• 沒有找到結果。

講義請點此

N/A
N/A
Protected

Academic year: 2021

Share "講義請點此"

Copied!
44
0
0

加載中.... (立即查看全文)

全文

(1)

11.1 Sequences goo.gl/wjLEY9 1

Chapter 11

Infinite Sequences and Series

11.1

Sequences (page 694)

v09mtwnVL7A 這一章的終極目標 是泰勒展開式的理 論, 為徹底了解泰 勒展開式, 必須先 從數列開始, 然後 討論級數, 最後介 紹函數項級數。 這章用到非常多的 數學論述與邏輯推 演, 必須反覆思考 以逐漸體會。 Definition 1 (page 694).

(1) A sequence (數列) is a list of numbers written in a definite order: a1, a2, a3, a4, . . . , an, . . . .

The number a1 is called the first term (第一項), a2 is the second term (第二項), and in

general an is the n-th term (第 n項).

(2) An infinite sequence (無窮數列) is a sequence that each term anhas a successor an+1.

(3) The sequence {a1, a2, a3, . . .} is also denoted by {an} or {an}∞n=1.



微積分課程感興趣的是無窮數列, 若將無窮數列依序寫下時, 在一般項後面還會再加上 「點點點」。 認識數列的幾種表 示法。 微積分課程 中主要探討的數列 有兩種類型, 一種 是有明確表達式的 數列, 另一種是利 用遞迴式定義出的 數列。

Example 2(page 694). Some sequences can be defined by giving a formula for the n-th term. There are three methods to describe a sequence. Notice that n doesn’t have to start at 1.

(a) {n+1n }∞n=1, an= n+1n , {12,23,34, . . . ,n+1n , . . .}. (b) {(−1)n3n(n+1)}∞n=1, an= (−1) n(n+1) 3n , {−23,39, −274, . . . ,(−1) n(n+1) 3n , . . .}. (c) {√n − 3}∞ n=3, an=√n − 3, n ≥ 3, {0, 1, √ 2,√3, . . . ,√n − 3, . . .}. (d) {cosnπ6 }∞n=0, an= cosnπ6 , n ≥ 0, {1, √ 3 2 ,12, 0, . . . , cos nπ 6 , . . .}.



數列不見得一定要從第一項開始寫起, 可以從第三項或是第零項開始, 例如 (c)與 (d)的說明。

Example 3 (page 695). Here are some sequences that don’t have a simple defining equation. (a) The Fibonacci sequence (費波那契數列) {fn} is defined recursively by the conditions

f1 = f2 = 1, fn= fn−1+ fn−2, n ≥ 3.

The first few terms are {1, 1, 2, 3, 5, 8, 13, 21, 34, 55, . . .}. This sequence arose when the 13th-century Italian mathematician known as Fibonacci solved a problem concerning the breeding of rabbits.

(b) If we let an be the digit in the n-th decimal place of the number

2, then {an} is a

well-defined sequence whose first few terms are {4, 1, 4, 2, 1, 3, 5, 6, 2, . . .}.

(2)

2 11.1 Sequences goo.gl/wjLEY9 Definition 4 (page 696). (數列極限之收斂或發散) b-SyPoWuIDk 數列收斂意思是極 限值存在 (實數), 否 則稱 發 散。 而 lim n→∞an= ∞的 情形, 是發散的數 列。

(1) A sequence {an} has the limit L and we write

lim

n→∞an= L or an→ L as n → ∞

if we can make the terms an as close to L as we like by taking n sufficiently large.

(2) If lim

n→∞an exists, we say the sequence converges (or is convergent,收斂). Otherwise, we

say the sequence diverges (or is divergent, 發散).

(3) If anbecomes large as n becomes large, we use the notation lim

n→∞an= ∞. 數列的極限相關定 理與函數的極限雷 同, 也有四則運算 與夾擠定理。 Theorem 5. If lim

n→∞an exists, then it is unique.

Property 6(Limit Laws for Sequences, page 697). If {an} and {bn} are convergent sequences

andc is a constant, then (1) lim

n→∞(an+ bn) = limn→∞an+ limn→∞bn.

(2) lim

n→∞(an− bn) = limn→∞an− limn→∞bn.

(3) lim

n→∞c an= c limn→∞an. In particular, n→∞lim c = c.

(4) lim n→∞(anbn) = limn→∞an· limn→∞bn. (5) lim n→∞ an bn = lim n→∞an lim n→∞bn if lim n→∞bn6= 0. (6) lim n→∞a p n=  lim n→∞an p if p > 0 and an> 0.

The Squeeze Theorem (夾擠定理, page 698). If an ≤ bn ≤ cn for n ≥ n0 and lim n→∞an = lim n→∞cn = L, then limn→∞bn= L. D8lA-tzwsEk 子數列必須依序挑 選 原 數 列 的 數 字, 因為子數列有保持 順序, 所以收斂性 也會被繼承。 當一個數列有正有 負, 可以先不考慮 符號 分 析 極 限 值, 若極限為零, 則原 數 列 限 極 亦 為 零。 這個定理算是很常 使用。 Theorem 7. If lim

n→∞an= L, then the limit of any subsequences limk→∞ank = L.



極限若存在,真相 (極限值) 只有一個!



數列的極限與函數的極限一樣有 「四則運算」 以及 「夾擠定理」。



夾擠定理,只要確定某一項之後三個數列有大小關係即可, 和前面有限項的大小無關。



子數列存在性定理一般的應用是考慮其否逆命題 — 證明原數列極限不存在。

Theorem 8 (page 698). If lim

n→∞|an| = 0, then limn→∞an= 0.

Proof. Since , by the , we have lim

n→∞an= 0.

(3)

11.1 Sequences goo.gl/wjLEY9 3

Theorem 9(page 697). If lim

x→∞f (x) = L and f (n) = anwhenn is an integer, then limn→∞an=

L. A2kSGbdhV4A 這兩個定理將函數 的極限與數列的極 限串聯起來, 定理 下方的三個註記也 應好好體會。

Theorem 10 (page 699). If lim

n→∞an= L and the function f is continuous at L, then

lim n→∞f (an) = f (L). x x y y

Figure 1: Limit relations between functions and sequences.



有了定理9、 定理 10, 就可以將上學期學過函數的極限運用到數列的極限, 超好用!



定理 10 意義: 「連續函數」 才可以和數列的 「極限」 交換順序。



若 lim n→∞an= 0, 則n→∞lim |an| = limn→∞an = 0。 (因為絕對值函數為連續函數) xr1_qaplJMU 例題示範如何利用 學到的定理處理數 列的極限; 對於例 題(d)應再強調的 是:因為n是自然 數, 並沒有離散型 的羅必達法則, 所 以若要使用羅必達 法則, 必須過渡到 改成變數為實數x 之後再使用。

Example 11. Discuss the convergence or divergence of the following sequences: (a) an= −n 2+1 2n2+3n (b) bn= nn!n (c) cn = (−1) n n (d) dn= ln nn (e) en= sin( π n). Solution.

(4)

4 11.1 Sequences goo.gl/wjLEY9

Theorem 12 (page 700). The sequence {rn}n=1 is convergent if −1 < r ≤ 1 and divergent for all other values of r. Furthermore, we have

MIBfc5shmKs 關於等比數列, 不 論是定理的所有推 論, 還有它的結論 都很重要, 必須好 好體會。 lim n→∞r n= ( 0 if − 1 < r < 1 1 if r = 1. Proof. Consider f (x) = ax. We know lim

x→∞a x= ∞ if a > 1; lim x→∞a x= 0 if 0 < a < 1. (1) Put a = r, we have (2) If r = 1, (3) If r = 0, (4) If −1 < r < 0, (5) If r = −1, (6) If r < −1,

Exercise. Show that lim

n→∞nr n= 0 if |r| < 1. Oa3TiKp31hU 認識遞增、 遞減、 單調數列與有界數 列的意義。

Definition 13 (page 700). A sequence {an} is called increasing (遞增) if an < an+1 for all

n ≥ 1, that is, a1 < a2 < a3< · · · . It is called decreasing (遞減) if an> an+1 for all n ≥ 1. A

sequence is monotonic (單調) if it is either increasing or decreasing.

Definition 14 (page 701). A sequence {an} is bounded above (有上界) if there is a number

M such that an ≤ M for all n ≥ 1. It is bounded below (有下界) if there is a number m such

that m ≤ anfor all n ≥ 1. If it is bounded above and below, then {an} is a bounded sequence

(5)

11.1 Sequences goo.gl/wjLEY9 5

Monotonic Sequence Theorem (page 702). Every bounded, monotonic sequence is con-vergent. (單調有界數列必收斂。 ) 定理應拆解成兩句 話理解: 遞增有上 界的數列收斂; 遞 減有下界的數列收 斂。 有界與單調兩 條件缺一不可。 x Figure 2: Monotonic sequence theorem.



有界數列未必收斂, 例如: 。



單調數列未必收斂, 例如: 。



定理證明要用到實數的完備性公設 (completeness axiom), 會在高等微積分的課程中詳細討論。 fyLUTsG19Q4 數學上必須先證明 數 列 的 極 限 存 在, 最後一段的極限值 找法才有意義。 也 就是說, 利用數學 歸納法證明這個遞 迴數列的極限存在 的證明是必須的。

Example 15 (page 703). Investigate the sequence {an}∞n=1 defined by the recurrence relation

(遞迴關係): a1 = 2, an+1 = 12(an+ 6) for n = 1, 2, 3, . . ..

Solution. Monotone: We claim: an+1 > an for all n ∈ N.

(1) When n = 1,

(2) Assume that it is true for n = k, that is, ak+1> ak.

