作业P143 1 ;4 ; 5 ; 7 ; 8;10(1),(2)泰勒(1685–1731)英国数学家,他早期是牛顿学派最优秀的代表人物之一 , 重要著作有:《正的和反的增量方法》(1715)《线性透视论》(1719)他在1712 年就得到了现代形式的泰勒公式 .他是有限差分理论的奠基人 .麦克劳林(1698–1746)英国数学家,著作有:《流数论》(1742)《有机几何学》(1720)《代数论》(1742)在第一本著作中给出了后人以他的名字命名的麦克劳林级数 .,]1,0[)(上具有三阶连续导数在设函数x f ,0)(,2)1(,1)0(21=′==f f f .24)(,≥′′ξξf 使一点=)(x f )(21之间与在其中x ζ,]1,0[∈x 由题设对证:备用题 1.321))((!31−′′′+x f ζ)(21f221)(−x )(!2121f ′′+))((2121−′+x f有)(21f =221)(−x )(!2121f ′′+321))((!31−′′′+x f ζ内至少存在证明)1,0(且得分别令,1,0=x)),0((211∈ζ)(21f=))1,((212∈ζ3211)(!3)(−′′′+ζf 3212)(!3)(ζf ′′′+)0(1f =)(21f =22121)(!2)(−′′+f )1(2f =22121)(!2)(f ′′+=1下式减上式 , 得[])()(48112ζζf f ′′−′′[])()(48112ζζf f ′′+′′≤)(241ξf ′′≤)10(<<ξ令))(,)((max )(12ζζξf f f ′′′′=′′24)(≥′′ξfe )10(!)1(!1!2111<<++++++=θθn en L 两边同乘 n !e n != 整数 +)10(1<<+θθn e假设 e 为有理数qp( p , q 为正整数) ,则当时,q n ≥等式左边为整数;矛盾 !2. 证明 e 为无理数 .证:2≥n 时,当故 e 为无理数 .等式右边不可能为整数.。
泰勒公式的应用英语文献
泰勒公式的应用英语文献Taylor's Theorem is a powerful mathematical tool that can be applied in many different areas. It enables us to solve various problems related to integral calculus, differential equations, and optimization. In particular, Taylor's theorem can be used to approximate functions, derivatives and integrals and to solve certain optimization problems.Taylor's theorem is a classical result in mathematics which states that any function can be written as the sum of its Taylor series. This theorem is most useful when applied to functions which are smooth and continuous, that is, the derivatives of the function are well-defined. According to the theorem, given an arbitrary point in a domain, it is possible to interpolate the function around this point by an appropriate Taylor polynomial.One of the most important applications of Taylor's theorem is in approximating integrals. Approximatingintegrals using Taylor's theorem requires a specific kind of convergence, which guarantees that the approximation is as accurate as possible. This is especially useful for computing integrals that have complex and frequently changing boundaries. By using Taylor's theorem, the integral can be approximated with the corresponding Taylor series, which can be quickly and accurately computed. This can reduce computation time significantly.Taylor's theorem can also be used to solve solvingcertain types of optimization problems. One of the most common uses of Taylor's theorem in optimization is to find the minimum or maximum of a function without having tocalculate the derivatives. By approximating the function with a Taylor polynomial, it is possible to find the location at which the minimum or maximum occurs. This is especially useful when the derivatives of the function are difficult to calculate.Finally, Taylor's theorem can also be used to solve certain types of differential equations. The theorem can be used to approximate the solutions to these equations, which makes it easier to find the solutions. Applying Taylor's theorem in this way can help to reduce the complexity of problems that involve solving differential equations.Overall, Taylor's theorem is a powerful tool that can be used in many different areas of mathematics. Its applications range from approximating integrals to solving optimization problems to finding solutions to differential equations. The theorem provides an efficient and accurate means of approximating functions, derivatives, and integrals, making it an invaluable tool in the world of mathematics.。
Taylor Series 泰勒级数 英文版
Definition:
Let f be a function with derivatives of all orders throughout some interval containing as interior point. Then theTaylorseries generated by f at x= is:
So
3. Find theMaclaurin series of
Solution : =
4. issecond order differentiable in the interval ,and ,proof that:
Because ,
, is between and 。
References:
1."mathematical analysis" (part ii) on [M].Central Chinanormal university press, 2001134-141.
Tayseries and Taylor formula are very important in approximate calculation.In this paper,I will discuss the application of Taylor series in many ways in the calculuation of duel function.
