【高等数学】第九章 多元函数微分法及其应用——第九节 二元函数的泰勒公式
上一节:【高等数学】第九章 多元函数微分法及其应用——第八节 多元函数的极值及其求法
总目录:【高等数学】 目录
文章目录
- 1. 二元函数的泰勒公式
1. 二元函数的泰勒公式
- 二元函数f(x,y)f(x, y)f(x,y)在点(x0,y0)(x_0, y_0)(x0,y0)的nnn阶泰勒公式
设z=f(x,y)z = f(x, y)z=f(x,y)在点(x0,y0)(x_0, y_0)(x0,y0)的某一邻域内连续且有(n+1)(n + 1)(n+1)阶连续偏导数,(x0+h,y0+k)(x_0 + h, y_0 + k)(x0+h,y0+k)为此邻域内任一点,则有f(x0+h,y0+k)=f(x0,y0)+(h∂∂x+k∂∂y)f(x0,y0)+12!(h∂∂x+k∂∂y)2f(x0,y0)+⋯+1n!(h∂∂x+k∂∂y)nf(x0,y0)+Rn,Rn=1(n+1)!(h∂∂x+k∂∂y)n+1f(x0+θh,y0+θk)(0<θ<1).\begin{aligned} f(x_0 + h, y_0 + k) &= f(x_0, y_0) + \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right) f(x_0, y_0) + \dfrac{1}{2!} \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^2 f(x_0, y_0) + \cdots + \dfrac{1}{n!} \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^n f(x_0, y_0) + R_n,\\R_n&=\dfrac{1}{(n + 1)!} \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^{n + 1} f(x_0 + \theta h, y_0 + \theta k) \quad (0 < \theta < 1). \end{aligned} f(x0+h,y0+k)Rn=f(x0,y0)+(h∂x∂+k∂y∂)f(x0,y0)+2!1(h∂x∂+k∂y∂)2f(x0,y0)+⋯+n!1(h∂x∂+k∂y∂)nf(x0,y0)+Rn,=(n+1)!1(h∂x∂+k∂y∂)n+1f(x0+θh,y0+θk)(0<θ<1).其中RnR_nRn称为拉格朗日余项
记号(h∂∂x+k∂∂y)f(x0,y0)→hfx(x0,y0)+kfy(x0,y0),\left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right) f(x_0, y_0)\to h f_x(x_0, y_0) + k f_y(x_0, y_0),(h∂x∂+k∂y∂)f(x0,y0)→hfx(x0,y0)+kfy(x0,y0), (h∂∂x+k∂∂y)2f(x0,y0)→h2fxx(x0,y0)+2hkfxy(x0,y0)+k2fyy(x0,y0),\left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^2 f(x_0, y_0)\to h^2 f_{xx}(x_0, y_0) + 2hk f_{xy}(x_0, y_0) + k^2 f_{yy}(x_0, y_0),(h∂x∂+k∂y∂)2f(x0,y0)→h2fxx(x0,y0)+2hkfxy(x0,y0)+k2fyy(x0,y0), (h∂∂x+k∂∂y)mf(x0,y0)→∑p=0mCmphpkm−p∂mf∂xp∂ym−p∣(x0,y0).\left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^m f(x_0, y_0) \to \sum_{p = 0}^m \mathrm{C}_m^p h^p k^{m - p} \left. \dfrac{\partial^m f}{\partial x^p \partial y^{m - p}} \right|_{(x_0, y_0)}.(h∂x∂+k∂y∂)mf(x0,y0)→p=0∑mCmphpkm−p∂xp∂ym−p∂mf(x0,y0).为了利用一元泰勒公式,引入函数Φ(t)=f(x0+ht,y0+kt)(0⩽t⩽1)\varPhi(t) = f(x_0 + ht, y_0 + kt) \quad (0 \leqslant t \leqslant 1)Φ(t)=f(x0+ht,y0+kt)(0⩽t⩽1)
