Scribd
Upload
89 views

Introduction To Numerical Analysis For Engineers: - Initial Value Problems 9.1

This document provides an introduction and overview of numerical methods for solving ordinary differential equations. It discusses initial value problems, including Euler's method, Taylor series methods, and Runge-Kutta methods. Boundary value problems and systems of differential equations are also mentioned. Specific numerical techniques are explained, such as discretization, recursion algorithms, error analysis, and convergence. Matlab code examples are provided to demonstrate implementations of Euler's method and the Runge-Kutta method.

Uploaded by

Mitesh Wakte
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
89 views

Introduction To Numerical Analysis For Engineers: - Initial Value Problems 9.1

This document provides an introduction and overview of numerical methods for solving ordinary differential equations. It discusses initial value problems, including Euler's method, Taylor series methods, and Runge-Kutta methods. Boundary value problems and systems of differential equations are also mentioned. Specific numerical techniques are explained, such as discretization, recursion algorithms, error analysis, and convergence. Matlab code examples are provided to demonstrate implementations of Euler's method and the Runge-Kutta method.

Uploaded by

Mitesh Wakte
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

Introduction to Numerical Analysis for Engineers

Mathews

• Ordinary Differential Equations 9


– Initial Value Problems 9.1
• Euler’s Method 9.2
• Taylor Series Methods 9.4
– Error analysis
• Runge-Kutta Methods 9.5
– Systems of differential equations 9.7
– Boundary Value Problems 9.8
• Shooting method 9.8
• Direct Finite Difference methods 9.9

13.002 Numerical Methods for Engineers Lecture 10


Ordinary Differential Equations
Initial Value Problems
Differential Equation

Linear Differential Equation

Non-Linear Differential Equation a b x

non-linear in y

Linear differential equations can often be solved analytically

Non-linear equations require numerical solution

13.002 Numerical Methods for Engineers Lecture 10


Ordinary Differential Equations
Initial Value Problems
Euler’s Method
Differential Equation

Example

Discretization

Finite Difference (forward)

Recurrence

euler.m

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Taylor Series Methods

Initial Value Problem Truncate series to k terms

Taylor Series

Derivatives

Choose Step Size h

+ Discretization

Recursion Algorithm
Partial Derivatives

with

Local Error

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Taylor Series Methods

General Taylor Series Method Example – Euler’s Method

Example
Error Analysis?

Euler’s Method

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Taylor Series Methods
Derivative Bounds
Error Analysis
Error Estimates and Convergence

Euler’s Method

O(h)
13.002 Numerical Methods for Engineers Lecture 10
Initial Value Problems
Taylor Series Methods
Example – Euler’s Method Error Analysis

Exact solution

Derivative Bounds

Error Bound

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Runge-Kutta Methods

Taylor Series Recursion

Substitute k2 in Runge Kutta

Runge-Kutta Recursion
Match 2nd order Taylor series

Match a,b,DE to match Taylor series amap.

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Runge-Kutta Methods
Initial Value Problem

mid-point

2nd Order Runge-Kutta

Predictor-corrector method
4th Order Runge-Kutta

13.002 Numerical Methods for Engineers Lecture 10


Initial Value Problems
Runge-Kutta Methods
h=1.0;
x=[0:0.1*h:10]; rk.m
Euler’s Method y0=0;
y=0.5*x.^2+y0;
figure(1); hold off
a=plot(x,y,'b'); set(a,'Linewidth',2);
% Euler's method, forward finite difference
xt=[0:h:10]; N=length(xt);
yt=zeros(N,1); yt(1)=y0;
for n=2:N
yt(n)=yt(n-1)+h*xt(n-1);
Recurrence end
hold on; a=plot(xt,yt,'xr'); set(a,'MarkerSize',12);
% Runge Kutta
fxy='x'; f=inline(fxy,'x','y');
[xrk,yrk]=ode45(f,xt,y0);
4th Order Runge-Kutta a=plot(xrk,yrk,'.g'); set(a,'MarkerSize',30);
a=title(['dy/dx = ' fxy ', y_0 = ' num2str(y0)])
set(a,'FontSize',16);
b=legend('Exact',['Euler, h=' num2str(h)],
'Runge-Kutta (Matlab)'); set(b,'FontSize',14);

Matlab ode45 automatically ensures convergence


Matlab inefficient for large problems –> Convergence Analysis
13.002 Numerical Methods for Engineers Lecture 10

You might also like