(3) When n = k + 1,

(4) By , we know {an} is monotone.

Bounded: We claim: an< 6 for all n ∈ N.

(1) When n = 1,

(2) Assume that it is true for n = k, that is, ak< 6.

(3) When n = k + 1,

(4) By , we know {an} is bounded above by 6.

Limit: By , we know lim

(6)

6 11.2 Series goo.gl/FgfEX6

11.2

Series (page 707)

WG08MDVpINc 前一節討論的無窮 數列理論的目的是 為 了 研 究 無 窮 級 數, 因為數字有無 限多個, 我們永遠 無法確實地把所有 數字都加起來, 所 以利用部份和還有 極限的概念去理解 無窮級數。

Definition 1 (page 707–708). Let {an}∞n=1 be an infinite sequence.

(1) The partial sums (部份和) of the sequence {an}

n=1 is defined as sn= n X k=1 ak= a1+ a2+ · · · + an.

These partial sums form a new sequence {sn}∞n=1 (部份和數列).

(2) An infinite series (or just a series無窮級數) is denoted by

∞ X n=1 an def. = lim n→∞ n X k=1 ak = lim n→∞sn= limn→∞(a1+ a2+ · · · + an),

which means the limit of the partial sums of the sequence {an}∞n=1.

(3) If the limit lim

n→∞sn= s exists (or convergent) as a finite number, then we say the series ∞

P

n=1

anconvergent(收斂), and the number s is called the sum of the infinite series ∞

P

n=1

an

(級數和).

(4) If the sequence {sn}∞n=1 is divergent, then the series ∞

P

n=1

an is called divergent (發散).



微積分課程中感興趣的是 「無窮級數」,透過 「部份和數列的極限」 來定義無窮級數收斂或發散。

Example 2(page 708). In this chapter, we are not interested in the infinite arithmetic series (等差級數、 算數級數):

X

n=1

(a + (n − 1)d)def.= a + (a + d) + (a + 2d) + · · · + (a + (n − 1)d) + · · · ,

where each term is obtained from the preceding one by adding it by the common difference (公差) d. This is because the arithmetic series is convergent if and only if a = 0 and d = 0.



無窮等差級數除了每一項都是零的級數和收斂外, 其他情況都發散, 故不值得研究它。

Example 3 (page 709). The geometric series (等比級數、 幾何級數) is an infinite series

X

n=1

arn−1def.= a + ar + ar2+ ar3+ · · · + arn−1+ · · · , a 6= 0.

Each term is obtained from the preceding one by multiplying it by the common ratio (公比) r. We will discuss the convergence or divergence of the geometric series in the following theorem.

(7)

11.2 Series goo.gl/FgfEX6 7

Theorem 4 (page 710). The geometric series

oWYmQx95UIQ 等比級數是級數理 論的標準模型, 它 的推論還有結果都 很 重 要。 特 別 注 意: 無窮等比數列 的收斂發散結果與 無窮等比級數的收 斂發散結果是不一 樣的,要區分清楚。 ∞ X n=1 arn−1= a + ar + ar2+ ar3+ · · · + arn−1+ · · · , a 6= 0.

is convergent if |r| < 1 and its sum is

X

n=1

arn−1= a

1 − r if |r| < 1. If |r| ≥ 1, the geometric series is divergent.

Proof.

T98QRCLcasM

用這個例子想清楚

0.9和1兩者是否 一樣?

Example 5. Write the number 0.142857 = 0.142857142857 . . . as a ratio of integers (fraction). Solution.

Theorem 6 (page 713). If the series P∞

n=1

an is convergent, then lim

n→∞an= 0. 若級數收斂數 列 會 趨 近 於 零。, 則原 這個定理只是級數 收 斂 的 必 要 條 件, 不是充要條件。 我 們比較常使用的是 這個定理的否逆命 題, 也就是下面所 寫的級數發散判別 法。 Proof.

Test for Divergence (page 713). If lim

n→∞an does not exist or if n→∞lim an 6= 0, then the series ∞

P

n=1

(8)

8 11.2 Series goo.gl/FgfEX6

Example 7 (page 713). The harmonic series (調和級數) is an infinite series

R06Bfd4f-2U 調和級數的發散證 明是利用11.1的 定理7: 證明調和 級數的某個部份和 子數列發散, 則原 部份和數列發散這 個定理。 ∞ X n=1 1 n def. = 1 +1 2 + 1 3 + 1 4+ · · · + 1 n + · · · . Show that it is divergent.

Proof.



若 lim n→∞an= 0,則級數 ∞ P n=1 an 收斂與否仍舊無法判定。 例如: 比較調和級數 P∞ n=1 1 n、 等比級數 ∞ P n=1 arn−1 或歐拉級數 P∞ n=1 1 n2。 v6CVD_pFvGk 若用線性代數的語 言 來 看 這 個 定 理, 則 收 斂 的級 數 具 有 「線性」 的性質。 雖然級數的概念是 源自於部份和數列 的極限, 但是級數 的乘與除並沒有相 關的定理。 此外, 只有兩級數都收斂 的時候才有定理的 結論, 若有一個級 數發散則結論不一 定。 Theorem 8 (page 714). If P∞ n=1 an and ∞ P n=1

bn are convergent series, then so are the series ∞ P n=1 c an (where c is a constant), ∞ P n=1 (an+ bn), and ∞ P n=1 (an− bn), and (a) P∞ n=1 c an= c ∞ P n=1 an. (b) P∞ n=1 (an+ bn) = ∞ P n=1 an+ ∞ P n=1 bn. (c) P∞ n=1 (an− bn) = ∞ P n=1 an− ∞ P n=1 bn.



各別的級數和 P∞ n=1 an 與 ∞ P n=1 bn之 「收斂」 很重要。



各項相加後得到的新的級數和與各別的級數和再相加相同。



注意! P∞ n=1 anbn6= ∞ P n=1 an· ∞ P n=1 bn。 兩數列相乘的級數和不會等於各別級數和再相乘!



級數和的收斂與否和前面有限項無關。



若 P∞ n=1 an 收斂而 ∞ P n=1 bn 發散,則 ∞ P n=1 (an+ bn) 發散。(習題 11.2, #83。)



若 P∞ n=1 an 與 ∞ P n=1 bn發散,則 ∞ P n=1 (an+ bn) 不一定收斂也不一定發散。(習題 11.2, #84。)

(9)

11.3 The Integral Test and Estimates of Sums goo.gl/4dRAFG 9

11.3

The Integral Test and Estimates of Sums (page

719)

JgpkrbFR7ew 瑕積分的收斂發散 性可以幫助我們了 解級數的收斂與發 散。 這個定理是等 價敘述, 瑕積分與 級數同享收斂或發 散的性質。 積分判 別法只適用於 「正 項級數」。

The Integral Test (page 721). Suppose f (x) is a continuous, positive, decreasing function on [1, ∞) and let an= f (n). Then the series

P

n=1

an is convergent if and only if the improper

integral R∞

1 f (x) dx is convergent. In other words,

(a) IfR∞ 1 f (x) dx is convergent, then ∞ P n=1 an is convergent. (b) IfR∞ 1 f (x) dx is divergent, then ∞ P n=1 an is divergent. x x y y

Figure 1: The integral test.



函數 f (x)必須 「恆正」 與 「遞減」,函數的連續性是要讓積分比較好處理。



定理使用時不見得要 「從頭n = 1, x = 1 開始」;收斂和發散和前面有限項無關。



定理只是說明瑕積分與級數享有相同的斂散性, 並不代表兩者具有相同的值。 aRHTPX5D7XI p-級數也是級數理 論的標準模型, 其 論述與結論都必須 確實理解。

Theorem 1 (page 721). The p-series P∞

n=1 1

np (p-級數) is convergent if p > 1 and divergent if

p ≤ 1.

Proof. If p < 0, If p = 0,

(10)

10 11.3 The Integral Test and Estimates of Sums goo.gl/4dRAFG

Example 2 (page 722). Determine whether the series P∞

n=1 ln n n converges or diverges. 9jI7TEanSjg 例題示範用積分判 別法驗證級數的收 斂或發散。 另一方 面, 各位應該要體 會到這個級數是發 散的, 因為分母對 應到p= 1,而分 子終將大於 1, 所 以級數比 ∞ P n=1 1 n 還要差。 Solution.



先觀察當指標改成x 時有沒有辦法用瑕積分驗證斂散性, 可以的話再逐一檢查條件。

Estimating the Sum of a Series, page 723

ERoRqkUnhV8 利用積分判別法確 定的收斂級數, 其 級數和可以進行估 計: 給定一個誤差 之下, 可以確定要 加的項數使得級數 和與部份和之差小 於給定的誤差。 Suppose a series P∞ n=1

an is convergent by the Integral Test. We can also estimate the size of

the remainder (餘項) Rn= s − sn= an+1+ an+2+ an+3+ · · · = ∞ X k=n+1 ak.

Remainder Estimate for the Integral Test(page 718). Suppose f (k) = ak, wheref (x) is

a continuous, positive, decreasing function forx ≥ n and P∞

n=1 an is convergent. IfRn = s −sn, then Z ∞ n+1f (x) dx ≤ R n≤ Z ∞ n f (x) dx. (1)

If we add sn to each side of the inequalities (1), because sn+ Rn= s, we get

sn+ Z ∞ n+1f (x) dx ≤ s ≤ s n+ Z ∞ n f (x) dx. x x y y

(11)

11.3 The Integral Test and Estimates of Sums goo.gl/4dRAFG 11

Example 3 (page 723). Approximate the sum of the series P∞

n=1 1

n3. How many terms are

tXW4uKcot00

例題示範如何用積 分判別法進行級數 餘項估計。

required to ensure that the sum is accurate to within 0.005? Solution.