If f is differentiable through order n+1 inan open interval I containing , the for each x in I, there exists a number between x and such that
第五节 泰勒公式与泰勒级数讲稿 、第六节函数的间接展开(泰勒级数)2013-3-26(修改)
例(1)(3) (90.5) 求级数21(3)nn x n ∞=-∑的收敛域. 解 令3t x =-,级数21nn t n∞=∑,由212lim lim 1(1)n n n n a n a n +→∞→∞==+知1t R =,因此当131x -<-<即24x <<时原级数收敛.当2x =时,原级数为21(1)nn n ∞=-∑收敛, 当4x =时,原级数为211n n∞=∑收敛.所以原级数收敛域为[2,4].(2)(92.3) 级数21(2)4nnn x n ∞=-⋅∑的收敛域为)4,0(.答 令2(2)t x =- 对于14nnn t n ∞=⋅∑, 由1141lim lim (1)44n n n n n na n a n ++→∞→∞⋅==+⋅, 于是收敛半径4t R =,则20(2)404x x ≤-<⇒<<内收敛.当0x =和4x =时,原级数都为11n n∞=∑发散,所以收敛域为(0,4).例4求幂级数1(21)nn x n ∞=+∑的收敛半径与收敛域.(中心不在原点的级数求收敛域时先作变量替换)解 令21t x =+,幂级数变形为1nn t n∞=∑,111lim lim lim 1111n t t n n n n a nn R R a n n→∞→∞→∞++====⇒=+12x R ⇒=1111022t x x <⇒+<⇒-<<,当1x =-时原级数为11(1)n n n ∞=-∑收敛,当0x =时,11n n∞=∑发散,故 原级数收敛半径12R =,收敛域为[1,0)-.注意:一般幂级数求收敛半径时作变量代换.§7.5 泰勒公式与泰勒级数教学目的:掌握泰勒公式与TaylorTh ,了解函数的Taylor级数与Taylor 展式的关系.重点:泰勒公式与泰勒定理成立的条件,理解泰勒公式的推导方法.难点: 理解泰勒公式的推导方法.教学方法:启发式讲授与指导练习相结合 教学过程:引例:近似表达函数的多项式的特性无论是函数的性态还是近似计算,多项式函数总是比较简单.为此可以考虑在一个局部范围内用多项式来近似表示一个复杂函数引例:当x 很小时,1xe x ≈+,设()x f x e =,1()1P x x =+,则11(0)(0)1,(0)(0)1f P f P ''====.用22()12x P x x =++表示 212x x e x ≈++在0x =处值更为接近.猜想将1()P x 换成()n P x 则在0x x =处两函数有直到n 阶相同的导数,其在0x x =处接近的程度更高,即212!n xx x e x n ≈++++ .为用多项式表示更复杂的函数:设有函数()f x 在0x x =的某一邻域内有直到1n +阶的导数,令()f x ≈0100()()()n n n P x a a x x a x x =+-++- , 再令 1()()n f x D I +∈,0(,)x I a b ∈=, 若 ()()00()()k k n f x P x =,0,1,,k n = .((0)(0)00()()n f x P x =表示0k =的函数值相等)则 ()01()!k k a f x k =(0,1,,k n = ),于是()f x ≈0100()()()n n n P x a a x x a x x =+-++- .证明:因0100()()()n n n P x a a x x a x x =+-++- ,10()()(1)n P x a x x O '=+-,20()2!()(1)n P x a x x O ''=+-…… ,()0()!()(1)k n k P x k a x x O =+- …… ,()()!n n n P x n a =,那么 ()()00()()!k k n k f x P x k a ==,所以 ()01()!k k a f x k =, 0,1,,k n = .一、泰勒(Taylor )公式在讲第三章微分的应用时我们导出了近似公式000()()()()f x f x f x x x '≈+-( 当0x x -很小时)从几何上看,这是在点0x 附近用切线的一段近似地代替曲线弧.在函数改变量的表达式0000()()()()()f x f x f x x x o x x '=+-+-中 略去了一个关于(0x x -)的高阶无穷小量(0x x →时).但公式000()()()()f x f x f x x x '≈+-在实际计算中的精度不高,其误差为1000()()()()()R x f x f x f x x x '=---,可以推出()2100()()(),,2!f R x x x x x ξξ''=-∈.如果需要精度更高些,可将(0x x -)的高阶无穷小分离成两部分()220200()()()o x x a x x o x x -=-+-(0x x →时).保留与20()x x -同阶的无穷小量,略去20()x x -的高阶无穷小量,此时有200020()()()()()f x f x f x x x a x x '≈+-+-,以此类推,为达到一定精确度的要求,可考虑用n 次多项式()P x 近似表示()f x ,当0x x -很小时,将多项式()P x 写成以(0x x -)的方幂展开的形式2010200()()()()n n P x a a x x a x x a x x =+-+-++- ,其中012,,,a a a 是待定系数.我们知道()P x 具有任意阶的连续导数,将()P x 的多项式两边求一阶到n 阶导数,并令0x x =可得000102(),(),()2!,,P x a P x a P x a '''===()0()!n n P x n a = 于是()P x 可以写成200000()()()()()()2!P x P x P x P x x x x x '''=+-+-+()00()()!n n P x x x n +-若函数()f x 在0x x =的某一邻域内一阶到n 阶的导数都存在,可以做出一个n 次多项式200000()()()()()()2!n P x P x P x P x x x x x '''=+-+-+()00()()!n n P x x x n +- ()n P x 不一定等于()f x ,但它可以近似表示()f x ,它的近似程度可以由误差()()()n n R x f x P x =-来确定. 设10()()(1)!n n kR x x x n +=-+,如果能确定k 的值,则()n R x 就确定了.【定理7.10】(泰勒公式)设()f x 在含有0x 的区间(,)I a b =内有直到1n +阶的连续导数,则(,)x a b ∀∈,()f x 可以按(0x x -)的方幂展开为()()()n n f x P x R x =+000()()()f x f x x x '=+-+()001()()()!n n n f x x x R x n +-+. 此式称为按0x x -的幂展开n 阶泰勒公式.其中(1)10()()()(1)!n n n f R x x x n ξ++=-+ 称为拉格朗日型余项,ξ介于0x 与x 之间. 证明:不妨设0x x >.令()()()n n R t f t P t =-,10)()(+-=n n x t t G ,由条件知:(连续1n +次使用柯西中值定理可以证明)()()0(),()[,]k k n n R t G t C x x ∈,()()0(),()(,)k k nn R t G t D x x ∈,显然 ()()00()()0k k n n R x G x ==, 0,1,,k n = .那么011001()()()()()()()()n n n nn n n nR x R x R x R x x G x G x G ξξ+'-=='-- 1010()()()()n n nn R R x G G x ξξ''-=''-22()()n n R G ξξ''==''(1)(1)1(1)1()()()(1)!n n n n n n n R f G n ξξξ+++++==+, 其中 0121n x x ξξξξ+<=<<<< ,所以(1)10()()()(1)!n n n f R x x x n ξ++=-+, ξ介于0x 与x 之间. 另证:因为()f x 在含有0x 的区间(,)I a b =内有直到1n +阶的连续导数,所以对于0(,)x a b ∈,可将()f x 写成200000()()()()()()2!f x f x f x f x x x x x '''=+-+-+ ()10001()()()!(1)!n n n kf x x x x x n n ++-+-+为求出k 的值,引进辅助函数2()()()()()()()2!f t t f x f t f t x t x t ϕ'''=------()11()()()!