Forgot password?
 Register account
View 1285|Reply 11

2019 AMC 10年级(相当于高一) 样题一

[Copy link]

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

业余的业余 Posted 2019-2-19 22:38 |Read mode
$ p,q, $ 和 $ r $ 是多项式 $ x^3-22x^2+80x-67  $ 的不同根。已知存在实数 $ A,B, $ 和 $ C $ 使得

$$\cfrac 1{s^3-22s^2+80s-67}=\cfrac A{s-p}+\cfrac B{s-q}+\cfrac C{s-r}$$

对所有 $s\not\in \{p, q, r\}$ 都成立。那么 $\cfrac 1A+\cfrac 1B +\cfrac 1C$ 的值为:

(A) 243
(B) 244
(C) 245
(D) 245
(E) 247

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

 Author| 业余的业余 Posted 2019-2-19 22:54
通分,待定系数,和韦达定理,得到如下六个方程

$A+B+C=0$
$A(q+r)+B(p+r)+C(p+q)=0$
$Aqr+Bpr+Cpq=1$
$p+q+r=22$
$pq+pr+pr=80$
$pqr=67$

应该不是叫直接把 $A,B,C$ 解出来吧?需要强大的代数变形能力{:curse:}

686

Threads

110K

Posts

910K

Credits

Credits
91229
QQ

Show all posts

kuing Posted 2019-2-19 23:04
依题意 `s^3-22s^2+80s-67=(s-p)(s-q)(s-r)`,所以等式去分母后为\[A(s-q)(s-r)+B(s-p)(s-r)+C(s-p)(s-q)=1,\]这时是对任意 `s` 成立,故此分别令 `s=p`, `q`, `r` 可得\[A=\frac1{(p-q)(p-r)},B=\frac1{(q-p)(q-r)},C=\frac1{(r-p)(r-q)},\]所以所求式为\[(p-q)(p-r)+(q-p)(q-r)+(r-p)(r-q),\]亦即\[(p+q+r)^2-3(pq+qr+rp),\]代韦达定理 `p+q+r=22`, `pq+qr+rp=80` 即得结果为 `244`。

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

 Author| 业余的业余 Posted 2019-2-19 23:07
有点意思。

从前三个方程,把 $A,B,C$ 作为未知数,解得

$A=\cfrac 1{qr-qp+p^2-pr}$

即 $\cfrac 1A=qr-qp+p^2-pr$, 由对称性,有

$\cfrac 1B=pr-rq+q^2-qp$

$\cfrac 1C=qp-pr+r^2-rq$

三式相加,得$\cfrac 1A+\cfrac 1B+\cfrac 1C=p^2+q^2+r^2-pq-qr-rp=(p+q+r)^2-3(pq+qr+pr)=22^2-3\cdot 80=484-240=244$

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

 Author| 业余的业余 Posted 2019-2-19 23:08
回复 3# kuing

太牛了。 这也是套路题吧?

你的办法,省去了成吨的计算,直指要害,佩服。

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

 Author| 业余的业余 Posted 2019-2-19 23:17
题目说的其实是 $s\notin\{p,q,r\}$, 但是 $s=p,q,r$ 是极限情形,取了某种意义上的倒数后,就是可去间断点了。总之这个办法快准狠,学习了。

770

Threads

4692

Posts

310K

Credits

Credits
35048

Show all posts

isee Posted 2019-2-19 23:21
印象中论坛里某帖有一道类似题

686

Threads

110K

Posts

910K

Credits

Credits
91229
QQ

Show all posts

kuing Posted 2019-2-20 13:19
题目说的其实是 $s\notin\{p,q,r\}$, 但是 $s=p,q,r$ 是极限情形,取了某种意义上的倒数后,就是可去间断点 ...
业余的业余 发表于 2019-2-19 23:17
如果想避免解释可去间断点啥的,可以用拉格朗日插值公式。