(12)

12 11.4 The Comparison Tests goo.gl/X7im2b

11.4

The Comparison Tests (page 727)

X0P1V6VewNE 比較判別法可以用 來討論複雜級數之 斂散性。 但是比較 定理只有單邊的性 質, 也就是說, 大 的級數收斂可推得 小的級數收斂; 小 的級數發散可推得 大的級數發散。 而 且, 它 只 適 用 於 「正項級數」。

The Comparison Test (page 727). Suppose that P∞

n=1

an and ∞

P

n=1

bn are series with positive

terms andan≤ bn for all n.

(a) If P∞ n=1 bn is convergent, then ∞ P n=1 an is also convergent. (b) If P∞ n=1 an is divergent, then ∞ P n=1 bn is also divergent. Proof. Let sn= n P k=1 ak, tn= n P k=1 bk, and t = ∞ P k=1 bk.

(a) Monotone: Since both series have positive terms, the sequences {sn}∞n=1 and {tn}∞n=1

are increasing.

Bounded: Since ak≤ bk for all k, we have sn≤ tn≤ t.

By the , P∞

n=1

an converges.

(b) If P∞

n=1

an is divergent, then sn→ ∞, thus tn→ ∞. Therefore ∞ P n=1 bn diverges. N7NT0rbbPzg 我們經常利用等比 級數與 p-級數這 兩個標準模型來判 定其它級數的斂散 性, 所以標準模型 的斂散性結果必須 熟知。

Most of time we use p-series and geometric series for the purpose of comparison.

(1) p-series: P∞

n=1 1

np. It is convergent if and divergent if .

(2) geometric series: P∞

n=1

arn−1. It is convergent if and divergent if .

Example 1. Show that the series P∞

n=1 n!

nn is convergent.

(13)

11.4 The Comparison Tests goo.gl/X7im2b 13

The Limit Comparison Test (page 729). Suppose that P∞

n=1

an and ∞

P

n=1

bn are series with

positive terms. If 95E2eNPi1qw 極限比較判別法相 較於比較判別法來 說 適 用 的 層 面 更 廣, 因為兩級數之 間不需建立明確的 大小關係, 只要確 定兩數列之比的極 限為正數, 則兩級 數享有一樣的斂散 性。 lim n→∞ an bn = c,

where c is a finite number and c > 0, then either both series converge or both diverge. Proof. Let m and M be positive numbers such that m < c < M . Since an

bn is close to c for

large n, there is an integer N such that m < an

bn < M ⇒ mb

n< an< M bn when n > N.

By the , we know both series converge or both diverge.



比較判別法與極限比較判別法只適用於 「正項級數」。 3fRmnVgSWBE 例題示範如何用極 限比較判別法證明 級數的斂散性。 如 何選取要比較的級 數呢? 想法是把一 般項 「最重要」 的 部份抓出來, 這個 概念與 「等級」 (or-der)有關,會在之 後的單元闡明。

Example 2 (page 730). Determine whether the following series converges or diverges.

(a) ∞ X n=1 9n 3 + 10n (b) ∞ X n=1 2n2+ 3n √ 5 + n5. Solution.

Exercise (page 726). Determine whether the following series converges or diverges.

(a) ∞ X n=1  1 + 1 n 2 e−n (b) ∞ X n=1 1 n1+n1 (c) ∞ X n=1 en1 n .

(14)

14 11.4 The Comparison Tests goo.gl/X7im2b

Estimating Sums, page 730

O0_YHbDvZg8 若用比較判別法得 知級數收斂, 則可 進行誤差估計, 誤 差的精準度會依賴 於比較的級數。

If we have used the Comparison Test to show that a series P∞

n=1

an converges by comparison

with a series P∞

n=1

bn, then we may be able to estimate the sum ∞

P

n=1

anby comparing remainders.

Consider the remainder Rn= s−sn= an+1+an+2+· · · and Tn = t−tn= bn+1+bn+2+· · · .

Since an≤ bn for all n, we have Rn≤ Tn.

(1) If P∞

n=1

bnis a p-series, we can estimate its remainder Tn as in Section 11.3.

(2) If P∞

n=1

bnis a geometric series, we can sum it exactly.

LNUObzAzm5c 以下兩個例題分別 示範比較判別法的 餘項估計, 其中一 個是用 p-級數進 行比較, 另一個是 用等比級數進行比 較。

Example 3 (page 730). Use the sum of the first 100 terms to approximate the sum of the series P∞

n=1 1

n3+1. Estimate the error involved in this approximation.

Solution.

Example 4(page 731). Use P10

n=1 cos2n

5n

.

= 0.07393 to estimate the error of the sum of the series

∞ P n=1 cos2n 5n . Solution. Exercise. Use P10 n=1 1 3n+4n .

= 0.19788 to estimate the error of the sum of the series P∞

n=1 1 3n+4n.

(15)

11.5 Alternating Series goo.gl/y9pRb5 15

11.5

Alternating Series (page 732)

gIEY2_ZYvQQ 當級數的每一項正 負 交 錯 時, 其 斂 散性可以用交錯級 數判別法測試。 當 忘掉正負符號的一 般項遞減且趨近於 零, 則交錯級數收 斂。

Definition 1 (page 732). An alternating series (交錯級數) is a series whose terms are alter-nately positive and negative.

Example 2 (page 732). Two examples of alternating series are

∞ X n=1 (−1)n−11 n = 1 − 1 2 + 1 3 − 1 4+ 1 5 − 1 6 + · · · ∞ X n=1 (−1)nn + 1n = −12+2 3 − 3 4 + 4 5 − 5 6+ 6 7 − · · · .

Alternating Series Test (page 727). If the alternating series

∞ X n=1 (−1)n−1b n= b1− b2+ b3− b4+ b5− b6+ · · · , where bn> 0, satisfies

(a) bn+1 ≤ bn for all n

(b) lim

n→∞bn= 0,

then the series is convergent.

Figure 1: Alternating series test.



交錯級數只要忘掉符號的 「某一項之後遞減」 並且 「趨近於零」,則級數收斂。 iPDMJLw-pGQ 注意到交錯級數判 別 法 只 是 必要 條 件, 當判別的條件 不滿足時, 定理不 適用; 必須要用別 的理論判定交錯級 數發散。

Example 3 (page 734). Determine whether the following series converges or diverges.

(a) ∞ X n=1 (−1)n−1 n (b) ∞ X n=1 (−1)n3n 4n − 1 . Solution.

(16)

16 11.5 Alternating Series goo.gl/y9pRb5

Example 4 (page 734). Test the series P∞

n=1(−1) n+1 n2

n3+1 for convergence or divergence.

k0D1mz4TGro 這個例題要從中學 到的是如何確定交 錯級數忘掉符號的 一般項為遞減。 將 上學期的微分理論 結合。 Solution.

Exercise. Test the series P∞

n=1(−1)

nen1 − 1 for convergence or divergence.

Estimating Sums, page 735

sP9rRIVfLiI

交錯級數的誤差估 計結果還蠻簡潔明 瞭的, 誤差不會超 過第一個餘項。

Alternating Series Estimation Theorem (page 735). If s = P∞

n=1(−1) n−1b

n is the sum of

an alternating series that satisfies (a) bn+1 ≤ bn (b) lim n→∞bn= 0, then |Rn| = |s − sn| ≤ bn+1.



「好的」 交錯級數 (滿足 (a)與 (b)),則級數和與有限項和之誤差只要看第一個餘項。



此定理只適用於 「交錯級數」,其他類型的級數不適用。

Example 5 (page 735). Find the sum of the series P∞

n=0 (−1)n

n! correct to three decimal place.

Solution. Since (n+1)!1 = (n+1)n!1 < n!1 and 0 ≤ lim

n→∞ 1

n! ≤ limn→∞1n = 0, the series ∞

P

n=0 (−1)n

n!

converges by the . By the Alternating Series Estimation Theorem we hope |s − sn| ≤ bn+1 < 0.0005, so (n + 1)! > 2000 and n ≥ 6. Hence s ≈ s6 = 0!1 − 1!1 +

1

2! − 3!1 +4!1 − 5!1 + 6!1 = 0.368056 . . .

.

= 0.368 correct to three decimal places with maximum error less than 0.001.

Exercise (page 736). How many terms of the series do we need to add in order to find the sum of the series P∞

n=1(−1) n+1 1

(17)

11.6 Absolute Convergence and the Ratio and Root Tests goo.gl/4ByLV9 17

11.6

Absolute Convergence and the Ratio and Root

Tests (page 737)

Txpji7oaJLA 級數絕對收斂顧名 思義是把每一項加 絕對值之後的級數 收斂, 將每一項先 加絕對值, 它就形 成正項級數。 Definition 1 (page 737-738). (1) A series P∞ n=1

an is called absolutely convergent (絕對收斂) if the series of absolute values ∞

P

n=1|an| is convergent.

(2) A series P∞

n=1

an is called conditionally convergent (條件收斂) if it is convergent but not

absolutely convergent. Example 2 (page 737).

(a) The series P∞

n=1 (−1)n n2 is absolutely convergent. (b) The series P∞ n=1 (−1)n n is conditionally convergent. HBAZYlvpB2Q 上學期曾經學過的 lim x→0 sin x x = 1 應該要和這個例子 聯想, 得知 sin1 n 當 n 很大的時候 和 1 n 差不多。 由 此可預測級數的斂 散性。

Example 3. Determine the series P∞

n=1(−1) nsin 1

n is absolutely convergent, conditionally

con-vergent, or divergent. Solution.