(1)!n n n k f t x t x t n n +----+显然 0()()0x x ϕϕ==,()t ϕ在区间0[,]x x 上连续(设0x x >),在区间0(,)x x 内可导,由罗尔中值定理可知,至少存在一点0(,)x x ξ∈,使得()0ϕξ'=,因为 ()()()()[()()()]t f t f t x t f t x t f t ϕ''''''=------2()[()()()]2!f t x t f t x t '''''---- (4)32()()[()()]3!2!f t f t x t x t '''-----(1)()(1)()()[()()]()!(1)!!n n nn n f t f t k x t x t x t n n n +-----+--化简整理得 (1)()()[()]!nn x t t k f t n ϕ+-'=- 所以(1)()[()]0!nn x k f n ξξ+--=,而 ()0n x ξ-≠ 由 (1)(1)()0()n n k f k f ξξ++-=⇒=,于是10)1()()!1()()(++-+=n n n x x n f x R ξ,ξ介于0x 与x 之间.在公式中当00x =时,公式可化为麦克劳林公式2(0)()(0)(0)2!f f x f f x x '''=+++()(0)()!n n n f x R x n ++其中 (1)1()()(1)!n n n f R x x n ξ++=+或令,01x ξθθ=<<,则 (1)1()()(1)!n n n f x R x x n θ++=+例1 求()xf x e =的n 阶麦克劳林公式.解 因()()k x f x e =,()0(0)1k f e ==, 其中 0,1,,1k n n =+ ,那么()(0)(0)x e f x f f x '==++(1)()11()(0)!(1)!n n n n f x f x x n n θ+++++ 211112!!(1)!xn n e x x x x n n θ+=++++++ ,(01θ<<).例2 求()sin f x x =的麦克劳林公式.解 因()()sin()2n n f x x π=+, ()(0)sin()2n n f π=.有 (0)0,(0)1,(0)0,(0)1,f f f f ''''''====- (2)(0)0k f =,(21)(0)(1)k k f +=- ,0,1,2k = , 那么sin ()x f x =(1)()11()(0)(0)(0)!(1)!n n nn f x f f x f x xn n θ++'=+++++ 3521121(1)()3!5!(21)!k k k x x x x R x k -+-=-+-+-+- ,(或2()k R x 都可以)其中:221sin[(2)]2()(2)!k k x k R x x k πθ-+=,01θ<<. (或 212sin[(21)]2()(21)!k k x k R x x k πθ+++=+,01θ<<)特别地:1k =时,sin x x ≈, 32||||3!x R ≤;2k =时,3sin 3!x x x ≈-, 54||||5!x R ≤;3k =时,35sin 3!5!x x x x ≈-+, 76||||7!x R ≤. 例3 按(4)x -的乘幂展开多项式432()523f x x x x x =-+-.解 (4)60,f =-324(4)(41523)|21,x f x x x ='=-+-= 24(4)(12302)|74,x f x x =''=-+=4(4)(2430)|66,x f x =''=-=(5)(4)24,()0,()0n f f x R x '''===,所以432()(4)11(4)37(4)21(4)60f x x x x x =-+-+-+--.二、泰勒级数1.通过前面的学习我们知道,级数在其收敛域内一定有和函数. 由泰勒公式的学习知道,我们可以用多项式近似表示函数.现在我们想知道函数是否一定可以展开为幂级数,需不需要附加条件?2.问题:已知函数有 01,(1)1n n x x x ∞==<-∑收敛域11ln(1)(1)(11)n n n x x x n∞-=+=--<≤∑.问:(1) 对于一般的函数()f x 是否也有00()()n n n f x a x x ∞==-∑?(2) 如果能展开,项的系数n a 如何确定?(3) 展开式是否唯一?(4) 在什么条件下函数才能展开成幂级数?3.【定理】(Taylor Th ) 设()f x 在0(,)U x δ内具有任意阶导数,且lim ()0n n R x →∞=,则在0(,)U x δ内有()000()()()!n n n f x f x x x n ∞==-∑.其中()n R x 为()f x 的拉格朗日型余项(1)10()()()(1)!n n n f R x x x n ξ++=-+.证明 由于 ()000()()()()()()!n nn n n n n f x f x x x R x P x R x n ==-+=+∑. 所以等式两边取极限()000()()()lim ()!n n n n n f x f x x x P x n ∞→∞==-=∑⇔lim ()lim[()()]0n n n n R x f x P x →∞→∞=-=,),(0δx U x ∈.4.函数()f x 在点0x x =有泰勒展式⇔()f x 在0(,)U x δ有任意阶导数且lim ()0n n R x →∞=.注意:1)函数在点处可以展开为Taylor 级数时,其展式是唯一的. 因为泰勒系数()0()(0,1,2,)!n f x n n = 是唯一的. 2)()000()()!n n n f x x x n ∞=-∑为 ()f x 在0x x =点的 Taylor 级数,等式0()()nnn f x a x x ∞==-∑在lim ()0n n R x →∞=时成立.5.泰勒级数与麦克劳林级数设()f x 在0x x =点具有任意阶导数,则称(1) ()000()()!n n n f x x x n ∞=-∑为()f x 在点0x 的泰勒级数,记作 ()000()()~()!n n n f x f x x x n ∞=-∑. (2) ()0(0)!n nn f x n ∞=∑称为()f x 的麦克劳林级数, 记作 ()0(0)()~!n nn f f x x n ∞=∑. 0(0)x =注意问题: ()f x 在0x x =点具有任意阶导数,那么 级数()000()()!n n n f x x x n ∞=-∑在收敛区间内是否收敛于()f x ?例: 函数21,0,()0,0.x e x f x x -⎧⎪≠=⎨⎪=⎩在0x =点处任意可导,且()(0)0,0,1,n f n == ,于是()~f x ()0(0)!n nn f x n ∞==∑000n n x ∞=⋅=∑,x -∞<<+∞ 显然()f x ≠()0(0)0!n nn f x n ∞==∑, 0x ≠.结论:当级数()000()()!n n n f x x x n ∞=-∑收敛于()f x 时,即lim ()0n n R x →∞=时有泰勒展式.应用举例:例4 求函数在点0x =处的泰勒级数: (1)()xf x e =, (2)()sin f x x =提示:0,!nxn x e x n ∞==-∞<<+∞∑210sin (1),(21)!n nn x x x n +∞==--∞<<+∞+∑小结:1.函数()f x 在点0x x =的泰勒公式为()000()()()()!k nk n k f x f x x x R x k ==-+∑其中余项为(1)10()()()(1)!n n n f R x x x n ξ++=-+,ξ介于0x 与x 之间.公式成立的条件是:()f x 在点0x x =的邻域内有直到1n +阶的导数.2. 函数()f x 在点0x x =的泰勒展式为()000()()()!n nn f x f x x x n ∞==-∑ ,其系数()0()!n n f x a n =为泰勒系数.当00x =时,()f x 的上述展式为麦克劳林展式.注意:函数在一点的泰勒展式唯一.泰勒定理成立的条件是:()f x 在点0x x =邻域内的各阶导数存在且lim ()0n n R x →∞=.3.在近似计算中先要写出函数的级数表示式,再取n 的特殊值即可得到所要近似值.课后记:存在问题:不能区分泰勒公式与泰勒级数.§7.6 某些初等函数的幂级数展开式教学目的:熟练掌握Taylor 公式、TaylorTh 展式;能灵活运用导出公式间接求出函数的泰勒展式.重难点:能灵活运用导出公式间接求出所给函数的泰勒展式以及麦克劳林展式.教学方法:启发式讲授与指导练习相结合 教学过程:一、某些初等函数的幂级数展开式由泰勒定理的学习可知一个函数()f x 对区间[,]a b 内一个特定值0x ,是否可以展开为幂级数,取决于它在0x x =处的各阶导数是否存在,以及当n →∞时,余项()n R x 是否趋于0.1.