设 `x_i`(`i=1`, `2`, `3`)互不相同,设常数函数 `f(x)=1`,由拉格朗日插值公式,`f(x)` 可以表示成
\[f(x)=\frac{(x-x_2)(x-x_3)}{(x_1-x_2)(x_1-x_3)}f(x_1)+\frac{(x-x_1)(x-x_3)}{(x_2-x_1)(x_2-x_3)}f(x_2)+\frac{(x-x_1)(x-x_2)}{(x_3-x_1)(x_3-x_2)}f(x_3),\]
故此有恒等式
\[1=\frac{(x-x_2)(x-x_3)}{(x_1-x_2)(x_1-x_3)}+\frac{(x-x_1)(x-x_3)}{(x_2-x_1)(x_2-x_3)}+\frac{(x-x_1)(x-x_2)}{(x_3-x_1)(x_3-x_2)},\quad(*)\]
当 `(x-x_1)(x-x_2)(x-x_3)\ne0` 时,两边除以它就是
\begin{align*}
&\frac1{(x-x_1)(x-x_2)(x-x_3)}\\
={}&\frac1{(x-x_1)(x_1-x_2)(x_1-x_3)}+\frac1{(x-x_2)(x_2-x_1)(x_2-x_3)}+\frac1{(x-x_3)(x_3-x_1)(x_3-x_2)}.
\end{align*}

显然这对于多元也是一样的,而且这里 `x`, `x_i` 都不一定要是实数。

686

Threads

110K

Posts

910K

Credits

Credits
91229
QQ

Show all posts

kuing Posted 2019-2-20 13:25
回复 8# kuing

恒等式 (*),正好就是昨天那个动点不等式里的引理二的证法,等会补充到那边。(本来昨天已经懒得写,结果还是写了……

46

Threads

323

Posts

2834

Credits

Credits
2834

Show all posts

 Author| 业余的业余 Posted 2019-2-20 22:56
回复 9# kuing

其实没有关系。这是确定待定系数的简洁方法。做积分题时,partial fraction  (比如把 $\cfrac {1}{(x-a)(x-b)(x-c)}$ 分解为 $\cfrac {A}{x-a}+\cdots$ 的技术) 会用到这个技巧,这样的 $A,B,C$ 必定存在,怎么简单怎么来。你的办法真没有问题的。

3159

Threads

7941

Posts

610K

Credits

Credits
63770
QQ

Show all posts

hbghlyj Posted 2023-1-22 01:17
Last edited by hbghlyj 2023-1-31 21:31
业余的业余 发表于 2019-2-20 15:56
回复 9# kuing

其实没有关系。这是确定待定系数的简洁方法。做积分题时,partial fraction  (比如把 $\cfrac {1}{(x-a)(x-b)(x-c)}$ ...
相关
PlanetMath–Heaviside formula

Let $P(s)$ and $Q(s)$ be polynomials with the degree of the former less than the degree of the latter.
  • If all complex zeroes $a_1,a_2,\ldots,a_n$ of $Q(s)$ are simple, then
    \begin{align}
    \mathcal{L}^{-1}\left\{\frac{P(s)}{Q(s)}\right\} = \sum_{j=1}^n\frac{P(a_j)}{Q'(a_j)}e^{a_jt}.
    \end{align}
  • If the different zeroes $a_1,a_2,\ldots,a_n$ of $Q(s)$ have the multiplicities $m_1,m_2,\ldots,m_n$, respectively,
    we denote $F_j(s) := (s-a_j)^{m_j}P(s)/Q(s)$; then
    \begin{align}\label2
    \mathcal{L}^{-1}\left\{\frac{P(s)}{Q(s)}\right\} =
    \sum_{j=1}^ne^{a_jt}\sum_{k=0}^{m_j-1}\frac{F_j^{(k)}(a_j)\;t^{m_j-1-k}}{k!(m_j-1-k)!}.
    \end{align}

A special case of the Heaviside formula (1) is
\begin{align}
\mathcal{L}^{-1}\left\{\frac{Q'(s)}{Q(s)}\right\}
= \sum_{j=1}^ne^{a_jt}.
\end{align}

