压缩映射
定义4.1 压缩映射
{∣φ(x2)−φ(x1)∣=L∣x2−x1∣L<1⇒φ(x)为压缩映射.\left\{\begin{matrix} \left | \varphi (x_2) - \varphi (x_1) \right | = L\left | x_2 - x_1 \right | \\ L < 1 \end{matrix}\right. \Rightarrow \varphi (x)为压缩映射.{∣φ(x2)−φ(x1)∣=L∣x2−x1∣L<1⇒φ(x)为压缩映射.
性质
- 若φ(x)\varphi (x)φ(x)为压缩映射 ⇒φ(x)连续\Rightarrow \varphi(x) 连续⇒φ(x)连续
- {φ(x)连续∣φ′(x)∣⩽L<1⇒φ(x)为压缩映射.\left\{\begin{matrix} \varphi(x) 连续\\ \left | \varphi'(x) \right | \leqslant L < 1 \end{matrix}\right. \Rightarrow \varphi (x)为压缩映射.{φ(x)连续∣φ′(x)∣⩽L<1⇒φ(x)为压缩映射.
收敛
收敛条件
{φ(x)∈C1[a,b]∣φ′(x)∣⩽L<1a⩽L⩽b⇒φ(x)收敛于唯一根α,α∈[a,b]\left\{\begin{matrix} \varphi (x) \in C^1 \left [ a, b \right ] \\ \left | \varphi'(x) \right | \leqslant L < 1 \\ a \leqslant L \leqslant b \end{matrix}\right. \Rightarrow \varphi (x)收敛于唯一根\alpha,\alpha \in \left [ a, b \right ]⎩⎨⎧φ(x)∈C1[a,b]∣φ′(x)∣⩽L<1a⩽L⩽b⇒φ(x)收敛于唯一根α,α∈[a,b]
ppp阶收敛
定义4.2 迭代法p阶收敛
limk→∞∣ek+1∣∣ek∣p=C≠0\lim_{k\rightarrow \infty} \frac{\left | e_{k+1} \right |} {\left | e_k \right |^p}=C \neq 0k→∞lim∣ek∣p∣ek+1∣=C=0
或
∣xk+1−α∣≈C∣xk−α∣p\left | x_{k+1}-\alpha \right | \approx C \left | x_k - \alpha \right |^p∣xk+1−α∣≈C∣xk−α∣p
这里ek=xk−αe_k = x_k -\alphaek=xk−α.则称迭代法是p阶收敛的.
定理4.1
{φ(x)在α邻域内充分光滑φ(α)=αφ′(α)=φ′′(α)=...=φ(p−1)(α)=0φ(p)(α)≠0p⩾2⇒limk→∞∣ek+1∣∣ek∣p=1p!∣φ(p)(α)∣≠0 \left\{\begin{matrix} \varphi (x)在\alpha 邻域内充分光滑\\ \varphi (\alpha)=\alpha\\ \varphi'(\alpha)=\varphi''(\alpha)=...=\varphi^{(p-1)}(\alpha)=0\\ \varphi^{(p)}(\alpha) \neq0 \\ p \geqslant 2 \end{matrix}\right. \Rightarrow \lim_{k\rightarrow \infty} \frac{\left | e_{k+1} \right |} {\left | e_k \right |^p}= \frac{1}{p!} \left | \varphi^{(p)}(\alpha) \right | \neq 0⎩⎨⎧φ(x)在α邻域内充分光滑φ(α)=αφ′(α)=φ′′(α)=...=φ(p−1)(α)=0φ(p)(α)=0p⩾2⇒k→∞lim∣ek∣p∣ek+1∣=p!1φ(p)(α)=0
Newton方法
切线法
形式
xk+1=xk−f(xk)f′(xk)x_{k+1}=x_k - \frac{f(x_k)} {f'(x_k)}xk+1=xk−f′(xk)f(xk)
收敛
当∣x0−α∣<2m1M2\left | x_0 - \alpha \right | < \frac{2m_1}{M_2}∣x0−α∣<M22m1时切线法收敛,其中m1m_1m1为f′(x)f'(x)f′(x)的最小值,M2M_2M2为f′′(x)f''(x)f′′(x)的最大值。切线法为一阶收敛(或线性收敛),即p=1p=1p=1.
简单Newton法
形式
xk+1=xk−f(xk)M,M=f′(x0)x_{k+1}=x_k - \frac{f(x_k)} {M},M=f'(x_0)xk+1=xk−Mf(xk),M=f′(x0)
收敛
简单Newton法为线性收敛,即p=1p=1p=1.
割线法
形式
xk+1=xk−f(xk)f(xk)−f(xk−1)xk−xk−1x_{k+1}=x_k - \frac{f(x_k)} {\frac{f(x_k)-f(x_{k-1})}{x_k-x_{k-1}}}xk+1=xk−xk−xk−1f(xk)−f(xk−1)f(xk)
收敛
limk→∞ek+1ekek−1=f′′(α)2f′(α)\lim_{k\rightarrow \infty} \frac{e_{k+1}}{e_k e_{k-1}}=\frac{f''(\alpha)}{2f'(\alpha)}k→∞limekek−1ek+1=2f′(α)f′′(α)
割线法收敛阶为p=1+52p=\frac{1+\sqrt 5}{2}p=21+5.
带参mmm重根
因mmm重根,故设f(x)=(x−α)mh(x)f(x)=(x-\alpha)^mh(x)f(x)=(x−α)mh(x),对f(x)f(x)f(x)求1m\frac{1}{m}m1次幂有[f(x)]1m=(x−α)[h(x)]1m\left [ f(x)\right ]^{\frac{1}{m}}=(x-\alpha) \left[ h(x) \right ]^{\frac{1}{m}}[f(x)]m1=(x−α)[h(x)]m1.变成了单根。因此,便有了以下的迭代公式:
形式
xk+1=xk−[f(x)]1m([f(x)]1m)′=xk−mf(xk)f′(xk)x_{k+1}=x_k - \frac{\left [ f(x)\right ]^{\frac{1}{m}}} {\left ( \left [ f(x)\right ]^{\frac{1}{m}} \right )'}=x_k-m\frac{f(x_k)}{f'(x_k)}xk+1=xk−([f(x)]m1)′[f(x)]m1=xk−mf′(xk)f(xk)
收敛
该方法为二阶收敛(或平方收敛),即p=2p=2p=2.
无参mmm重根
设辅助函数u(x)=f(x)f′(x)=(x−α)mh(x)[(x−α)mh(x)]′=(x−α)hˉ(x)u(x)=\frac{f(x)}{f'(x)}=\frac{(x-\alpha)^mh(x)}{\left [ (x-\alpha)^mh(x) \right ]'}=(x-\alpha) \bar h(x)u(x)=f′(x)f(x)=[(x−α)mh(x)]′(x−α)mh(x)=(x−α)hˉ(x).
形式xk+1=xk−u(x)u′(x)x_{k+1}=x_k-\frac{u(x)}{u'(x)}xk+1=xk−u′(x)u(x)
收敛
该方法为二阶收敛(或平方收敛),即p=2p=2p=2.
本文介绍了压缩映射的概念及其性质,并探讨了压缩映射与函数连续性的关系。此外,还详细阐述了几种数值迭代法,包括Newton方法、割线法及针对多重根的改进方法,分析了它们的收敛性和收敛阶。
2337

被折叠的 条评论
为什么被折叠?



