<p>This paper deals with numerical computation and analysis for the initial boundary problems of two dimensional (2D) Sobolev equations with piecewise continuous argument. Firstly, a two-level high-order compact difference method (HOCDM) with computational accuracy <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="10473_2025_505_Article_IEq1.gif" Format="GIF" Height="22" Rendition="HTML" Resolution="72" Type="Linedraw" Width="117" /> </InlineMediaObject> <EquationSource Format="TEX">\(\mathcal{O}(\tau^2+h_x^4+h_y^4)\)</EquationSource> </InlineEquation> is suggested, where <i>τ</i>; <i>h</i><sub><i>x</i></sub>; <i>h</i><sub><i>y</i></sub> denote the temporal and spatial stepsizes of the method, respectively. In order to improve the temporal computational accuracy of this method, the Richardson extrapolation technique is used and thus a new two-level HOCDM is derived, which is proved to be convergent of order four both in time and space. Although the new two-level HOCDM has the higher computational accuracy in time than the previous one, it will bring a larger computational cost. To overcome this deficiency, a three-level HOCDM with computational accuracy <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="10473_2025_505_Article_IEq2.gif" Format="GIF" Height="22" Rendition="HTML" Resolution="72" Type="Linedraw" Width="117" /> </InlineMediaObject> <EquationSource Format="TEX">\(\mathcal{O}(\tau^4+h_x^4+h_y^4)\)</EquationSource> </InlineEquation> is constructed. Finally, with a series of numerical experiments, the theoretical accuracy and computational efficiency of the above methods are further verified.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

High-order compact difference methods for 2D Sobolev equations with piecewise continuous argument

  • Chengjian Zhang,
  • Bo Hou

摘要

This paper deals with numerical computation and analysis for the initial boundary problems of two dimensional (2D) Sobolev equations with piecewise continuous argument. Firstly, a two-level high-order compact difference method (HOCDM) with computational accuracy \(\mathcal{O}(\tau^2+h_x^4+h_y^4)\) is suggested, where τ; hx; hy denote the temporal and spatial stepsizes of the method, respectively. In order to improve the temporal computational accuracy of this method, the Richardson extrapolation technique is used and thus a new two-level HOCDM is derived, which is proved to be convergent of order four both in time and space. Although the new two-level HOCDM has the higher computational accuracy in time than the previous one, it will bring a larger computational cost. To overcome this deficiency, a three-level HOCDM with computational accuracy \(\mathcal{O}(\tau^4+h_x^4+h_y^4)\) is constructed. Finally, with a series of numerical experiments, the theoretical accuracy and computational efficiency of the above methods are further verified.