Exercise. Determine the series (a) P∞

n=1(−1) n−1√n+1−√n−1 n and (b) ∞ P n=2 (−1)n n(ln n) is absolutely

convergent, conditionally convergent, or divergent.

x_wgYkVW7Os

若 級 數 絕 對 收 斂,

則原級數收斂, 反 之不一定成立。

Theorem 4 (page 738). If a series P∞

n=1

an is absolutely convergent, then it is convergent.

(18)

18 11.6 Absolute Convergence and the Ratio and Root Tests goo.gl/4ByLV9

The Ratio Test (page 739).

比值判別法的想法 是類於於等比級數 的操作, 後項比前 項就像是公比的概 念, 加絕對值再取 極限之後會以1為 分界。 注意到極限 值若為 1, 則判別 法失效。 (a) If lim n→∞ an+1 an

= L < 1, then the series

P

n=1

an is absolutely convergent (and therefore

convergent). (b) If lim n→∞ an+1 an = L > 1 or n→∞lim an+1 an

= ∞, then the series

∞ P n=1 an is divergent. (c) If lim n→∞ an+1 an

= 1, the Ratio Test is inconclusive (無法確定的); that is, no conclusion can be drawn about the convergence or divergence of P∞

n=1 an.



加上絕對值後, 級數的 「行為」 被公比為r 的等比級數控制, 其中L < r < 1。 5_EGI1__1Bk 例題示範如何用比 值判別法得知級數 是 絕 對 收 斂 或 發 散。

Example 5(page 740). Determine whether the series is absolutely convergent, conditionally convergent, or divergent. (a) P∞

n=1(−1) n n3 3n (b) ∞ P n=1(−1) n+1 n22n n! . Solution.



帶有 「指數」 或 「階乘」 的級數,比值法 (Ratio Test)通常很好用。



帶有 「多項式」、「有理函數」 或帶有 「三角函數」,通常用比較判別法。



比值判別法無法確定的例子: P∞ n=1 1 n 發散,而 ∞ P n=1 1 n2 收斂。

Exercise (page 743). Determine whether the series is absolutely convergent, conditionally convergent, or divergent. (a) P∞

n=1 2n2 n! (b) ∞ P n=1 (n!)2 (5n)!. OfRNffD1IGw 根式判別法的想法 也是來自於等比級 數, 因為將等比級 數的一般項取絕對 值開 n 次根號後 就會出現公比。

The Root Test (page 741). (a) If lim

n→∞

n

p|an| = L < 1, then the series ∞

P

n=1

an is absolutely convergent (and therefore

convergent). (b) If lim n→∞ n p|an| = L > 1 or lim n→∞ n

p|an| = ∞, then the series ∞ P n=1 an is divergent. (c) If lim n→∞ n

(19)

11.6 Absolute Convergence and the Ratio and Root Tests goo.gl/4ByLV9 19

Example 6 (page 741). Test the convergence of the series P∞

n=1  2n+3 3n+2 n . 例題示範用根式判 別法確定級數的斂 散性。 Solution.



通常級數型如 P∞ n=1 (an)n 可考慮用根式法(Root Test)。



比值法比根式法重要一些 (11.8 之後)。

Exercise (page 743). Determine whether the series is absolutely convergent, conditionally convergent, or divergent. (a) P∞

n=1 1 + n1n2 (b) P∞ n=1 (√n −n − 1)2n.

Rearrangements, page 742

9Skjoj7F2Hk 有限個數字相加則 加 法 具 有 交 換 律。 但是無限多個數字 相加是否有交換律 呢? 這是非常深刻 的數學問題, 直到 黎曼 (Riemann) 才確 立 了 以 下 定 理: 絕對收斂的級 數 具 有 加 法 交 換 律; 條件收斂的級 數可以經過調整加 到 任 何 想 加 的 數 字。

If we rearrange the order of the terms in a finite sum, then the value of the sum remains unchanged. But it is not always the case for an infinite series.

By a rearrangement of an infinite series P∞

n=1

an (更序級數) we mean a series obtained by

simply changing the order of the terms. Formally, we will write P

σ(n)

aσ(n)where σ(n) is an one-to-one map from the natural number N to itself. For instance, a rearrangement of P

σ(n)

aσ(n) could start as follows:

a2+ a7+ a3+ a32+ a15+ a10+ a200+ · · · .

It turns out that

Theorem 7 (page 742).

(a) If P∞

n=1

anis an absolutely convergent series with sums, then any rearrangement of ∞

P

n=1

an

has the same sum s. (b) If P∞

n=1

an is a conditionally convergent series and r is any real number whatsoever, then

there is a rearrangement of P∞

n=1

(20)

20 11.6 Absolute Convergence and the Ratio and Root Tests goo.gl/4ByLV9

Example 8 (page 742). Consider the alternating harmonic series

hAVP2l9aPs4 例題示範調和級數 (條件收斂), 記和 為S,則經過順序 的調整後, 和變成 3 2S。 S = 1 − 12+ 1 3− 1 4 + 1 5 − 1 6 + 1 7− 1 8 + · · · (2) If we multiply this series by 12 and insert 0 between the terms of new series, we get

1 2S = 0 + 1 2 + 0 − 1 4 + 0 + 1 6 + 0 − 1 8 + · · · (3) Now we add the series in (2) and (3) to get

3 2S = 1 + 1 3− 1 2 + 1 5 + 1 7− 1 4+ · · · (4) Notice that the series in (4) contains the same terms as in (2).

兩個絕對收斂的級 數相乘, 也有乘法 對加法的分配律。 Theorem 9. If P∞ n=0 an and ∞ P n=0

bn are two absolutely convergent series with sum A and B,

respectively, then the product series P∞

n=0 n

P

k=0

akbn−k and any rearrangement of ∞ P n=0 n P k=0 akbn−k

has a sum equal to AB.

Appendix

aWmWSnWmzro

比 值 判 別 法 的 證 明, 設法和等比級 數建立比較關係。

Proof of Ratio Test, page 739.

(a) Since L < 1, we can choose a number r such that L < r < 1. Since lim

n→∞ an+1 an = L and L < r, the ratio an+1 an

will eventually be less than r; that is, there exists an integer N such that an+1 an < r ⇐⇒ |an+1| < |an|r whwnever n ≥ N. In general, we get |aN+k| < |aN+k−1|r < |aN+k−2|r2 < · · · < |aN|rk for all k ≥ 1.

By the Comparison Test, we know

∞ X n=N +1 |an| = ∞ X k=1 |aN+k| < ∞ X k=1 |aN|rk= |aN|r 1 − r. Hence P∞ n=1|an| is convergent, and ∞ P n=1 anis absolutely convergent. (b) If lim n→∞ an+1 an = L > 1 or limn→∞ an+1 an

= ∞, then the ratio an+1 an will eventually be greater than 1; that is, there exists an integer N such that

an+1 an > 1 |an+1| > |an| whenever n ≥ N. Since lim n→∞an6= 0, the series ∞ P n=1

(21)

11.7 Strategy for Testing Series goo.gl/1HDd1D 21

11.7

Strategy for Testing Series (page 739)

-Z4Z5J7E3ZA 前面幾個單元都是 介紹各種級數判別 法該如何使用, 並 寫出完整地數學論 述。 但 是隨 便 一 個級 數, 你要 怎 麼預判它是絕對收 斂、 條件收斂、 還 是發散? 並且要 用哪個級數比較還 有要用什麼定理論 述? 我們可以透過 等級(order)來了 解級數收斂或發散 的理論。 各位應好 好體會級數理論的 奧義, 這是高等數 學的一個很重要的 概念。

本文將討論如何利用直覺的方式判斷級數是絕對收斂

(Absolutely Convergent)

、 條件收斂

(Conditionally Convergent)

或發散

(Divergent),

以及歸納出幾個心得以快速找到證明級

數收斂或發散的判別法。

判斷級數的收斂或發散並沒有完整的標準程序

(Standard Operation Procedure),

以下

只是提供幾個經驗分享。 以下的原則大體上可以涵蓋各位將面臨到的

90%

的級數。 剩下的

10%

算是比較特殊的級數

,

例如第

25, 35, 36, 38, 50, 62, 65

,

各位需額外花時間仔細研

究其性質

,

再將結果納入心得。

(1)

心中一定要非常清楚以下兩類基本的級數收斂與發散

: • p-

級數

(p-series) P∞ n=1 1 np:

此級數當

p >1

時收斂

,

0 < p ≤ 1

時發散。

另一個是等比級數

(geometric series) P∞ n=1 arn−1:

此級數當

|r| < 1

時收斂

,

|r| ≥ 1

時發散。

(2)

利用等級

(order)

的觀念

猜測

級數是絕對收斂

(A.C.)

、 條件收斂

(C.C.)

或是發散

(Div.)

