# General Solution to Linear, Homogenius, Second-Order Differential Equation

1. Sep 16, 2010

### nivekious

I need to solve d/dx(e^(-(x^2))*(du/dx))=u+x*(du/dx)
I get e^(-(x^2))*u''-2xe^(-(x^2))*u'=u+x*u'
u''-2x*u'=e^(x^2)*u+xe^(x^2)*u'
u''+(2x-xe^(x^2))*u'-e^(x^2)*u=0

2. Sep 16, 2010

### jackmell

I can suggest a route to study it. First, get a DE text that deals with second-order DEs with variable coefficients. I use "Intermediate Differential Equations" by Rainville. There are some simple substitutions to try with these types of equations, putting it in it's normal form $v''+Iv=0$ and sometimes that results in a constant-coefficient equation. Some other substitution method to get a constant-coefficient equation. I tried those quickly and did not seem to do it. But you may want to try yourself. My next step would be to turn to Mathematica in desperation using the DSolve command. I tried that and Mathematica can't solve it. My next step, even more desperate but not too uncommon is to use power series which of course in this case will involve Cauchy-products but that's ok according to Rainville. If I had to solve it analytically, and I could think of nothing else, I'd do it with power series.

3. Sep 17, 2010

### JJacquelin

I don't think that the ODE is on the homogenius kind. It doesn't matter since direct integration directly leads to a first order ODE :

#### Attached Files:

• ###### EDO.JPG
File size:
30.5 KB
Views:
93
4. Sep 17, 2010

### jackmell

. . . schooled. Very nice though. :)