Solve y'+y=(2xe^-x)/(1+ye^x) using variation of parameters followed by separation of variables.
@SithsAndGiggles
dy/dx=(2xe^-x)/(1+ye^x)-y?
So you have to solve the same equation twice, using a different method each time?
Same equation twice? I don't think so...
I guess I just don't understand what the directions mean by "...followed by..." Anyway, to use VoP, you'll need two known fundamental solutions. From the homogeneous equation \[y'+y=0\] you have the characteristic equation \[r+1=0~~\iff~~r=-1\] So one solution is \(C_1e^{-x}\) (the fundamental solution is then \(e^{-x}\)). Not sure about the second solution just yet...
I'll come back. Let me look for problems like this.
I don't know if it's relevant or helpful, but I stumbled across an interesting way to express the right hand side in terms of partial fractions: \[\begin{align*}\frac{xe^{-x}}{1+ye^x}&=\frac{x}{e^x(1+ye^x)}\\\\ &=\frac{f(x,y)}{e^x}+\frac{g(x,y)}{1+ye^x}\\\\ x&=f(x,y)(1+ye^x)+g(x,y)e^x\\\\ x&=f(x,y)+e^x\bigg(yf(x,y)+g(x,y)\bigg) \end{align*}\] which means \(f(x,y)=x\) and \(yf(x,y)+g(x,y)=0\). Substituting gives \(g(x,y)=-xy\), and so \[RHS=\frac{xe^{-x}}{1+ye^x}=\frac{x}{e^x}-\frac{xy}{1+ye^x}\]
So now you can split up the equation into two non-homogeneous ones: \[y'+y=xe^{-x}~~~~~~~~~~~\text{(a simple linear equation)}\\ y'+y=-\frac{xy}{1+ye^x}~~\text{(not quite sure yet)}\]
Mathematica 9 came up with the follwoing:\[y(x)=-e^{-x}\pm e^{-2 x} \sqrt{c_1 e^{2 x}+2 e^{2 x} x^2+e^{2 x}} \]
The linear equation gives \(e^{-x}x^2\) as a fundamental solution (I'll leave the details of finding it to you). This solution shows up under the square root in rob's comment
Thanks for the extra info.
You're welcome! So when you use VoP, you'll be using \(y_1=e^{-x}\) and \(y_2=x^2e^{-x}\), which should be linearly independent... Check to make sure that's true.
I'm not sure where the separation of variables comes in though...
Join our real-time social learning platform and learn together with your friends!