。 常見也常用的等級順序如下

: 1 ≪ ln n ≪ nk ≪ an ≪ n! ≪ nn,

其中

k >0, a > 1

(3)

尋找適當的定理

(

判別法

),

通常來說

, •

只有單一類型

,

或是不同類型的 「相加」

比較判別法

(CT, LCT)

兩種以上類型 「相乘」

,

或是帶有階乘

比值法

(Ratio T)

級數型如

(bn)n ⇒

根式法

(Root T)

級數正負交錯

交錯級數法

(AST)

特殊函數

,

例如

ln n,

觀察它是否連續化之後可以積分

積分法

(IT)

發散

⇒ (DT),

除了

AST

以外的判別法都有可能用到。

(3)

剩下的

10%

會遇到比較特殊或不顯而易見的等級

(order),

必須重新理解

,

並設法納

心得

(4)

注意到

sin n, cos n, sin 1 n,cos 1 n,tan 1 n

對待的方式不同

,

可見第

14, 21, 22, 23, 24, 34, 45, 55, 73, 87

題的分析。

(5)

熟悉以下極限也有助於判斷級數收斂或發散

: lim x→0cos x = 1 x→0lim sin x x = 1 x→0lim tan x x = 1 lim x→∞ x √ a= 1 lim x→∞ x √ x= 1 lim x→∞ 1 + 1 x x = e lim x→0(1 + x) 1 x = e

(22)

22 11.7 Strategy for Testing Series goo.gl/1HDd1D

11.7 Exercises and 11 Review

Determine whether the series is conditionally convergent, absolutely convergent, or divergent. (page 743, 746) BGxlrFgH3ZU 試著利用等級的概 念先預判級數是絕 對收斂、 條件收斂 或是發散。 然後再 挑選適當的定理或 判別法給予嚴格的 論述。 1. ∞ X n=1 1 n+ 3n 2. ∞ X n=1 (2n + 1)n n2n 3. ∞ X n=1 (−1)n n n+ 2 4. ∞ X n=1 (−1)n n n2+ 2 5. ∞ X n=1 n22n−1 (−5)n 6. ∞ X n=1 1 2n + 1 7. ∞ X n=2 1 n√ln n 8. ∞ X k=1 2kk! (k + 2)! 9. ∞ X k=1 k2e−k 10. ∞ X n=1 n2e−n3 11. ∞ X n=1  1 n3 + 1 3n  12. ∞ X k=1 1 k√k2+ 1 13. ∞ X n=1 3nn2 n! 14. ∞ X n=1 sin 2n 1 + 2n 15. ∞ X k=1 2k−13k+1 kk 16. ∞ X n=1 n2+ 1 n3+ 1 17. ∞ X n=1 n! 2 · 5 · · · (3n + 2) 18. ∞ X n=2 (−1)n−1 √ n − 1 19. ∞ X n=1 (−1)nln n√n 20. ∞ X n=1 √ n2− 1 n3+ 2n2+ 5 21. ∞ X n=1 (−1)ncos  1 n2  22. ∞ X k=1 1 2 + sin k 23. ∞ X n=1 tan 1 n  24. ∞ X n=1 nsin 1 n  25. ∞ X n=1 n! en2 26. ∞ X n=1 n2+ 1 5n 27. ∞ X k=1 kln k (k + 1)3 28. ∞ X n=1 e1n n2 29. ∞ X n=1 (−1)n cosh n 30. ∞ X j=1 (−1)j √j j+ 5 31. ∞ X k=1 5k 3k+ 4k 32. ∞ X n=1 (n!)n n4n 33. ∞ X n=1  n n+ 1 n2 34. ∞ X n=1 1 n+ n cos2n 35. ∞ X n=2 1 n1+n1 36. ∞ X n=2 1 (ln n)ln n 37. ∞ X n=1  n √ 2 − 1n 38. ∞ X n=1  n √ 2 − 1 39. ∞ X n=1 n n3+ 1 40. ∞ X n=1 n2 n+ 1 n n 41. ∞ X n=1 n3 5n 42. ∞ X n=1 (−1)n √ n+ 1 43. ∞ X n=2 1 n2√ln n 44. ∞ X n=2 ln  n 3n + 1 

(23)

11.7 Strategy for Testing Series goo.gl/1HDd1D 23 45. ∞ X n=1 cos 3n 1 + (1.2)n 46. ∞ X n=1 n2n (1 + 2n2)n 47. ∞ X n=1 1 · 3 · · · · (2n − 1) 5nn! 48. ∞ X n=1 (−5)2n n29n 49. ∞ X n=1 (−1)n−1 √n n+ 1 50. ∞ X n=1 √ n+ 1 −√n − 1 n 51. ∞ X n=1 (−1)n−1n−1 3 52. ∞ X n=1 (−1)n−1n−3 53. ∞ X n=1 (−1)n(n + 1)3n 22n+1 54. ∞ X n=1 (−1)n√n ln n 55. ∞ X n=1 cos nπ 3  n! 56. ∞ X n=1 (−2)n nn 57. ∞ X n=1  n2+ 1 2n2+ 1 n 58. ∞ X n=2  −2n n+ 1 5n 59. ∞ X n=1  1 + 1 n n2 60. ∞ X n=1 (2n)! (n!)2 61. ∞ X n=1 n100100n n! 62. ∞ X n=1 n! 2n2 63. ∞ X n=1 2nn! nn 64. ∞ X k=1 3 √ k − 1 k(√k+ 1) 65. ∞ X n=1 √ n+ 11 − cosπn 66. ∞ X n=1 n2e−n3 67. ∞ X n=0 (−1)n π 2n (2n)! 68. ∞ X n=1 (−1)n−1n 4 4n 69. ∞ X n=0 (tan−1n)n 70. ∞ X n=2 (−1)n−1 (ln n)n 71. ∞ X n=3 1 ln(n!) 72. ∞ X n=2  n ln n n 73. ∞ X n=1 ntan 1 2n 74. ∞ X n=1 1 (ln(n + 1))n 75. ∞ X n=1  1 n− sin 1 n  76. ∞ X n=1 nn−1 (2n2+ n + 1)n+22 77. ∞ X n=1 (−1)n n − ln n 78. ∞ X n=2 (−1)n√n n − 1 79. ∞ X n=1 (−1)n(n + 1)! nn+1 80. ∞ X n=1 (−1)nln n + 1 n  81.1 −1 · 3 3! + 1 · 3 · 5 5! − 1 · 3 · 5 · 7 7! + · · · + (−1) n−11 · 3 · 5 · · · · (2n − 1) (2n − 1)! + · · · 82.2 5+ 2 · 6 5 · 8+ 2 · 6 · 10 5 · 8 · 11+ 2 · 6 · 10 · 14 5 · 8 · 11 · 14+ · · · 83. ∞ X n=1 2 · 4 · 6 · · · (2n) n! 84. ∞ X n=1 (−1)n 2nn! 5 · 8 · 11 · · · (3n + 2) 85. 1 ln 2− 1 ln 3+ 1 ln 4− 1 ln 5+ · · · 86.1 + 1 + 2 1 + 22 + · · · + 1 + n 1 + n2 + · · · 87. sin π 2 + sin π 22 + · · · + sin π 2n + · · ·

(24)

24 11.8 Power Series goo.gl/Cyk7iL

11.8

Power Series (page 746)

2kojP3VuFN4 前幾節所學的是無 窮級數, 現在要把 函 數 的 概 念結 合, 變成每代入一個數 都要問一次級數的 收斂或發散。 那些 使級數收斂的點與 級 數 和 的 對 應 關 係變 成 函 數 的 概 念。 冪級數較為特 別, 它的一般項是 cnxn的形式。

Definition 1 (page 746). A power series (冪級數) is a series of the form

X

n=0

cnxn= c0+ c1x + c2x2+ c3x3+ · · · ,

where x is a variable and the cn’s are constants called the coefficients (係數) of the series.

A power series may converge for some values of x and diverge for other values of x. The sum of the series is a function

f (x) = c0+ c1x + c2x2+ c3x3+ · · · + cnxn+ · · ·

whose domain (定義域) is the set of all x for which the series converges.



「冪級數」 可想成是 「多項式」 的推廣— 多了極限的運算。



「冪級數」 是一個函數f (x), 函數的定義域是級數收斂所成的集合。

Example 2 (page 746). If cn≡ 1, the power series becomes the geometric series ∞

X

n=0

xn= 1 + x + x2+ · · · + xn+ · · · ,

which converges when and diverges when .

我們也可以討論中 心移到 x= a的 冪級數。 注意這裡 有一些記號上的約 定。

Definition 3 (page 747). A series of the form

X

n=0

cn(x − a)n = c0+ c1(x − a) + c2(x − a)2+ · · ·

is called a power series in (x − a) (以 (x − a) 形式的冪級數) or a power series centered at a (以 a為中心的冪級數) or power series about a (關於 a的冪級數).



約定 (x − a)0 ≡ 1,即使 x = a也是如此。



任何關於 a的冪級數,必在 x = a 收斂,所以冪級數的定義域是非空集合。 S3GVhXpAEZo 冪級 數 的 收 斂 定 理, 直接利用比值 判別法的結果順勢 而得。 注意端點的 收斂性總是要另外 討論。

Theorem 4 (page 749). For a given power series P∞

n=0

cn(x − a)n there are only three

possi-bilities:

(a) The series converges only when x = a. (b) The series converges for all x.

(c) There is a positive number R such that the series converges if |x − a| < R and diverges if |x − a| > R. (注意此定理還不完整,端點收斂行為因級數而異。)

(25)

11.8 Power Series goo.gl/Cyk7iL 25

Definition 5 (page 749).

(1) The number R in case (c) is called the radius of convergence (收斂半徑) of the power series.

(2) By convention, the radius of convergence is R = 0 in case (a) and R = ∞ in case (b). (3) The interval of convergence (收斂區間) of a power series is the interval that consists

of all values of x for which the series converges. When x is an endpoint (端點) of the interval, that is, x = a ± R, anything can happen – the interval of convergence could be

(a − R, a + R) (a − R, a + R] [a − R, a + R) [a − R, a + R]. uUmfwcHLv7s 這裡的學習, 除了 要 會 確 實 論 述 冪 級 數 的 收 斂 或 發 散, 也要會從等級 (order) 的概念去 感 受 冪 級 數 的 特 性。

Example 6 (page 747). Find the interval of the convergence of the following series:

(a) ∞ X n=0 n!xn (b) ∞ X n=0 (−1)nx2n

22n(n!)2 (Bessel function of order 0) (c) ∞ X n=1 1 n(x − 3) n. Solution.