直接展开法(利用泰勒级数与麦克劳林级数展开函数)将函数()f x 展成麦克劳林级数步骤:(1) 求()()n f x ,进而求出()(0)n f ;如果()f x 在00x =的某一阶导数不存在,则()f x 不能在00x =展成幂级数.(2)写出()f x 的麦克劳林级数()0(0)()~!n nn f f x x n ∞=∑,并求出级数的收敛半径R 、收敛域;(3) 讨论lim ()0n n R x →∞=或()(),n f x M ≤ ||x R <,(4) 在收敛区间I 上有 ()0(0)()!n nn f f x x n ∞==∑, x I ∈.例1 将()xf x e =展开成x 的幂级数.解:(1) 00x =,(2) 由于()()n x f x e =,所以()(0)1n f =, 1,2,n = ;(0)(0,1,2,)!n n f a n n ==()00(0)!!n nn n n f x x n n ∞∞===∑∑, 由于收敛半径1(1)!limlim lim(1)!n n n n n a n R n a n →∞→∞→∞++===+=+∞; (3) ∴201!2!!n n xn x x x e x n n ∞===+++++∑ , x -∞<<+∞.近似计算: 1xe x ≈+;212xx e x ≈++;23126xx x e x ≈+++.例2 将()sin f x x =展开成x 的幂级数. 解 (1) ()()sin()2n fx x n π=+⋅, 0,1,2,n = ;()(0)n f 依次循环取0,1,0,1,0,1,0,1,(0,1,2,)n --=即(21)(0)(1)n n f +=-,(2)(0)0n f = (0,1,2,)n = ;(2)()2100(0)(1)!(21)!n n n nn n f x x n n +∞∞===-+∑∑ 【或2111(1)(21)!n n n x n -∞-==--∑】3521(1)3!5!(21)!n nx x x x n +=-++-++ ,而211(21)!lim lim (23)!n n n n u n x R u n +→∞→∞+==+ 21lim0(23)(22)n x n n →∞==++;所以收敛域为 x -∞<<+∞.(3) 所以2121110sin (1)(1)(21)!(21)!n n n nn n x x x n n -∞∞-==⎡⎤=--⎢⎥-⎣⎦∑∑+=+ 35211sin (1)3!5!(21)!n n x x x x x n --=-++-+- ,x -∞<<+∞.例3 将函数 ()(1)f x x α=+展开成麦克劳林级数,其中α是任意不为零的常数. 分析:因为 1()(1)f x x αα-'=+,2()(1)(1)f x x ααα-''=-+()()(1)(1)(1)n n f x n x αααα-=--++所以 (0)1,(0),(0)(1),f f f ααα'''===-()(0)(1)(1)n f n ααα=--+ 得麦克劳林级数公式:1(1)(1)(1)!n n n x x n αααα∞=--++=∑,收敛域为 1x <(结果为二项式级数)当1x =±时,级数是否收敛于()1x α+取决于α的取值.可以证明:当1α≤-时,收敛域为()1,1-;当10α-<< 时,收敛域为(1,1]-;当0α>时,收敛域为[]1,1-. 取111,,,22ααα=-==- 等不同的值可以得到相应的公式.001()(1)1nn n n n x x x ∞∞===-=-+∑∑,(11x -<<). 2311111224246x x x x +=+-+⋅⋅⋅41[1,1]2468x x -+∈-⋅⋅⋅ 23111313512242461x x x x⋅⋅⋅=-+-⋅⋅⋅+41357(1,1]2468x x ⋅⋅⋅+-∈-⋅⋅⋅ 011n n x x ∞==-∑,(11x -<<).可以由无穷递缩等比数列求和公式得到.特别地,当α是正整数n 时,可以看出含有nx 项以后的各项的系数都为零.从而得到二项式公式21(1)(1)12!n n n n n x nx x nx x --+=+++++ . 2.间接法根据函数的泰勒展式的唯一性,利用常见展开式如sin x ,xe ,11x-,(1)n x +的公式,通过变量代换、四则运算、恒等变形、逐项求导、逐项积分等方法,求函数的幂级数(泰勒)展开式.例4 (1) 将()cos f x x =展开成x 的幂级数. 解:已知2121110sin (1)(1)(21)!(21)!n n n nn n x x x n n -+∞∞-===-=--+∑∑,x ∈R . 那么210cos (sin )(1)(21)!n nn x x x n +∞='⎡⎤'==-⎢⎥+⎣⎦∑20(1)(2)!n n n x n ∞==-∑,x <+∞ (2) 将21()1f x x =+展开成x 的幂级数.(注意收敛区间的间接求法)解:已知11n n x x ∞==-∑, 11x -<<. 那么 22220011()(1)11()nn n n n x x x x ∞∞====-=-+--∑∑, 11x -<<.例5 (1)将()ln(1)f x x =+展开成x 的幂级数.解:已知001[ln(1)]()(1)1nn n n n x x x x ∞∞=='+==-=-+∑∑, ||1x <.那么ln(1)[ln(1)]xx t dt '+=+⎰1000(1)(1)1n xnnnn n x t dt n +∞∞===-=-+∑∑⎰,||1x <. 又因为 1x =时,级数 01(1)1nn n ∞=-+∑收敛, ln(1)x +在1x =连续.1x =-时,级数 011n n ∞=-+∑发散, 于是1ln(1)(1)1n nn x x n +∞=+=-+∑ 231(1)231n n x x x x n +=-++-++ , 其中 收敛域为 11x -<≤.(2)将()arctan f x x =展开成x 的幂级数.解 221()()1n n f x x x ∞='==-+∑ 20(1),(1,1)n n n x x ∞==-∈-∑,(0)0f =21arctan 1xx dt t=+⎰212000(1)(1),(1,1)21n xn nnn n x t dt x n +∞∞===-=-∈-+∑∑⎰当2100(1)1,(1)2121n n n n n x x n n +∞∞==-=±-=±++∑∑均收敛, 故 21arctan (1),[1,1]21n nn x x x n +∞==-∈-+∑.注意:对于不需要通过积分与求导就可以的得到的级数,其收敛域可以直接由原收敛域间接求出,但对于要积分或求导才能得到的级数,端点要单独考察一下敛散性. 提问:用间接法将下列函数展开为为x 的幂级数,并确定收敛域: (1)()e 2x f x -=解 因为e ()!xnn 01xx n ∞==-∞<<+∞∑,所以有e 222001()(1)!!n x n n n n x x n n ∞∞-===-=-∑∑, 并由2x -∞<-<+∞得)(x f 的收敛域为(,)-∞+∞.同理可得e 3001()(1)!33!x n n nnn n x x n n ∞∞-===-=-⋅∑∑,(,)x ∈-∞+∞. (2)2()cos f x x =解 因为)()!2()1(cos 02+∞<<-∞-=∑∞=x n x x n nn,所以有21cos (1cos 2)2x x =+220111(2)(2)(1)1(1)22(2)!2(2)!n n n n n n x x n n ∞∞===+-=+-∑∑, 并由+∞<<∞-x 2得)(x f 的收敛域为),(+∞-∞. 同理可得21sin (1cos 2)2x x =-2011(2)(1)22(2)!nn n x n ∞==--∑211(2)(1)()2(2)!nn n x x n ∞-==--∞<<+∞∑,(3)e 3()x f x x -=解 因为)(!1e 0+∞<<-∞=∑∞=x x n n n x,所以有e 333001()(1)()!!n x n n n n x x x x x n n +∞∞-===-=--∞<<+∞∑∑. (4)1()3f x x=-解 由 ()2n 111t t t 1t 11t-=+++++-<<- ,有1113313x x =⋅--211[1()()]3333n x x x -=⋅+++++ 103n n n x ∞+==∑ 又由13<x得其收敛区间为)3,3(-.收敛域为 [3,3)-解211()()23(3)(1)431x x x f x x x x x x x ===----+-+并由 2111(11)1n t t t t t-=+++++-<<-,知 1001111()3333313nn n n n x x x x ∞∞+===-⋅=-=---∑∑(其中33)x -<<和0001()()(1)(11)1n nn n n n n x x x x x ∞∞∞====-=-=--<<+∑∑∑,所以2100()[(1)]2343nn n n n n x x x f x x x x ∞∞+====-----∑∑1111[(1)],1143n n n n x x ∞-==-+--<<∑.