Example. Since the zeros of the binomial $s^4+4a^4$ are $s = (\pm1\pm i)a$, we can calculate by (3) as follows:
$$\mathcal{L}^{-1}\left\{\frac{s^3}{s^4+4a^4}\right\}
= \frac{1}{4}\mathcal{L}^{-1}\left\{\frac{4s^3}{s^4+4a^4}\right\} = \frac{1}{4}\sum_\pm e^{(\pm 1\pm i)at}
= \frac{e^{at}+e^{-at}}{2}\cdot\frac{e^{iat}+e^{-iat}}{2}
= \cosh{at}\cos{at}$$
Proof of (1). Without hurting the generality, we can suppose that $Q(s)$ is monic. Therefore
$$Q(s) = (s-a_1)(s-a_2)\cdots(s-s_n).$$
For $j = 1,2,\ldots,n$, denoting
$$Q(s) := (s-a_j)Q_j(s),$$
one has $Q_j(a_j) \neq 0$. We have a partial fraction expansion of the form
\begin{align}
\frac{P(s)}{Q(s)} = \frac{C_1}{s-a_1}+\frac{C_2}{s-a_2}+\ldots+\frac{C_n}{s-a_n}
\end{align}
with constants $C_j$. According to the linearity and the formula 1 of Laplace Transform, one gets
\begin{align}
\mathcal{L}^{-1}\left\{\frac{P(s)}{Q(s)}\right\}
= \sum_{j=1}^nC_je^{a_jt}.
\end{align}
For determining the constants $C_j$, multiply (3) by $s-a_j$. It yields
$$\frac{P(s)}{Q_j(s)}
= C_j+(s-a_j)\sum_{\nu \neq j}\frac{C_\nu}{s-a_\nu}.$$
Setting to this identity $s := a_j$ gives the value
\begin{align}
C_j = \frac{P(a_j)}{Q_j(a_j)}.
\end{align}
But since $Q'(s) = \frac{d}{ds}((s-a_j)Q_j(s))
= Q_j(s)+(s-a_j)Q_j'(s)$,
we see that $Q'(a_j) = Q_j(a_j)$; thus the equation (6) may be written
\begin{align}
C_j = \frac{P(a_j)}{Q'(a_j)}.
\end{align}
The values (6) in (4) produce the formula (1).

Kreyszig E - Advanced Engineering Mathematics - Wiley 2006 9Ed (工科数学) page 247 Problem 15
untitled.png 15. PROJECT. Heaviside Formulas. (a) Show that for a simple root $a$ and fraction $A /(s-a)$ in $F(s) / G(s)$ we have the Heaviside formula
$$
A=\lim _{s \rightarrow a} \frac{(s-a) F(s)}{G(s)} .
$$
(b) Similarly, show that for a root $a$ of order $m$ and fractions in
$$
\frac{F(s)}{G(s)} =\frac{A_m}{(s-a)^m}+\frac{A_{m-1}}{(s-a)^{m-1}}+\cdots  +\frac{A_1}{s-a}+\text {further fractions}
$$
we have the Heaviside formulas for the first coefficient
$$
A_m=\lim _{s \to a} \frac{(s-a)^m F(s)}{G(s)}
$$
and for the other coefficients
$$
A_k=\frac{1}{(m-k) !} \lim _{s \to a} \frac{d^{m-k}}{d s^{m-k}}\left[\frac{(s-a)^m F(s)}{G(s)}\right], \quad
k=1, \cdots, m-1 .
$$
Heaviside cover-up method
Brilliant wiki
cover-up即 “把分母中的0覆盖”, 在h.pdf把分母中的0填充\phantom, 节录如下:$$\frac{x-7}{(x-1)(x+2)}=\frac{A}{x-1}+\frac{B}{x+2}$$
To determine $A$ by the cover-up method, on the left-hand side we mentally remove (or cover up with a finger) the factor $x - 1 $ associated with $A$, and substitute $x = 1$ into what’s left; this gives $A$:
为了通过覆盖法确定$A$,在左侧,我们在脑中删除(或用手指覆盖)与$A$相关的因子$x - 1$,并将$x = 1$代入剩下的; 这给出 $A$:
$$\left.\frac{x-7}{\hphantom{(x-1)}(x+2)}\right|_{x=1}=\frac{1-7}{1+2}=-2=A$$而在HeavisideCoverup2008.pdf用\box填充. 在第4页介绍了如何对重根使用此方法:
Extension to Multiple Roots. Heaviside’s method can be extended to the case of repeated roots. The basic idea is to factor–out the repeats. To illustrate, consider the partial fraction expansion details
\begin{array}{rlrl}R & =\frac{1}{(s+1)^{2}(s+2)} & & \text {A sample rational function having repeated roots.} \\ & =\frac{1}{s+1}\left(\frac{1}{(s+1)(s+2)}\right) & & \text {Factor–out the repeats.} \\ & =\frac{1}{s+1}\left(\frac{1}{s+1}+\frac{-1}{s+2}\right) & & \text {Apply the cover–up method to the simple root fraction} \\ & =\frac{1}{(s+1)^{2}}+\frac{-1}{(s+1)(s+2)} & & \text {Multiply.} \\ & =\frac{1}{(s+1)^{2}}+\frac{-1}{s+1}+\frac{1}{s+2} & & \text {Apply the cover–up method to the last fraction on the right.}\end{array}Special Methods. Heaviside’s method has a useful extension for the case of roots of multiplicity two. To illustrate, consider these details:
\begin{array}{rll} R & =\frac{1}{(s+1)^{2}(s+2)} &\boxed1\text{A fraction with multiple roots.}\\ & =\frac{A}{s+1}+\frac{B}{(s+1)^{2}}+\frac{C}{s+2} &\boxed2\text{ See equation (5), page 2.}\\ & =\frac{A}{s+1}+\frac{1}{(s+1)^{2}}+\frac{1}{s+2}&\boxed3\text{ Find $B$ and $C$ by Heaviside’s cover–up method.} \\ & =\frac{-1}{s+1}+\frac{1}{(s+1)^{2}}+\frac{1}{s+2}&\boxed4\text{ Details below.}\end{array}We discuss $\boxed4$ details. Multiply the equation $\boxed1 =\boxed 2$ by $s + 1$ to partially clear fractions, the same step as the cover-up method:$$\frac{1}{(s+1)(s+2)}=A+\frac{B}{s+1}+\frac{C(s+1)}{s+2}$$We don't substitute $s+1=0$, because it gives infinity for the second term. Instead, set $s=\infty$ to get the equation $0=A+C$. Because $C=1$ from $\boxed3$, then $A=-1$.