(26)

26 11.9 Representations of Functions as Power Series goo.gl/R1h9Dc

11.9

Representations of Functions as Power Series

(page 752)

YMXmwlzqXZY 這個單元要介紹幾 個從等比級數出發 透過一些基本運算 下就可以順勢寫出 的 冪 級 數。 這 裡 必須先對 「公式」 確實理解才有辦法 做出千變萬化的結 果。

In this section, we learn how to represent certain types of functions as sums of power series. We will see that it is useful for integrating functions that don’t have elementary antiderivatives, for solving differential equations, and for approximating functions by polynomials.

Example 1 (page 752). Recall that the geometric series: 1 1 − x = 1 + x + x 2+ x3 + · · · + xn+ · · · = ∞ X n=0 xn if |x| < 1.

We can express the following functions by manipulating geometric series:

(1) 1 1 + x2 =

(2) x 2 + x =

Differentiation and Integration of Power Series, page 754

冪級數在收斂區間 的內部有很好的操 作特性, 就像多項 式一般地操作微分 與積分。 注意 (a) 的情況, 微分後養 成習慣把取和的下 指標改成n = 1, 在之後的一些計算 上比較不會產生困 擾。

Theorem 2 (page 754). If the power series P∞

n=0

cn(x − a)n has radius of convergence R > 0,

then the function f (x) defined by

f (x) = c0+ c1(x − a) + c2(x − a)2+ · · · = ∞

X

n=0

cn(x − a)n

is differentiable (and therefore continuous) on the interval (a − R, a + R) and

(a) f′(x) = c 1+ 2c2(x − a) + 3c3(x − a)2+ · · · = ∞ X n=1 ncn(x − a)n−1. (b) Z f (x) dx = C + c0(x − a) + c1(x − a) 2 2 + c2 (x − a)3 3 + · · · = C + ∞ X n=0 cn(x − a) n+1 n + 1 .



「冪級數」 和 「多項式」 一樣,可以逐項微分、 逐項積分,而且收斂 「半徑」 不變。

(27)

11.9 Representations of Functions as Power Series goo.gl/R1h9Dc 27



重新看待定理中的 (a), (b), 對於收斂的冪級數: PKUbCdvWTrg 用這樣的表示法可 以把定理看得很清 楚, 它是微分或積 分與求和之間的互 換, 在冪級數的情 況下是合法的。 (a) d dx ∞ X n=0 cn(x − a)n ! = ∞ X n=0 d dx(cn(x − a) n) 「微分」 和 「求和、 極限」 可交換。 (b) Z ∞ X n=0 cn(x − a)n ! dx = ∞ X n=0 Z cn(x − a)ndx 「積分」 和 「求和、 極限」 可交換。



「收斂半徑」 相同不代表 「收斂範圍」 相同 (端點的收斂性會改變), 所以端點一律重新檢查。 這三個例子非常經 典, 特別是對數函 數 與 反 正 切 函 數, 它們的微分正好可 以和等比級數公式 對應, 所以可以順 勢地改寫。

Example 3 (page 745). Express the following function as a power series and find its interval of convergence.

(1) f (x) = 1

(1 − x)2 (2) g(x) = ln(1 + x) (3) h(x) = tan−1x.

(28)

28 11.9 Representations of Functions as Power Series goo.gl/R1h9Dc

Example 4 (page 758). Find the sum of each of the following series.

0njTFhZc8yc 有些級數的求和問 題可以用冪級數的 觀點處理, 試著透 過係數與次方中的 n的關係推理出它 和冪級數及其微分 或積分的關聯。 (1) ∞ X n=1 nxn, |x| < 1 (2) ∞ X n=1 n 2n. Solution. aDV_lSIgMKk 對於那些積不出來 的函數或是積分處 理 很 困 難 的 函 數, 若要尋求定積分或 瑕積分的值, 退而 求其次地, 在允許 一個誤差之下, 改 用冪級數展開, 研 究冪級數要加到多 少項之下的積分值 與真正值之間的差 在誤差範圍內, 用 這種方式理解積分 之意義。

Example 5 (page 750). Evaluate R 1

1+x7dx as a power series and approximate

R0.5

0 1 1+x7 dx

correct to within 10−7.

Solution. We express the integrand and then integrate term by term: 1

1 + x7 =

Z 1

1 + x7dx =

This series converges for , that is .

Z 0.5

0

1

1 + x7dx =

When we choose n = 3, by the Alternating Series Estimation Theorem, the error is smaller than the term with b4 = 29·2129 ≈ 6.4 × 10−11, so we have

Z 0.5

0

1

(29)

11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN 29

11.10

Taylor and Maclaurin Series (page 759)

O2stBBo5Wvs 這一節的目的是要 研究其它函數能不 能重新表示成冪級 數的形式。 研究的 方法第一步是 「假 設」 函數可以順利 地寫成冪級數, 先 得到冪級數每一項 係數的表達式。 這一節是級數理論 的重頭戲, 應徹底 理解與體會。

In this section, we will answer two questions: Which functions have power series representa-tion? How can we find such representarepresenta-tion?

First, suppose that a smooth function f (x) can be represented by a power series:

f (x) = c0+ c1(x − a) + c2(x − a)2+ c3(x − a)3+ · · · , if |x − a| < R. (1)

• Put x = a, then we get .

• Since f′(x) = ,

we put x = a and get .

• Since f′′(x) = , we put x = a and get .

• By induction, since f(k)(x) = , we have .

Theorem 1 (page 759). If f (x) has a power series representation (expansion) at a:

f (x) =

X

n=0

cn(x − a)n for |x − a| < R

then its coefficients are given by the formula cn=

f(n)(a) n! . 由剛才的討論, 對 於函數在某一點可 以無限次微分的函 數, 我們就可以定 義泰勒級數或是馬 克勞林級數。 之後 的篇幅是要研究原 函數及其泰勒級數 或馬克勞林級數的 關係。 通常我們會先研究 馬克勞林級數, 對 於中心不同的泰勒 級數之情況, 只要 再知道一些平移的 理論或轉換式, 就 可以把式子改寫。

Definition 2 (page 760). Given a smooth function f (x), define the Taylor series of the function f (x) at a (or about a or centered at a) (函數 f (x)在 x = a處的泰勒級數) by

T (x)def.= ∞ X n=0 f(n)(a) n! (x − a) n= f (a) + f′(a) 1! (x − a) + f′′(a) 2! (x − a) 2+ · · · . (2)

For the special cases a = 0 the series (2) becomes

M (x)def.= ∞ X n=0 f(n)(0) n! x n= f (0) + f′(0) 1! x + f′′(0) 2! x 2 + · · · .

This case the function M (x) is given the special name Maclaurin series (馬克勞林級數).



由前面討論知道: 「若f (x) 可表示成冪級數時」,則 f (x)和它的泰勒級數 T (x)一致。



我們必須追問 (研究): 有哪些函數 「可以」 寫成冪級數?(存在函數無法表示成冪級數。)

(30)

30 11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN

Example 3 (page 760). Find the Maclaurin series of the function f (x) = ex and its radius

of convergence. 4VY_jxCsk5g 這個例題是先求出 指數函數的馬克勞 林級數。 注意到例 題中的後半段討論 收斂性只是在了解 冪級數的定義域而 已。

Solution. Since f(n)(x) = , we know that f(n)(0) = for all n ∈ N or n = 0.

Therefore the Maclaurin series of f (x) = ex is ∞ X n=0 f(n)(0) n! x n =

To find the radius of convergence, we let an= , then

an+1 an =

By the , the radius of convergence is .

xvg6yDTPmq8 給了函數及其泰勒 級 數, 現 在 要 開 始研究兩者是否相 等。 首先把泰勒級 數分成兩部份, 一 個是 n 階泰勒多 項式, 另一部份是 餘項。

Question 4 (page 761). Under what circumstances is a function equal to the sum of its Taylor series? In other words, if f (x) has derivatives of all orders, when is it true that

f (x)= T (x)? def.= ∞ X n=0 f(n)(a) n! (x − a) ndef. = lim n→∞Tn(x), where Tn(x) = n X k=0 f(k)(a) k! (x − a) k= f (a) +f′(a) 1! (x − a) + · · · + f(n)(a) n! (x − a) n. (3) Definition 5 (page 761).

(a) The polynomial Tn(x) in (3) is called n-th degree Taylor polynomial of f (x) at a (f (x)

在x = a的 n-階泰勒多項式).

(b) Define the remainder (餘項) of the Taylor series as rn(x)

def. = f (x) − Tn(x). 這個定理告知函數 及其泰勒級數相等 的等價條件是餘項 趨近 於 零。 如 果 你把事情想清楚的 話, 就會覺得這個 定理是蠻顯然的。

Theorem 6 (page 761). A smooth function f (x) = T (x) on the interval |x − a| < R if and only if lim

n→∞rn(x) = 0 for |x − a| < R.

Proof. (⇒) Since f (x) = lim

n→∞Tn(x) and rn(x) = f (x) − Tn(x), we have

lim

n→∞rn(x) = limn→∞(f (x) − Tn(x)) = f (x) − limn→∞Tn(x) = f (x) − f (x) = 0.

(⇐) Conversely, since lim

n→∞rn(x) = 0 and Tn(x) = f (x) − rn(x), we have

T (x) = lim

n→∞Tn(x) = limn→∞(f (x) − rn(x)) = f (x) − limn→∞rn(x) = f (x) − 0 = f (x).



想清楚:函數是否與其泰勒級數 「相同」,和泰勒級數的 「收斂範圍」 是兩回事。

(31)

11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN 31

Question 7 (page 762). How do we show that lim

n→∞rn(x) = 0 for a specific function f (x)?

