高数笔记 P08:无穷级数

1 定义

  • 无穷级数:\(\displaystyle \sum_{n = 1}^\infty u_n = u_1 + u_2 + \cdots + u_n + \cdots\)
  • 部分和数列\(\{ S_n\}\),其中\(\displaystyle S_n = \sum_{ n= 1}^n u_n\)
  • 无穷级数的和:\(\displaystyle S = \lim_{n \to \infty} S_n\),若\(S\)存在,则无穷级数收敛;\(S\)不存在,则无穷级数发散。
  • 余部\(r_n\),若无穷级数收敛,\(\displaystyle \lim_{n \to \infty} r_n = 0\)
  • 绝对收敛:\(\displaystyle \sum_{n = 1}^\infty |u_n|\)\(\displaystyle \sum_{n = 1}^\infty u_n\)都收敛;条件收敛:\(\displaystyle \sum_{n = 1}^\infty u_n\)收敛,而\(\displaystyle \sum_{n = 1}^\infty |u_n|\)发散。

2 性质

  • 若级数\(\displaystyle \sum_{n = 1}^\infty u_n\)收敛于和\(S\),则级数\(\displaystyle \sum_{n = 1}^\infty ku_n\)收敛于和\(kS\)
  • 若级数\(\displaystyle \sum_{n = 1}^\infty u_n, \displaystyle \sum_{n = 1}^\infty v_n\)分别收敛于和\(S, \sigma\),则级数\(\displaystyle \sum_{n = 1}^\infty (u_n \pm v_n)\)收敛于\(S \pm \sigma\)
  • 在级数中去掉、添加、改变有限项,级数的收敛性不变。
  • 如果级数收敛,则对这个级数中的项任意添加括号形成的新级数仍收敛。
  • 若级数\(\displaystyle \sum_{n = 1}^\infty u_n\)收敛,则其一般项趋向于零。\(\displaystyle \lim_{n \to \infty} u_n = 0\).
  • 绝对收敛的级数一定收敛。条件收敛的级数的所有正项(或负项)组成的级数一定发散。
  • 补充:柯西审敛原理 级数\(\displaystyle \sum_{n = 1}^\infty u_n\)收敛的充要条件为对于任意给定的正数\(\varepsilon\),总存在正整数\(N\),使得当\(n \gt N\)时,对于任意的正整数\(p\),都有\(|u_{n+1} + u_{n+2} + \cdots + u_{n+p}| \lt \varepsilon\)成立。
  • 补充:设级数\(\displaystyle \sum_{n = 1}^\infty u_n, \sum_{n = 1}^\infty v_n\)都绝对收敛,其和分别是\(S, \sigma\),则它们的柯西乘积 \(u_1 v_1 + (u_1 v_2 + u_2 v_1) + \cdots + (u_1 v_n + u_2 v_{n-1} + \cdots + u_n v_1) + \cdots\) 也绝对收敛,且其和为 \(S\sigma\)

3 常数项级数

正项级数

  • 正项级数\(\displaystyle \sum_{n = 1}^\infty u_n\)收敛的充要条件是它的部分和数列\(\{S_n\}\)有界。

  • 比较审敛法:若\(0 \le u_n \le v_n\),则\(\displaystyle \sum_{n = 1}^\infty v_n\)收敛,\(\implies \displaystyle \sum_{n = 1}^\infty u_n\)收敛;\(\displaystyle \sum_{n = 1}^\infty u_n\)发散,\(\implies \displaystyle \sum_{n = 1}^\infty v_n\)发散。

    \(\displaystyle \lim_{n \to \infty} \cfrac {u_n}{v_n} = l \quad (0 \le l \le +\infty)\),(1)\(0 \lt l \lt +\infty\),两级数同敛散性;(2)\(l = 0\),则\(\displaystyle \sum_{n = 1}^\infty v_n\)收敛,\(\implies \displaystyle \sum_{n = 1}^\infty u_n\)收敛;(3)\(l = +\infty\),则\(\displaystyle \sum_{n = 1}^\infty u_n\)发散,\(\implies \displaystyle \sum_{n = 1}^\infty v_n\)发散。

  • 比值审敛法:\(\displaystyle \lim_{n \to \infty} \cfrac {u_{n+1}}{u_n} = \rho\)\(\rho \lt 1\),收敛;\(\rho \gt 1\),发散;\(\rho = 1\),此方法失效。

  • 根值审敛法:\(\displaystyle \lim_{n \to \infty} \sqrt[n]{u_n} = \rho\)\(\rho \lt 1\),收敛;\(\rho \gt 1\),发散;\(\rho = 1\),此方法失效。

