What is the power series solution to this differential equation?

Ted123
Messages
428
Reaction score
0

Homework Statement



[PLAIN]http://img59.imageshack.us/img59/2091/diffeq.png
[PLAIN]http://img684.imageshack.us/img684/6748/diffeqp.png

The Attempt at a Solution



Making the substitutions [tex]y= \sum_{n=0}^{\infty} a_n x^n[/tex] and [tex]y^{\prime} = \sum_{n=0}^{\infty}na_nx^{n-1},[/tex]

[tex]\begin{align*}<br /> y'-2xy & = \sum_{n=0}^{\infty} (na_nx^{n-1} - 2xa_nx^n )\\<br /> &= \sum_{n=0}^{\infty} na_nx^{n-1} - \sum_{n=0}^{\infty} 2xa_nx^n\\<br /> &= \sum_{n=1}^{\infty} na_nx^{n-1} - 2\sum_{n=0}^{\infty} a_nx^{n+1}\\<br /> &= \sum_{n=1}^{\infty} na_nx^{n-1} - 2\sum_{n=0}^{\infty} a_nx^{n+1}<br /> \end{align*}[/tex]

Now I'm having trouble seeing how to shift the summation index to combine the 2 sums into one and to have one coefficient of [tex]x^{n+1}[/tex]
 
Last edited by a moderator:
Physics news on Phys.org
Hi Ted123! :smile:

(on this forum, use "tex", or "itex" for inline latex, rather than "latex" …

and then no need for "begin" etc :wink:)

You need to get everything with the same exponent …

so change that ∑…xn-1 to ∑…xn :smile:
 
tiny-tim said:
Hi Ted123! :smile:

(on this forum, use "tex", or "itex" for inline latex, rather than "latex" …

and then no need for "begin" etc :wink:)

You need to get everything with the same exponent …

so change that ∑…xn-1 to ∑…xn :smile:


Just think it's easier to read if I align the equations :smile:

[tex]\begin{align*}<br /> y'-2xy & = \sum_{n=0}^{\infty} (na_nx^{n-1} - 2xa_nx^n )\\<br /> &= \sum_{n=0}^{\infty} na_nx^{n-1} - \sum_{n=0}^{\infty} 2xa_nx^n\\<br /> &= \sum_{n=1}^{\infty} na_nx^{n-1} - 2\sum_{n=0}^{\infty} a_nx^{n+1}\\<br /> &= \sum_{n=1}^{\infty} na_nx^{n-1} - 2\sum_{n=0}^{\infty} a_nx^{n+1}\\<br /> &= \sum_{n=0}^{\infty} (n+1)a_{n+1}x^n - 2\sum_{n=0}^{\infty} a_nx^{n+1}<br /> \end{align*}[/tex]

The problem is the RHS sum has [tex]x^{n+1}[/tex] not [tex]x^n[/tex]

EDIT: Just realized I can do this:

[tex]a_1+\sum\limits_{n=1}^{\infty} [(n+1)a_{n+1}x^n-2a_{n-1}x^n][/tex]
[tex]a_1+\sum\limits_{n=1}^{\infty} [(n+1)a_{n+1}-2a_{n-1}]x^n[/tex]

Equating the coefficients of [tex]x_n[/tex] at every n, we obtain the recurrence relation

[tex]a_1 = 0[/tex]
[tex](n+1)a_{n+1}-2a_{n-1} = 0,\;\;\;\;\;\;\;\;\;\;n=1,2,3,...[/tex]
or
[tex]a_{n+1} = -\frac{2a_{n-1}}{n+1}[/tex]
 
Last edited:
Equating the coefficients of [tex]x_n[/tex] at every n we obtain the recurrence relation

[tex]a_1 = 0[/tex]
[tex](n+1)a_{n+1}-2a_{n-1} = 0,\;\;\;\;\;\;\;\;\;\;n=1,2,3,...[/tex]
or
[tex]a_{n+1} = -\frac{2a_{n-1}}{n+1}[/tex]

Setting consecutively [tex]n=1,2,3,...[/tex] we determine [tex]a_2, a_3, a_4[/tex] etc.:

[tex]a_2 = -\frac{2a_0}{2} = -a_0[/tex]

[tex]a_3 = -\frac{2a_1}{3} = 0[/tex]

[tex]a_4 = -\frac{2a_2}{4} = \frac{1}{2}a_0[/tex]

[tex]a_5 = -\frac{2a_3}{5} = 0[/tex]

[tex]a_6 = -\frac{2a_4}{6} = -\frac{1}{6}a_0[/tex]

What is the general solution [tex]a_n?[/tex]

It seems to be of the form [tex]a_n = \left\{ \begin{array}{lr} <br /> 0, & \;n\;\text{odd}\\ <br /> \displaystyle \pm\frac{1}{ \left( \frac{n}{2} \right) !}a_0, & \;n\;\text{even}<br /> \end{array} <br /> \right.[/tex]
 
Last edited:
ok, so a2n = a0/n! …

total is ∑ (x2)n/n! …

remind you of anything? :wink:
 
tiny-tim said:
ok, so a2n = a0/n! …

total is ∑ (x2)n/n! …

remind you of anything? :wink:

Actually, I clearly made an algebraic error and there should be no minus sign in the recurrence relation. Then if you follow it through it gives what you have and this equals [tex]a_0 e^{x^2}[/tex]

I have another:

[PLAIN]http://img685.imageshack.us/img685/2091/diffeq.png

[tex]\begin{align*}<br /> xy'-y & = \sum_{n=0}^{\infty} (na_nx^{n} - a_nx^n )\\<br /> &= \sum_{n=0}^{\infty} (n-1)a_nx^{n}<br /> \end{align*}[/tex]

This would imply [tex](n-1)a_n = 0 \Rightarrow a_n = 0[/tex] for all n.

But obviously the solution is [tex]y = a_0 x[/tex] so what discrepencies arise?
 
Last edited by a moderator:
Hi Ted123! :smile:
Ted123 said:
[tex](n-1)a_n = 0 \Rightarrow a_n = 0[/tex]

nooo :redface:

(n-1)an = 0 has two solutions :wink:
 
tiny-tim said:
Hi Ted123! :smile:


nooo :redface:

(n-1)an = 0 has two solutions :wink:

[tex]n=1[/tex] is also a solution...
 
yup! :biggrin:

so when n = 1, an (ie a1) can be anything :wink:
 
  • #10
tiny-tim said:
yup! :biggrin:

so when n = 1, an (ie a1) can be anything :wink:

So how do I express the coefficients [tex]a_n[/tex] in a power series?

[tex]a_n = 0[/tex] unless n=1 in which case [tex]a_1[/tex] is arbitrary.

so [tex]y = a_1 x[/tex]

and this is easily verified by solving the equation directly.

What does it mean when it says 'explain any discrepancies which arise'?
 
  • #11
Ted123 said:
So how do I express the coefficients [tex]a_n[/tex] in a power series?

Just y = a1x (or y = Cx). :smile:
What does it mean when it says 'explain any discrepancies which arise'?

Sorry … I've no idea. :redface:
 

Similar threads

Replies
8
Views
4K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K