903ROGCtJ1M 從 前 一 個 定 理 知 道: 函數及其泰勒 級數相等的等價條 件 是餘 項 趨 近 於 零。 但是實際給了 函數, 要怎麼證明 餘項趨近於零, 這 反而是相當困難的 問題。 這一頁主要 是在講如何估計餘 項。 先將餘項 (原 本 是 無 限 多 項 相 加) 表示成一個簡 單的型式; 再得到 下面 的 泰 勒 不 等 式。

Theorem 8. Suppose that f (x) has continuous derivative at x = a up to n + 1 order, then

f (x) = f (a) + f′(a) 1! (x − a) + · · · + f(n)(a) n! (x − a) n+ r n(x) = Tn(x) + rn(x), where rn(x) = f(n+1)(c) (n + 1)! (x − a)

n+1,c is a number between a and x.

Proof. Without loss of generality, we assume a < x. Consider the function

g(t) = f (x) − f (t) − f′1!(t)(x − t) − · · · − f

(n)(t)

n! (x − t)

n,

then g(t) is continuous on [a, x], and

g′(t) = − n X k=0 f(k+1)(t) k! (x − t) k − n X k=1 f(k)(t) k! k(x − t) k−1(−1) = − n X k=0 f(k+1)(t) k! (x − t) k+ n X k=1 f(k)(t) (k − 1)!(x − t) k−1 = − n X k=0 f(k+1)(t) k! (x − t) k+n−1X k=0 f(k+1)(t) k! (x − t) k = −f (n+1)(t) n! (x − t) n.

Let h(t) = (x − t)n+1, by the Cauchy Theorem (generalized Mean Value Theorem), then there exists c ∈ (a, x) such that

g′(c) h′(c) = g(x) − g(a) h(x) − h(a) ⇒ −f(n+1)(c)(x−c)n! n −(n + 1)(x − c)n = 0 − rn(x) 0 − (x − a)n+1, so rn(x) = f(n+1)(c) (n + 1)! (x − a) n+1.



想成是 「均值定理」 的高階版本, 餘項形式和泰勒多項式一樣, 只是高次微分處代入 c。

Once we have this expression of the remainder, we can estimate it by the following theorem.

泰勒不等式的一個 重點是函數n+ 1 次微分後的絕對值 小於等於M,這個 M不能和n有關, 這樣才有機會用階 乘去控制指數而得 到餘項趨近於零。

Taylor’s Inequality (page 762). If |f(n+1)(x)| ≤ M for |x − a| ≤ d, then the remainder rn(x) of the Taylor series satisfies the inequality

|rn(x)| = f(n+1)(c) (n + 1)! (x − a) n+1 ≤ M (n + 1)!|x − a| n+1 for |x − a| ≤ d.

(32)

32 11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN

Example 9 (page 763).

(1) Prove that ex is equal to the sum of Maclaurin series.

(2) Find the Taylor series for f (x) = ex at a = 2.

nrcD9xhLLjI 在認識函數及其泰 勒級數相等的等價 條 件 後, 現 在 要 開始研究基本函數 及其泰勒級數的關 係。 先以指數函數 為例, 確實驗證兩 者對所有實數都一 致。 此外, 也用中 心為2的例子了解 指數函數在不同點 的泰勒級數和原函 數的關係。 Solution.

(1) If f (x) = ex, then f(n)(x) = ex for all n ∈ N. Given x ∈ R, there is a positive number

d such that |x| ≤ d. Since |f(n+1)(x)| = ex≤ ed, we get

|rn(x)| = f(n+1)(c) (n + 1)! x n+1 ≤ for |x| ≤ d.

Notice that ed is a number independent of n, so we have

lim

n→∞

ed

(n + 1)!|x|

n+1 =

By the Squeeze Theorem lim

n→∞rn(x) = 0, and e

x= P∞

n=0 1

n!xn for all x ∈ R.

(2) We have f(n)(2) = e2, so the Taylor series for f (x) = ex at x = 2 is

Another viewpoint is . bNdXszACQ3s 這個例子是探討正 弦函數及其馬克勞 林級數的關係, 結 論也是非常好, 處 處收斂且相等。

Example 10(page 764). Find the Maclaurin series for f (x) = sin x. Prove that it represents sin x for all x.

Solution. We compute for k = 0, 1, 2, 3, . . .,

f(4k)(x) = f(4k+1)(x) = f(4k+2)(x) = f(4k+3)(x) = f(4k)(0) = f(4k+1)(0) = f(4k+2)(0) = f(4k+3)(0) = so the Maclaurin series for f (x) = sin x is

Since f(n+1)(x) is ± sin x or ± cos x, we know that |f(n+1)(x)| ≤ 1 for all x ∈ R. By Taylor’s

Inequality:

|rn(x)| =

Since lim

n→∞ , we have limn→∞rn(x) = 0 for all x ∈ R by the Squeeze

Theorem. Thus sin x is equal to the sum of its Maclaurin series P∞

n=0 (−1)n

(33)

11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN 33

Example 11 (page 764–765).

(1) Represent f (x) = sin x as the sum of its Taylor series centered at x = π3.

1VqXTH3fXfY 餘弦函數與其泰勒 級數的關係, 也可 以仿照之前的方法 再操作一次。 而這 個例題要示範的是 透過一些三角函數 的關係式還有冪級 數的逐項微分逐項 積分理論求得。 後 者的處理將有助於 快速變化出更多函 數的泰勒級數, 免 於總是土法煉鋼般 地枯燥討論。

(2) Find the Maclaurin series for cos x. (3) Find the Maclaurin series for x cos x. Solution. We have for k = 0, 1, 2, 3, . . .

f(4k)(x) = f(4k+1)(x) = f(4k+2)(x) = f(4k+3)(x) =

f(4k)(π

3) = f(4k+1)(π3) = f(4k+2)(π3) = f(4k+3)(π3) =

(1) The Taylor series at π 3 is

(2) Instead of computing derivatives and substituting in the Maclaurin series for cos x, we can differentiate the Maclaurin series for sin x:

cos x =

Since the Maclaurin series for sin x converges for all x, the differential series for cos x also converges for all x.

(3) We can multiply the series for cos x by x: x cos x = s7gCTjaAgkI 現在要討論的是二 項式函數及其泰勒 級數的關係。 第一 步仍然是要先把二 項式函數的泰勒級 數表示出來。 注意 到二項式函數的次 方 m可以是任何 的實數。

Example 12 (page 766). Find the Maclaurin series for f (x) = (1 + x)m, where m is any real

number. Solution.

(34)

34 11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN

Example 13 (page 766). Find the radius of convergence of the binomial series (二項式級數,

從上一個例子推得) P∞ n=0 m(m−1)···(m−n+1) n! xn. 接下來要確定由二 項式函數對應到的 馬克勞林級數之定 義域, 也就是級數 的收斂性。 收斂半 徑較容易理解, 但 是端點的收斂情況 與m有關,較為複 雜, 將留到這一章 的最後補充說明。 將函數稱為二項式 函數的原因是其級 數的係數將組合數 Cm n 的概念推廣, 這時 m 可以允許 是任何實數。

Solution. If m is a nonnegative integer, then the terms are eventually 0 and so the series is finite. For other values of m, if the n-th term is an, then

an+1 an =

By the , the binomial series converges if and diverges if , and the radius of convergence is .

The Binomial Series (page 767). If m is any real number and |x| < 1, then

(1 + x)m = ∞ X n=0 Cnmxn= 1 + mx +m(m − 1) 2! x 2+m(m − 1)(m − 2) 3! x 3 + · · · .

The interval of convergence depends on m: (−1, 1) if m ≤ −1; (−1, 1] if −1 < m < 0; [−1, 1] if m > 0.



直接估計餘項趨近於零比較麻煩, 有其他的方法證明二項式函數與二項式級數 「相同」。

Definition 14 (page 766). Numbers Cm n =

m(m−1)(m−2)···(m−n+1)

n! are called binomial

coeffi-cients (二項式係數). Remark that Cm

0 ≡ 1 for all m ∈ R. UMOZvHfMms4 例題以 m= −1 2 的二項式函數討論 其馬克勞林級數。

Example 15 (page 767). Find the Maclaurin series for g(x) = √1

4−x and its radius of

con-vergence.

Solution. We rewrite f (x) in a form where we can use the binomial series: 1

4 − x =

Using the binomial series with m = and with x replaced by , we have 1

4 − x =

(35)

11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN 35

Important Maclaurin series and their radii of convergence

ndu0zmvTjXc 這個部份總結基本 函數與其泰勒級數 的關係。 我們可以 透過等級、 奇偶性、 在原點附近的行為 把這七個函數的泰 勒級數熟記。 (1) 1 1 − x = ∞ X n=0 xn= 1 + x + x2+ x3+ · · · R = 1 (2) ex= ∞ X n=0 xn n! = 1 + x 1!+ x2 2! + x3 3! + · · · R = ∞ (3) sin x = ∞ X n=0 (−1)nx2n+1 (2n + 1)! = x − x3 3! + x5 5! − x7 7! + · · · R = ∞ (4) cos x = ∞ X n=0 (−1)nx2n (2n)! = 1 − x2 2! + x4 4! − x6 6! + · · · R = ∞ (5) tan−1x = ∞ X n=0 (−1)nx2n+1 2n + 1 = x − x3 3 + x5 5 − x7 7 + · · · R = 1 (6) ln(1 + x) = ∞ X n=1 (−1)n−1xn n = x − x2 2 + x3 3 − x4 4 + · · · R = 1 (7) (1 + x)m = ∞ X n=0 Cnmxn= 1 + mx + m(m − 1) 2! x 2+m(m − 1)(m − 2) 3! + · · · R = 1 _XBE_z7S114 現在要開始實際應 用, 從泰勒及數的 觀點重新理解微積 分理論。 級數和的 問題也可以與函數 的泰勒級數進行聯 想。

Example 16 (page 768). Find the sum of the series 1 1 · 2 − 1 2 · 22 + 1 3 · 23 − 1 4 · 24 + · · · . Solution. 極限問題也可以重 新認識, 用泰勒級 數的觀點了解函數 的基本行為, 可以 把極 限 看 得 更 清 楚。

Example 17 (page 769). Evaluate lim

x→0

ex− 1 − x

x2 .

