离散及分布型时滞微分方程的数值正则性.pdfVIP

摘 要 数值算法的动力学特征一直被众多学者所关注,它包括很多内容,如收敛性、稳定 性、耗散性、正则性、混沌、分叉等等. 本文在已有结果的基础上,着重研究一般线 性方法求解带离散与分布型时滞微分方程时的正则性问题. 1990年,Hairer 首次提出了数值方法的正则性. 随后,数值方法的正则性问题逐渐 成为众多文献讨论的主题之一. 这些文献主要讨论了混合方法,Runge-Kutta方法,一般 线性方法及单支方法用于求解时滞微分方程时的正则性. 本文在已有结果的基础上 讨论一般线性方法用于求解带离散与分布型时滞微分方程时的正则性. 文章结构大致如下：第一章中我们介绍正则性研究的背景、意义以及该领域的 主要研究成果. 第二章中,我们讨论一般线性方法求解离散型时滞微分方程时的正则 性问题。对于带离散时滞的微分方程,我们证明了essentially one-stage (EOS)方法及可 折叠成EOS 方法的正则性,并给出了非EOS 方法求解离散型时滞微分方程时满足正 则性与强正则性要求的充分必要条件. 第三章重点研究一般线性方法求解分布型时 滞微分方程时的正则性. 和第二章中得到的结论类似,我们证明了essentially one-stage (EOS)方法及可折叠成EOS 方法用于求解分布型时滞微分方程时的正则性,并给出了 非EOS 方法满足正则性与强正则性要求的充分必要条件. 最后,第四章中我们用数值 实验验证了本文的主要结果. 关键词：正则性，离散型时滞微分方程，分布型时滞微分方程，一般线性方法， EOS方法 I Abstract For numerical methods, there has been an increasing interest in the investigation of their dynamical properties, such as convergence, stability, dissipativity, regularity, chaos and bifurcation. This paper focus on the regularity of general linear methods for solving ODEs with discrete and distributed time delays. In 1990, Hairer firstly proposed the concept of regularity. Since then, regularity has been widely researched by many authors in the field of numerical methods. The existing results are mainly concerned with hybrid methods, Runge-Kutta methods and general linear methods for solving ODEs without time delays, and one-leg methods, Runge-Kutta methods for solving delay differential equations (DDEs). In this paper, we analyze the regularity of general linear methods for solving ODEs with discrete and distributed time delays. This paper is organized as follows. In chapter 1, we introduce the background, some excellent results of regularity of numerical methods. The regularity of general linear methods for solving ODEs with discrete delays i