(6)将1()x d e dx x-展成x 的幂级数. 解:因为 )(!1e 0+∞<<-∞=∑∞=x x n n n x,0111!()n x n x d e d n dx x dx x ∞=⎡⎤-⎢⎥-⎢⎥=⎢⎥⎢⎥⎣⎦∑ 121111()!!n n n n d n x x x dx n n ∞∞--==⎡⎤-==-∞<<+∞⎢⎥⎣⎦∑∑. 例6(1) (07.3.10)将函数21()34f x x x =-+展开为1x -的幂级数,并指出收敛区间.解: 21111()[]34541f x x x x x ==--+-+111[]53(1)2(1)x x =--+-+-11111[]115321132x x =-⋅-⋅---+ 001(1)1(1)(1)153102n n n n nn n x x ∞∞==--=--∑∑0123(1)[](1)3032nn n n n x ∞=-=+-∑ 由12x -<得收敛区间为()1,3-.(2) 将()sin f x x =展开成()4x π-的幂级数.解:由于 ()sin ()44f x x ππ⎡⎤=+-⎢⎥⎣⎦ sin cos()cos sin()4444x x ππππ=-+- 又已知21sin (1)(21)!n n n x x n +∞==-+∑, x -∞<<+∞, 20cos (1)(2)!n n n x x n ∞==-∑, x -∞<<+∞. 那么 1sin cos()sin()442x x x ππ⎡⎤=-+-⎢⎥⎣⎦ 2210()()144(1)(2)!(21)!2n n n n x x n n ππ+∞=⎡⎤--⎢⎥=-+⎢⎥+⎢⎥⎣⎦∑, 收敛域 x -∞<<+∞.(3)将21()f x x =展开为2x -的幂级数,并确定收敛区间.解 1111222212x x x ==⋅--++ 10(2)(1),222nn n n x x ∞+=-=--<∑ 1121111(2)()()(1),042n n n n n x f x x x x -∞++=-'==-=-<<∑类似可求211()()(1)1f x x x '==-- 11,11n n nx x ∞-==-<<∑ 小结:1.函数()f x 在点0x x =的泰勒展式为n n n x x n x f x f )(!)()(000)(-=∑∞= ,其系数()0()!n n f x a n = 为泰勒系数.当00x =时,()f x 的上述展式为麦克劳林展式.注意:函数在一点的泰勒展式唯一.2.利用公式中的已知收敛域,间接地求所求级数的收敛域比较方便.3.常用于间接展开的公式有1)01,11n n x x x ∞==<-∑ 2)21sin (1),(21)!n nn x x x n +∞==-<+∞+∑ 3)20cos (1),(2)!n n n x x x n ∞==-<+∞∑ 4)0,!nx n x e x n ∞==<+∞∑注意:有限个级数的代数和的收敛域应为各个收敛域的公共部分.课后记:存在问题:1.间接展开时不能灵活运用已知公式和级数的性质去正确写出套用公式所需的表示式.2.忽略了级数和的收敛域应为各个收敛域的公共部分.。
- 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
Taylor's Formula and the Study of Extrema1. Taylor's Formula for MappingsTheorem 1. If a mapping Y U f →: from a neighborhood ()x U U = of a point x in a normed space X into a normed space Y has derivatives up to order n -1 inclusive in U and has an n-th order derivative()()x f nat the point x, then()()()()()⎪⎭⎫ ⎝⎛++++=+n n n h o h x f n h x f x f h x f !1,Λ (1)as 0→h .Equality (1) is one of the varieties of Taylor's formula, written here for rather general classes of mappings.Proof. We prove Taylor's formula by induction. For1=nit is true by definition of ()x f ,.Assume formula (1) is true for some N n ∈-1.Then by the mean-value theorem, formula (12) of Sect. 10.5, and the induction hypothesis, we obtain.()()()()()()()()()()()()()⎪⎭⎫ ⎝⎛=⎪⎭⎫ ⎝⎛=⎪⎪⎭⎫ ⎝⎛-+++-+≤⎪⎭⎫ ⎝⎛+++-+--<<nn n n n n h o h h o h h x f n h x f x f h x f h x f n h f x f h x f 11,,,,10!11sup !1x θθθθθΛΛ,as 0→h .We shall not take the time here to discuss other versions of Taylor's formula, which are sometimes quite useful. They were discussed earlier in detail for numerical functions. At this point we leave it to the reader to derive them (see, for example, Problem 1 below). 2. Methods of Studying Interior ExtremaUsing Taylor's formula, we shall exhibit necessary conditions and also sufficient conditions for an interior local extremum of real-valued functions defined on an open subset of a normed space. As we shall see, these conditions are analogous to the differential conditions already known to us for an extremum of a real-valued function of a real variable.Theorem 2. Let R U f →: be a real-valued function defined on an open set U in a normed space X and having continuous derivatives up to order 11≥-k inclusive in a neighborhood of a point U x ∈and a derivative()()x f kof order k at the point x itself.If()()()0,,01,==-x f x f k Λand()()0≠x f k , then for x to be an extremum of the function f it is:necessary that k be even and that the form ()()k k h x fbe semidefinite,andsufficient that the values of the form()()k k h x fon the unit sphere 1=h be bounded awayfrom zero; moreover, x is a local minimum if the inequalities()()0>≥δk k h x f ,hold on that sphere, and a local maximum if()()0<≤δk k h x f ,Proof. For the proof we consider the Taylor expansion (1) of f in a neighborhood of x. The assumptions enable us to write()()()()()k k k h h h x f k x f h x f