Semilocal convergence of a sixth-order Jarratt method in Banach spaces (Q634726): Difference between revisions
From MaRDI portal
Latest revision as of 23:10, 9 December 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Semilocal convergence of a sixth-order Jarratt method in Banach spaces |
scientific article |
Statements
Semilocal convergence of a sixth-order Jarratt method in Banach spaces (English)
0 references
16 August 2011
0 references
Let \(X\) and \(Y\) be Banach spaces and \(F:\Omega\subset X\rightarrow Y\) be a three times Fréchet differentiable operator on an open convex subset \(\Omega\), and \(x_0 \in \Omega\) that satisfies C1) \(\left\|\Gamma_0F(x_0)\right\|\leq\eta\), C2) \(\left\|\Gamma_0\right\|\leq\beta\), C3) \(\left\|F''(x)\right\|\leq M, \;x \in \Omega\), C4) \(\left\|F'''(x)\right\|\leq N, \;x \in \Omega\), and C5) there is a positive real number \(L\) such that \(\left\|F'''(x)-F'''(y)\right\| \leq L\left\|x-y\right\|\), \( \forall x, y \in \Omega\), where \( \Gamma_n =F'(x_n)^{-1}\). A number of technical results lead the reader to the main existence and uniqueness theorem. Let \(a_0=M\beta\eta\), \(b_0=N\beta\eta^{2}\), \(c_0=L\beta\eta^{3}\) and \(d_0=h(a_0)\varphi(a_0,b_0,c_0)\) satisfy \(a_0<s\) and \(h(a_0)d_0<1\), where \(s\) is the smallest positive zero of the scalar function \(g(t)t-1\) and \(g,h,\varphi\) are defined by 1) \(g(t)=1+\frac{1}{2}\frac{t}{1-t}+\frac{1}{2}\frac{6-t}{3-2t}\left[\frac{1}{1-t}+\left(1+\frac{1}{2}\frac{t}{1-t}\right)^{2}\right]\), 2) \(h(t)=\frac{1}{1-tg(t)}\), 3) \(\varphi(t,u,v)= \left\{\frac{t}{3-2t}\left(t+\frac{1}{3}u\right)+\frac{1}{9}u+\frac{6-t}{3-2t}\left[\frac{11}{18}u+\frac{t^{2}}{2(1-t)}\right]\right\}\lambda+\frac{1}{2}(\frac{6-t}{3-2t}\lambda)^{2}\) 4) with \(\lambda=\frac{1}{8}\frac{t^{3}}{1-t}+\frac{17}{216}v+\frac{1}{12}\frac{tu}{1-t}\). Let \(\overline{B(x_0,R\eta)} \subset\Omega\) where \(R=\frac{g(a_0)}{1-d_0}\), then starting from \(x_0\), the sequence \((x_n)\) generated by an extension to conditions described by \textit{X. Wang, J. Kou} and \textit{Y. Li} [Appl. Math. Lett. 22, No. 12, 1798--1802 (2009; Zbl 1184.65054)] converges to a solution \(x^{*}\) of \(F(x)=0\) where \(x_n, x^{*}\) belong to \(\overline{B(x_0,R\eta)}\) and \(x^{*}\) is the unique solution of \(F(x)=0\) in \(B(x_0,\frac{2}{M\beta}-R\eta)\cap\Omega\). The result also gives \textit{a priori} estimate. The article finishes with an illustration.
0 references
nonlinear operator equations
0 references
Banach space
0 references
recurrence relation
0 references
semilocal convergence
0 references
Jarratt's method
0 references
numerical example
0 references
0 references