  • 极限审敛法:若\(\displaystyle \lim_{n \to \infty} nu_m = l \gt 0\),则级数发散;若\(p \gt 1\)\(\displaystyle \lim_{n \to \infty} n^p u_n = l \ \ (0 \le l \lt +\infty)\),则级数收敛。

  • 对数判别法:\(\displaystyle \lim_{n \to \infty} \cfrac {\ln {\cfrac 1{u_n}}}{\ln n} = p\)\(p \gt 1\),收敛;\(p \lt 1\),发散;\(p = 1\),此方法失效。

交错级数

莱布尼茨定理:如果级数\(\displaystyle \sum_{n = 1}^{\infty} (-1)^{n-1} u_n\)满足:\(u_n \ge u_{n+1}\)\(\displaystyle \lim_{n \to \infty} u_n = 0\),则级数收敛,且其和\(S \lt u_1\)

4 幂级数

概念:函数项级数、收敛点、和函数

定义

  • 幂级数\(\displaystyle \sum_{n = 0}^\infty a_n x^n = a_0 + a_1 x + a_2 x^2 + \cdots + a_n x^n + \cdots\)
  • 若级数\(\displaystyle \sum_{n = 0}^\infty a_n x^n\)\(x = x_0(x_0 \neq 0)\)时收敛,那么适合不等式\(|x| \lt |x_0|\)的一切\(x\)使这个幂级数绝对收敛。反之,若级数\(\displaystyle \sum_{n = 0}^\infty a_n x^n\)\(x = x_0(x_0 \neq 0)\)时发散,那么适合不等式\(|x| \gt |x_0|\)的一切\(x\)使这个幂级数发散。
  • 幂级数收敛的三种情况:
    • 仅在\(x = 0\)处收敛,\(x \neq 0\)时发散;
    • 对于\(x \in (-\infty, +\infty)\)都收敛且绝对收敛;
    • 对于\(x \in (-R, +R)\)时收敛且绝对收敛,之外的发散。\(R\)收敛半径
  • 一个幂级数的收敛半径总存在,包含\(0, +\infty\)
  • \(\displaystyle \lim_{n \to \infty} \left| \cfrac {a_{n+1}}{a_n} \right| = \rho\),那么幂级数的收敛半径\(R = \begin{cases} \cfrac 1\rho & \rho \neq 0 \\ +\infty & \rho = 0 \\ 0 & \rho = +\infty \\ \end{cases}\)
  • \(\displaystyle \lim_{n \to \infty} \sqrt[n]{\left| a_n \right|} = \rho\),那么幂级数的收敛半径\(R = \begin{cases} \cfrac 1\rho & \rho \neq 0 \\ +\infty & \rho = 0 \\ 0 & \rho = +\infty \\ \end{cases}\)

性质

  • \(\displaystyle \sum_{n = 0}^\infty a_n x^n \pm \sum_{n = 0}^\infty b_n x^n = \sum_{n = 0}^\infty (a_n \pm b_n)x^n = S_1(x) \pm S_2(x) \quad x\in (-R, +R)\)
  • \(\displaystyle (\sum_{n = 0}^\infty a_n x^n)(\sum_{n = 0}^\infty b_n x^n) = \sum_{n = 0}^\infty (a_0 b_n + a_1 b_{n - 1} + \cdots + a_n b_0)x^n = S_1(x) S_2(x) \quad x\in (-R, +R)\)
  • \(\displaystyle \cfrac {S_1(x)}{S_2(x)} = \cfrac {\sum_{n = 0}^\infty a_n x^n}{\sum_{n = 0}^\infty b_n x^n} = c_0 + c_1 x + \cdots + c_n x^n + \cdots\)
  • 和函数\(S(x)\)在收敛域上连续;
  • 和函数\(S(x)\)在收敛域上可导,且可逐项求导,\(S'(x) = \displaystyle \sum_{n = 0}^\infty n a_n x^{n-1}\)
  • 和函数\(S(x)\)在收敛域上可积,且可逐项积分,\(\displaystyle \int_0^x S(x)dx = \sum_{n = 0}^\infty \int_0^x a_n x^n dx = \sum_{n = 0}^\infty \cfrac {a_n}{n+1} x^{n+1}\)