α+=-+!1where ()h α is a real-valued function, and ()0→h α as 0→h . We first prove the necessary conditions. Since ()()0≠x f k , there exists a vector00≠h on which ()()00≠kk h x f . Then for values of thereal parameter t sufficiently close to zero,()()()()()()kk k th th th x f k x f th x f 0000!1α+=-+()()()k k k k t h th h x f k ⎪⎭⎫ ⎝⎛+=000!1αand the expression in the outer parentheses has the same sign as()()kk h x f 0.For x to be an extremum it is necessary for the left-hand side (and hence also the right-handside) of this last equality to be of constant sign when t changes sign. But this is possible only if k is even.This reasoning shows that if x is an extremum, then the sign of the difference ()()x f th x f -+0 is the same as that of ()()kk h x f 0for sufficiently small t; hence in that case there cannot be twovectors0h , 1hat which the form ()()x f kassumes values with opposite signs.We now turn to the proof of the sufficiency conditions. For definiteness we consider the case when()()0>≥δk k h x ffor 1=h . Then()()()()()kk k h h h x f k x f h x f α+=-+!1()()()k k k h h h h x f k ⎪⎪⎪⎭⎫ ⎝⎛+⎪⎪⎭⎫ ⎝⎛=α!1 ()kh h k ⎪⎭⎫ ⎝⎛+≥αδ!1and, since ()0→h α as 0→h , the last term in this inequality is positive for all vectors0≠h sufficiently close to zero. Thus, for all such vectors h,()()0>-+x f h x f ,that is, x is a strict local minimum.The sufficient condition for a strict local maximum is verified similiarly.Remark 1. If the space X is finite-dimensional, the unit sphere ()1;x S with center at X x ∈, being a closed bounded subset of X, is compact. Then the continuous function()()()()kk i i i i k k h h x f h x f ⋅⋅∂=ΛΛ11 (a k-form) has both a maximal and a minimal value on ()1;x S . Ifthese values are of opposite sign, then f does not have an extremum at x. If they are both of the same sign, then, as was shown in Theorem 2, there is an extremum. In the latter case, a sufficient condition for an extremum can obviously be stated as the equivalent requirement that the form()()k k h x fbe either positive- or negative-definite.It was this form of the condition that we encountered in studying realvalued functions on n R .Remark 2. As we have seen in the example of functions R R f n →:, the semi-definitenessof the form()()k k h x fexhibited in the necessary conditions for an extremum is not a sufficientcriterion for an extremum.Remark 3. In practice, when studying extrema of differentiable functions one normally uses only the first or second differentials. If the uniqueness and type of extremum are obvious from the meaning of the problem being studied, one can restrict attention to the first differential when seeking an extremum, simply finding the point x where ()0,=x f 3. Some Examples Example 1. Let()()RR C L ;31∈ and()[]()R b a C f ;,1∈.In other words,()()321321,,,,u u u L u u u αis a continuously differentiable real-valued function defined in 3Rand ()x f x α a smoothreal-valued function defined on the closed interval []R b a ⊂,. Consider the function()[]()R R b a C F →;,:1(2)defined by the relation()[]()()f F R b a C f α;,1∈ ()()()R dx x f x f x L ba∈=⎰,,, (3)Thus, (2) is a real-valued functional defined on the set of functions()[]()R b a C ;,1.The basic variational principles connected with motion are known in physics and mechanics. According to these principles, the actual motions are distinguished among all the conceivable motions in that they proceed along trajectories along which certain functionals have an extremum. Questions connected with the extrema of functionals are central in optimalcontrol theory. Thus, finding and studying the extrema of functionals is a problemof intrinsic importance, and the theory associated with it is the subject of a large area of analysis - the calculus of variations. We have already done a few things to make the transition from the analysis of the extrema of numerical functions to the problem