根据多元复合函数的求导法则,得到Φ(t)\varPhi(t)Φ(t)的各阶导数:
Φ′(t)=hfx(x0+ht,y0+kt)+kfy(x0+ht,y0+kt)=(h∂∂x+k∂∂y)f(x0+ht,y0+kt),\begin{aligned} \varPhi'(t) &= h f_x(x_0 + ht, y_0 + kt) + k f_y(x_0 + ht, y_0 + kt) \\ &= \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right) f(x_0 + ht, y_0 + kt), \end{aligned}Φ′(t)=hfx(x0+ht,y0+kt)+kfy(x0+ht,y0+kt)=(h∂x∂+k∂y∂)f(x0+ht,y0+kt),
Φ′′(t)=(h∂∂x+k∂∂y)2f(x0+ht,y0+kt),\begin{aligned} \varPhi''(t) = \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^2 f(x_0 + ht, y_0 + kt), \end{aligned}Φ′′(t)=(h∂x∂+k∂y∂)2f(x0+ht,y0+kt),
Φ(n+1)(t)=(h∂∂x+k∂∂y)n+1f(x0+ht,y0+kt),\begin{aligned} \varPhi^{(n+1)}(t) = \left( h \dfrac{\partial}{\partial x} + k \dfrac{\partial}{\partial y} \right)^{n+1} f(x_0 + ht, y_0 + kt), \end{aligned}Φ(n+1)(t)=(h∂x∂+k∂y∂)n+1f(x0+ht,y0+kt),
根据一元函数的麦克劳林公式,可以推得二元函数的泰勒公式 - 误差估计式
函数的各(n+1)(n + 1)(n+1)阶偏导数都连续,
故它们的绝对值在点(x0,y0)(x_0, y_0)(x0,y0)的某一邻域内都不超过某一正常数MMM.
于是,有下面的误差估计式:∣Rn∣⩽M(n+1)!(∣h∣+∣k∣)n+1=M(n+1)!ρn+1(∣h∣ρ+∣k∣ρ)n+1⩽M(n+1)!(2)n+1ρn+1\begin{aligned} |R_n| &\leqslant \dfrac{M}{(n + 1)!}(|h| + |k|)^{n + 1} = \dfrac{M}{(n + 1)!} \rho^{n + 1} \left( \dfrac{|h|}{\rho} + \dfrac{|k|}{\rho} \right)^{n + 1} \\ &\leqslant \dfrac{M}{(n + 1)!} (\sqrt{2})^{n + 1} \rho^{n + 1} \end{aligned} ∣Rn∣⩽(n+1)!M(∣h∣+∣k∣)n+1=(n+1)!Mρn+1(ρ∣h∣+ρ∣k∣)n+1⩽(n+1)!M(2)n+1ρn+1其中ρ=h2+k2\rho = \sqrt{h^2 + k^2}ρ=h2+k2.
误差∣Rn∣|R_n|∣Rn∣是当ρ→0\rho \to 0ρ→0时比ρn\rho^nρn高阶的无穷小. - 二元函数的拉格朗日中值公式
对于二元函数的nnn阶泰勒公式,令n=0n=0n=0,可得二元函数的拉格朗日中值公式f(x0+h,y0+k)=f(x0,y0)+hfx(x0+θh,y0+θk)+kfy(x0+θh,y0+θk).\begin{aligned} f(x_0 + h, y_0 + k) &= f(x_0, y_0) + h f_x(x_0 + \theta h, y_0 + \theta k) + k f_y(x_0 + \theta h, y_0 + \theta k). \end{aligned} f(x0+h,y0+k)=f(x0,y0)+hfx(x0+θh,y0+θk)+kfy(x0+θh,y0+θk). 如果函数f(x,y)f(x, y)f(x,y)的偏导数fx(x,y)f_x(x, y)fx(x,y)、fy(x,y)f_y(x, y)fy(x,y)在某一区域内都恒等于零,
那么函数f(x,y)f(x, y)f(x,y)在该区域内为一常数.
下一节:
总目录:【高等数学】 目录