Fourier Transform and Shifting in Solving ODEs?

Click For Summary
SUMMARY

The forum discussion focuses on solving the ordinary differential equation (ODE) given by u(x+1,t) - 2u(x,t) + u(x-1,t) = u_t with the initial condition u(x,0) = f(x). The hint provided involves using the shift formula for the Fourier transform, specifically F[f(ax-b)] = \frac{\exp{i\omega b/a}}{|a|} \overline{f}(\omega/a). The user applies the Fourier transform to each term but expresses uncertainty about the correctness of their approach, particularly in simplifying the expression. The expected solution is presented as an integral involving e^{-i\omega x} and e^{-2(1-\cos{\omega})t}.

PREREQUISITES
  • Understanding of Fourier transforms, specifically the shift formula.
  • Knowledge of ordinary differential equations (ODEs) and their solutions.
  • Familiarity with complex exponentials and their properties.
  • Ability to manipulate integrals in the context of Fourier analysis.
NEXT STEPS
  • Study the application of the Fourier transform in solving linear ODEs.
  • Learn about the properties of complex exponentials in Fourier analysis.
  • Investigate the derivation and application of the shift formula in Fourier transforms.
  • Explore examples of solving ODEs using integral transforms, particularly in physics contexts.
USEFUL FOR

Mathematicians, physicists, and engineers involved in solving differential equations, particularly those utilizing Fourier analysis for modeling and analysis.

cloud18
Messages
6
Reaction score
0
Find the solution (in integral form) of the equation:

[tex] u(x+1,t) - 2u(x,t) + u(x-1,t) = u_t [/tex]
[tex]u(x,0) = f(x)[/tex]

Hint: Use the shift formula

[tex] F[f(ax-b)] = \frac{\exp{i\omega b/a}}{|a|} \overline{f}(\omega/a)[/tex]

So I took the Fourier transform of each term using the shift formula:

[tex] \exp{(-i\omega)} \overline{u} - 2\overline{u} + \exp{(i\omega)}\overline{u} = \overline{u}_t[/tex]

But I don't think this is correct thus far...
 
Last edited:
Physics news on Phys.org
Why don't you think it is right? Also, that expression is not fully simplified. Maybe you were expecting something with sinusoids?
 
Well I guess I get stuck on how to solve the ODE:

[tex] (\exp{(-i\omega)} - 2 + \exp{(i\omega)})\overline{u} = \overline{u}_t[/tex]

So [tex](\exp{(-i\omega)} - 2 + \exp{(i\omega)})[/tex] can be treated as a constant with respect to t. But when I solve I get exponents raised to exponents...The answer is suppose to be (limits of integration -inf to +inf):

[tex] \frac{1}{2\pi}\int{e^{-i\omega x-2(1-\cos{\omega})t} \int{e^{i\omega\xi}f(\xi)d\xi}d\omega}[/tex]
 
Last edited:

Similar threads

Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 12 ·
Replies
12
Views
2K
Replies
3
Views
2K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K