of finding and studying extrema of functionals seem natural to the reader. However, we shall not go deeply into the special problems of variational calculus, but rather use the example of the functional (3) to illustrate only the general ideas of differentiation and study of local extrema considered above.We shall show that the functional (3) is a differentiate mapping and find its differential. We remark that the function (3) can be regarded as the composition of the mappings()()()()()x f x f x L x f F ,1,,= (4)defined by the formula()[]()[]()R b a C R b a C F ;,;,:11→(5)followed by the mapping[]()()()R dx x g g F R b a C g ba∈=∈⎰2;,α (6)By properties of the integral, the mapping 2Fis obviously linear and continuous, so thatits differentiability is clear. We shall show that the mapping1Fis also differentiable, and that()()()()()()()()()()x h x f x f x L x h x f x f x L x h f F ,,3,2,1.,,,∂+∂=(7)for()[]()R b a C h ;,1∈.Indeed, by the corollary to the mean-value theorem, we can write in the present case()()()ii iu u u L u u u L u u u L ∆∂--∆+∆+∆+∑=32131321332211,,,,,,()()()()()()∆⋅∂-∆+∂∂-∆+∂∂-∆+∂≤<<u L u L u L u L u L u L 3312211110sup θθθθ()()ii i i u L u u L i ∆⋅∂-+∂≤=≤≤=3,2,110max max 33,2,1θθ (8)where ()321,,u u u u = and ()321,,∆∆∆=∆.If we now recall that the norm ()1c fof the function f in()[]()R b a C ;,1is⎭⎬⎫⎩⎨⎧c c f f ,,max (wherecfis the maximum absolute value of the function on the closed interval []b a ,), then,setting x u =1,()x f u =2, ()x f u ,3=, 01=∆, ()x h =∆2, and ()x h ,3=∆, we obtain from inequality (8),taking account of the uniform continuity of the functions ()3,2,1,,,321=∂i u u u L i , on boundedsubsets of3R , that()()()()()()()()()()()()()()()()x h x f x f x L x h x f x f x L x f x f x L x h x f x h x f x L bx ,,3,2,,,0,,,,,,,,max ∂-∂--++≤≤()()1c h o = as()01→c hBut this means that Eq. (7) holds.By the chain rule for differentiating a composite function, we now conclude that the functional (3) is indeed differentiable, and()()()()()()()()()()⎰∂+∂=b adx x h x f x f x L x h x f x f x L h f F ,,3,2,,,,, (9)We often consider the restriction of the functional (3) to the affine space consisting of the functions()[]()R b a C f ;,1∈that assume fixed values ()A a f =, ()B b f = at the endpoints of theclosed interval []b a ,. In this case, the functions h in the tangent space ()1f TC , must have the value zero at the endpoints of the closed interval []b a ,. Taking this fact into account, we may integrate by parts in (9) and bring it into the form()()()()()()()()⎰⎪⎭⎫⎝⎛∂-∂=b a dx x h x f x f x L dx d x f x f x L h f F ,3,2,,,,,(10)of course under the assumption that L and f belong to the corresponding class ()2C .In particular, if f is an extremum (extremal) of such a functional, then by Theorem 2 we have()0,=h f Ffor every function()[]()R b a C h ;,1∈such that ()()0==b h a h . From this and relation (10)one can easily conclude (see Problem 3 below) that the function f must satisfy the equation()()()()()()0,,,,,3,2=∂-∂x f x f x L dxdx f x f x L (11)This is a frequently-encountered form of the equation known in the calculus of variations as the Euler-Lagrange equation.Let us now consider some specific examples. Example 2. The shortest-path problemAmong all the curves in a plane joining two fixed points, find the curve that has minimal length.The answer in this case is obvious, and it rather serves as a check on the formal computations we will be doing later.We shall assume that a fixed Cartesian coordinate system has been chosen in the plane, in which the two points are, for example, ()0,0 and ()0,1 . We confine ourselves to just the curves that are the graphs of functions()[]()R C f ;1,01∈assuming the value zero at both ends ofthe closed interval []1,0 . The length of such a curve()()()⎰+=12,1dx x f f F (12)depends on the function f and is a functional of the type considered in Example 1. In