函数展开成幂级数

  • 泰勒级数:\(\displaystyle \sum_{n = 0}^\infty \cfrac {f^{(n)}(x_0)}{n!} (x - x_0)^n\)
  • 麦克劳林级数:\(\displaystyle \sum_{n = 0}^n \cfrac {f^{(n)}(0)}{n!} x^n\)
  • 泰勒级数的收敛定理:设\(f(x)\)\(x = x_0\)处任意阶可导,则泰勒级数在\(|x - x_0| \lt R\)内收敛于\(f(x)\)的充要条件是\(\displaystyle \lim_{n \to \infty} R_n(x) = 0\),其中\(R_n(x) = \cfrac {f^{(n+1)}(x_0 + \theta (x - x_0))}{(n+1)!} (x - x_0)^{n+1}\)
  • 常用的麦克劳林公式
    • \(\cfrac 1{1-x} = 1 + x + x^2 + \cdots + x^n + \cdots \quad x \in (-1, 1)\)
    • \(\cfrac 1{1+x} = 1 - x + x^2 + \cdots + (-1)^n x^n + \cdots \quad x \in (-1, 1)\)
    • \(e^x = 1 + x + \cfrac {x^2}{2!} + \cdots + \cfrac {x^n}{n!} + \cdots \quad x \in (-\infty, +\infty)\)
    • \(\sin x = x - \cfrac {x^3}{3!} + \cdots + \cfrac {(-1)^n x^{2n+1}}{(2n+1)!} + \cdots \quad x \in (-\infty, +\infty)\)
    • \(\cos x = 1 - \cfrac {x^2}{2!} + \cdots + \cfrac {(-1)^n x^{2n}}{(2n)!} + \cdots \quad x \in (-\infty, +\infty)\)
    • \(\ln (1+x) = x - \cfrac {x^2}2 + \cdots + \cfrac {(-1)^{n-1} x^n}n + \cdots \quad x \in (-1. 1]\)
    • \((1+x)^\alpha = 1 + \alpha x + \cfrac {\alpha (\alpha -1)}{2!} x^2 + \cdots + \cfrac {\alpha (\alpha - 1)\cdots(\alpha - n + 1)}{n!} x^n + \cdots \quad x \in (-1,1)\)
  • 应用:近似计算、解微分方程

5 三角级数

定义 三角函数系的正交性

  • 三角级数:\(\cfrac {a_0}2 + \displaystyle \sum_{n = 1}^\infty (a_n \cos nx + b_n \sin nx)\)
  • 正交是指在区间积分为零:
    • \(\displaystyle \int_{-\pi}^\pi \cos nx dx = 0, \int_{-\pi}^\pi \sin nx dx = 0, (n = 1, 2, 3, \cdots)\)
    • \(\displaystyle \int_{-\pi}^\pi \sin kx \cos nx dx = 0, (k,n = 1, 2, 3, \cdots)\)
    • \(\displaystyle \int_{-\pi}^\pi \cos kx \cos nx dx = 0, \int_{-\pi}^\pi \sin kx \sin nx dx = 0, (k, n = 1, 2, 3, \cdots ,k \neq n)\)

函数展开为傅里叶级数

  • 傅里叶级数:\(\begin{cases} \displaystyle a_n = \cfrac 1\pi \int_{-\pi}^\pi f(x) \cos nx dx & n = 0, 1, 2, 3, \cdots \\ \displaystyle b_n = \cfrac 1\pi \int_{-\pi}^\pi f(x) \sin nx dx & n = 1, 2, 3, \cdots \\ \end{cases}\)称为\(f(x)\)的傅里叶级数。

    \(f(x) \sim \cfrac {a_0}2 + \displaystyle \sum_{n = 1}^\infty (a_n \cos nx + b_n \sin nx)\)

  • 狄利克雷定理:设\(f(x)\)为周期为\(2\pi\)的周期函数,如果它满足:在一个周期内连续或者只有有限个第一类间断点,在一个周期内至多有有限个极值点,那么\(f(x)\)的傅里叶级数收敛。并且

    • \(x\)是连续点时,级数收敛于\(f(x)\)
    • \(x\)是间断点时,级数收敛域\(\cfrac {f(x^-) + f(x^+)}2\)
  • 对于周期为\(2\pi\)的函数的傅里叶级数展开分两步进行:

    • 计算\(a_n, b_n\)
    • 讨论收敛情况。
  • 对于周期为\(2l\)的函数的傅里叶级数展开分两步进行:

    • 计算展开式:\(f(x) = \cfrac {a_0}2 + \displaystyle \sum_{n = 1}^\infty (a_n \cos \cfrac {n \pi x}l + b_n \sin \cfrac {n \pi x}l), (x \in C)\),其中\(\begin{cases} \displaystyle a_n = \cfrac 1l \int_{-l}^l f(x) \cos \cfrac {n \pi x}l dx & n = 0, 1, 2, \cdots \\ \displaystyle b_n = \cfrac 1l \int_{-l}^l f(x) \sin \cfrac {n \pi x}l dx & n = 1, 2, 3, \cdots \\ C = \left\{ x \mid f(x) = \cfrac 12 [f(x^-) + f(x^+)] \right\} \\ \end{cases}\)
    • 讨论收敛情况。
posted @ 2020-07-27 11:12  ixtwuko  阅读(3351)  评论(0编辑  收藏  举报