The illustration works for one root of multiplicity two, because $s=\infty$ will resolve the coefficient not found by the cover-up method.

In general, if the denominator in (1) has a root $s_0$ of multiplicity $k$, then the partial fraction expansion contains terms
$$
\frac{A_1}{s-s_0}+\frac{A_2}{\left(s-s_0\right)^2}+\cdots+\frac{A_k}{\left(s-s_0\right)^k}
$$
Heaviside's cover-up method directly finds $A_k$, but not $A_1$ to $A_{k-1}$.

Cover-up Method and Complex Numbers. Consider the partial fraction expansion
$$
\frac{10}{(s+1)\left(s^2+9\right)}=\frac{A}{s+1}+\frac{B s+C}{s^2+9} .
$$
The symbols $A, B, C$ are real. The value of $A$ can be found directly by the coverup method, giving $A=1$. To find $B$ and $C$, multiply the fraction expansion by $s^2+9$, in order to partially clear fractions, then formally set $s^2+9=0$ to obtain the two equations
$$
\frac{10}{s+1}=B s+C, \quad s^2+9=0 .
$$
The method applies the identical idea used for one real root. By clearing fractions in the first, the equations become
$$
10=B s^2+C s+B s+C, \quad s^2+9=0 .
$$
Substitute $s^2=-9$ into the first equation to give the linear equation
$$
10=(-9 B+C)+(B+C) s .
$$
Because this linear equation has two complex roots $s=\pm 3 i$, then real constants $B, C$ satisfy the $2 \times 2$ system
$$
\begin{aligned}
-9 B+C & =10 \\
B+C & =0 .
\end{aligned}
$$
Solving gives $B=-1, C=1$.
The same method applies especially to fractions with 3-term denominators, like $s^2+s+1$. The only change made in the details is the replacement $s^2 \rightarrow$ $-s-1$. By repeated application of $s^2=-s-1$, the first equation can be distilled into one linear equation in $s$ with two roots. As before, a $2 \times 2$ system results.

3159

Threads

7941

Posts

610K

Credits

Credits
63770
QQ

Show all posts

hbghlyj Posted 2023-1-22 02:10
\eqref{2}如何证明呢

Mobile version|Discuz Math Forum

2025-5-31 10:33 GMT+8

Powered by Discuz!

× Quick Reply To Top Edit