this case the function L has the form()()233211,,u u u u L +=and therefore the necessary condition (11) for an extremal here reduces to the equation()()()012,,=⎪⎪⎪⎭⎫⎝⎛+x f x f dx dfrom which it follows that()()()常数≡+x fx f 2,,1 (13)on the closed interval []1,0 Since the function21uu + is not constant on any interval, Eq. (13) is possible only if()≡x f ,const on []b a ,. Thus a smooth extremal of this problem must be a linear function whosegraph passes through the points ()0,0 and ()0,1. It follows that ()0≡x f , and we arrive at the closed interval of the line joining the two given points. Example 3. The brachistochrone problemThe classical brachistochrone problem, posed by Johann Bernoulli I in 1696, was to find the shape of a track along which a point mass would pass from a prescribed point 0Ptoanother fixed point1Pat a lower level under the action of gravity in the shortest time.We neglect friction, of course. In addition, we shall assume that the trivial case in whichboth points lie on the same vertical line is excluded. In the vertical plane passing through the points 0Pand1Pwe introduce a rectangularcoordinate system such that 0Pis at the origin, the x-axis is directed vertically downward,and the point1Phas positive coordinates ()11,y x .We shall find the shape of the track amongthe graphs of smooth functions defined on the closed interval []1,0x and satisfying the condition ()00=f ,()11y x f =. At the moment we shall not take time to discuss this by no means uncontroversial assumption (see Problem 4 below). If the particle began its descent from the point0Pwith zero velocity, the law of variationof its velocity in these coordinates can be written asgxv 2= (14)Recalling that the differential of the arc length is computed by the formula()()()()dx x f dy dx ds 2,221+=+=(15)we find the time of descent()()()⎰+=12,121x dx xx f gf F (16)along the trajectory defined by the graph of the function ()x f y =on the closed interval []1,0x .For the functional (16)()()1233211,,u u uu u L +=,and therefore the condition (11) for an extremum reduces in this case to the equation()()()012,,=⎪⎪⎪⎪⎪⎭⎫⎝⎛⎪⎭⎫ ⎝⎛+x f x x f dx d , from which it follows that()()()xc x fx f =+2,,1 (17)where c is a nonzero constant, since the points are not both on the same vertical line. Taking account of (15), we can rewrite (17) in the formx c dsdy= (18)However, from the geometric point of viewϕcos =ds dx ,ϕsin =dsdy (19)where ϕ is the angle between the tangent to the trajectory and the positive x-axis.By comparing Eq. (18) with the second equation in (19), we findϕ22sin 1cx =(20)But it follows from (19) and (20) thatdx dy d dy =ϕ,2222sin 2sin c c d d tg d dx tg d dx ϕϕϕϕϕϕϕ=⎪⎪⎭⎫ ⎝⎛==,from which we find()b c y +-=ϕϕ2sin 2212(21)Settinga c =221 and t =ϕ2, we write relations (20) and (21) as()()bt t a y t a x +-=-=sin cos 1 (22)Since 0≠a , it follows that 0=x only for πk t 2=,Z k ∈. It follows from the form of thefunction (22) that we may assume without loss of generality that the parameter value 0=t corresponds to the point()0,00=P . In this case Eq. (21) implies 0=b , and we arrive at thesimpler form()()t t a y t a x sin cos 1-=-= (23)for the parametric definition of this curve.Thus the brachistochrone is a cycloid having a cusp at the initial point0Pwhere thetangent is vertical. The constant a, which is a scaling coefficient, must be chosen so that the curve (23) also passes through the point1P .Such a choice, as one can see by sketching thecurve (23), is by no means always unique, and this shows that the necessary condition (11) foran extremum is in general not sufficient. However, from physical considerations it is clear which of the possible values of the parameter a should be preferred (and this, of course, can be confirmed by direct computation).泰勒公式和极值的研究1.映射的泰勒公式定理1 如果从赋范空间X 的点x 的邻域()x U U =到赋范空间Y 的映射Y U f→:在U中有直到n-1阶(包括n-1在内)的导数,而在点x 处有n 阶导数。