《微分方程的一种数值解法》-毕业设计（论文）.docVIP

PAGE 南京邮电大学 毕 业 设 计（论 文） 题 目 微分方程的一种数值解法I 专 业 信息与计算科学 学生姓名 班级学号 B070601 指导教师 指导单位 理学院 日期： 年 月 日至 年 月 日  摘 要 本文主要研究微分方程的各种数值解法。常用的的几种数值解法有时间步进法，频域方法，小波变换法等。Euler法、Rung-Kutta法是典型的时间步进法，在科学研究和工问题中使用较多。该方法的优点是计算简单方便，适用范围较广；该方法的缺点是时间步进法的误差随着计算会累积，所以如果计算初期的迭代步长过大，所产生的误差即使很小，经过积累后也将严重影响计算结果。小波配置方法是近年来发展的一中求解微分方程的数值解法，由于小波具有时域和频域双重局部化特性及多分辨分析的特性，使快速小波配置方法具有误差均匀分布，不存在误差累积，并可以有效的处理局部的奇异性。本文利用Matlab编程对上述几种算法进行数值实现，并对它们进行比较。 关键词：小波变换；微分方程；数值解法；Euler法；Rung-Kutta方法ABSTRACT This task is to study the numerical solution of differential equations. Several numerical method currently used time-stepping method, frequency domain methods, wavelet transform..The Euler method and the Runge-Kutta method inside time-stepping method.In the investigation of the above two methods have been widely used, but there are some drawbacks.This is also the shortcomings of the time-stepping method.As computing,the deviation will accumulate.Therefore, if the calculation of the initial step size is too large, the error generated by even a small, after accumulating the results will also be seriously affected. In this paper,the algorithm of the Euler method and the Runge-Kutta method is introduced.And through the error analysis of practical examples to point out the shortcomings of the time-stepping method.On this basis, the wavelet transform method is described in detail.The error of wavelet calculation does not accumulate, it is not present the drawbacks of time stepping method. Key words:wavelet transform ;differential equations;numerical method;the Euler method ; the Runge-Kutta method 目 录 第一章 绪论·····················································1 1.1 微分方程的数值解法···········································1 1.2 小波分析的发展···············································1 1.3 小波在求解微分方程数值算法中的应用··························2 1.4 小波与常用算法的比较………………·····························2 微分方程的时间步进法··································3 2.1 Euler法·······