Solution. Using the Maclaurin series for ex, we have

lim

x→0

ex− 1 − x

x2 =

=

(36)

36 11.10 Taylor and Maclaurin Series goo.gl/7ZQuaN

Multiplication and division of power series, page 770

Example 18 (page 770). Find the first three nonzero terms in the Maclaurin series for (1) exsin x and (2) tan x.

4rSByKcMCfk 兩函數相乘或相除 之泰勒級數也可以 直接透過像多項式 那樣直接四則運算 求得。 Solution. Example 19(page 768). MlVCJ930aac 例題中的積分是積 不出來的, 也就是 不可能透過積分技 巧把不定積分的結 果用初等函數的方 式表達。 然而這個 積分在統計、 物理 還有高等的數學理 論有著非常重要的 地位。 對於這個積 分的認識, 退而求 其次的方法就是將 函數改用泰勒級數 表示並進行誤差估 計。 (1) Evaluate Z e−x2dx as an infinite series. (2) Evaluate Z 1 0

e−x2dx correct to within an error of 0.001. Solution.

(1) We replace x with −x2 in the series for ex and get, for all x ∈ R,

e−x2 =

We integrate term by term: Z

e−x2

dx = .

The series is convergent because e−x2

is convergent . (2) We compute Z 1 0 e−x2dx = = ≈

(37)

11.11 Applications of Taylor Polynomials goo.gl/SbJNXE 37

11.11

Applications of Taylor Polynomials (page 774)

gaZEs0i5haI 這一單元將討論泰 勒級 數 的 各 式 應 用, 首先要把泰勒 級 數 的 概 念 與 上 學 期 所 學 的 微 分 (differential) 結 合。 微分只是一階 或線性的近似, 而 泰勒多項式就是高 階或是多項式的近 似。

In this section we explore some applications of Taylor polynomials. Computer scientists like them because polynomials are the simplest of functions. Physicists and engineers use them in such fields as relativity, optics, blackbody radiation, electric dipoles, the velocity of water waves, and building highways across a desert.

Approximating Functions by Polynomials, page 774

Recall that the linear approximation of f (x) at x = a (in section 3.10):

f (x) ≈ f (a) + f′(a)(x − a) (1)

Right hand side of (1), called the linearization of f (x) at x = a, is the first-degree Taylor polynomial T1(x). If f (x) is the sum of its Taylor series, then Tn(x) → f (x) as n → ∞, and

so Tn(x), nth-degree Taylor polynomial of f (x) at x = a, can be used as an approximation to

f (x): f (x) ≈ Tn(x) = n X k=0 f(k)(a) k! (x − a) k.

When using a Taylor polynomial Tn(x) to approximate a function f (x), we have to ask that

how good an approximation is it? How large should we take n to be in order to achieve a desired accuracy? To answer these questions we need to look at the absolute value of the remainder |rn(x)| = |Rn(x)| = |f (x) − Tn(x)|. Here we remark that if f (x) is the sum of its

Taylor series, then rn(x) = Rn(x).

There are three possible methods for estimating the size of the error:

(1) If the series is an alternating series, we can use the Alternating Series Estimation The-orem.

(2) In all cases we can use Taylor Inequality: If |f(n+1)(x)| ≤ M for |x − a| ≤ d, then

|rn(x)| = f(n+1)(c) (n + 1)! (x − a) n+1 ≤ M (n + 1)!|x − a| n+1 for |x − a| ≤ d.

(3) If a graphing device is available, we can use it to graph (estimate) |Rn(x)|.

GuYLS60ySIU 我們也可以用數學 繪圖軟體透過圖形 的方法感受函數及 其泰勒級數的相關 性。 看圖形變化時 應著重於他們的近 似範圍以及近似程 度。

Example 1. Desmos Graphing Calculator is a free, online, graphing calculator: https://www.desmos.com/calculator

https://desmos.s3.amazonaws.com/Desmos_User_Guide.pdf

We will illustrate Taylor polynomial approximations by Desmos Calculator with some impor-tant examples.

(38)

38 11.11 Applications of Taylor Polynomials goo.gl/SbJNXE

Example 2 (page 775).

(a) Approximate f (x) =√3x by a Taylor polynomial of degree 2 at a = 8.

Q8xluK-BGQE 在以前沒有計算機 的時代, 用泰勒級 數估計一些特別的 數 字 顯 得 非 常 重 要, 因為它是一個 用多項式還有極限 的方法認識一些數 學量。

(b) How accurate is this approximation when 7 ≤ x ≤ 9?

Solution.

(a) We compute

f (x) = f′(x) = f′′(x) = f′′′(x) =

f (8) = f′(8) = f′′(8) =

So the desired approximation is

3

x ≈ T2(x) =

=

(b) We can use Taylor’s Inequality with n = 2 at a = 8:

|r2(x)| ≤

Thus, if 7 ≤ x ≤ 9, the approximation in (a) is accurate to within .

VYKj43S7HpE 三角函數的估算也 可以用泰勒級數的 方法求得。 現在雖 然我們可以直接按 計算機求得其精確 值, 但是計算機的 構造原理是來自於 泰勒級數的理論。

Example 3 (page 776). What is the maximum error possible in using the approximation sin x ≈ x − x3!3 +

x5

5! when −0.3 ≤ x ≤ 0.3? Use this approximation to find sin 12◦ correct to

six decimal places.

Solution. Notice that the Maclaurin series sin x = x − x3!3 + x5!5 x7!7 + · · · is alternating for all x 6= 0, and the successive terms decrease in size because |x| < 1, so we can use the . The error in approximating sin x by the first three terms of its Maclaurin series is at most

x7 7! = |x| 7 5040 ≤

To find sin 12◦, we first convert to radian measure:

sin 12◦ = sin12 · π 180  = sinπ 15  ≈

(39)

11.11 Applications of Taylor Polynomials goo.gl/SbJNXE 39

Applications to Physics, page 778

-cXS6MnYExU 愛因斯坦的狹義相 對論與牛頓力學的 關係也可以用泰勒 級 數 的 方 法 理 解: 愛因斯坦理論在微 觀尺度下與牛頓力 學相當。

Example 4(page 778). In Einstein’s theory of special relativity the mass of an object moving with velocity v is

m = qm0 1 −vc22

,

where m0 is the mass of the object when at rest and c is the speed of light. The kinetic energy

of the object is the difference between its total energy and its energy at rest: K = mc2−m 0c2.

(a) Show that when v is very small compared with c, this expression for K agrees with classical Newtonian physics: K = 12m0v2.

(b) Use Taylor’s Inequality to estimate the difference in these expressions for K when |v| ≤ 100 m/s.

Solution.

(a) Using the expressions given for K and m, we get

K = mc2− m0c2 = m0c2 q 1 −vc22 − m0c2 = m0c2  1 − v 2 c2 −12 − 1 ! .

With x = −vc22, the Maclaurin series for (1 + x)− 1

2 is a binomial series with m = −1

2. Therefore we have (1 + x)−12 = 1 −1 2x + −12  −32  2! x 2+ −12  −32  −52  3! x 3 + · · · = 1 −12x +3 8x 2 −165 x3+ · · · , and K = m0c2  1 +1 2 v2 c2 + 3 8 v4 c4 + 5 16 v6 c6 + · · ·  − 1  = m0c2  1 2 v2 c2 + 3 8 v4 c4 + 5 16 v6 c6 + · · ·  .

If v is much smaller than c, then all terms after the first are very small when compared with the first term. If we omit them, we get

K = m0c2 1 2 v2 c2  = 1 2m0v 2. (b) Let f (x) = m0c2  (1 + x)−12 − 1 

with x = −vc22. We can use Taylor’s Inequality to

write r1(x) = f′′c) 2! x 2, where −v2 c2 ≤ ˜c ≤ 0.

數據

Figure 1: Limit relations between functions and sequences.
Figure 1: The integral test.
Figure 2: Remainder estimate for the Integral Test.
Figure 1: Alternating series test.

參考文獻

相關文件

In particular, there exist conditionally convergent series.. W EN -C HING L IEN Advanced

Articles of this Chapter, other than those of headings 96.01 to 96.06 or 96.15, remain classified in the Chapter whether or not composed wholly or partly of precious metal or metal

For importation of cinema film, no matter whether the film is positive or negative, an approval issued by Bureau of Audiovisual and Music Industry Development, Ministry of Culture

Whether possesses good or evil sīla, transgresses or breaks sīla, or defiles or purifies sīla follows. investigation and is determined by volition in the belongings of mind, which

Quadratically convergent sequences generally converge much more quickly thank those that converge only linearly.

“Find sufficiently accurate starting approximate solution by using Steepest Descent method” + ”Compute convergent solution by using Newton-based methods”. The method of

  Whether the complaint is made in written form or in person, the complainant should provide his/her name, correspondence/e-mail address and/or contact phone

Every convergent sequence is bounded.. W EN -C HING L IEN Advanced