4
$\begingroup$

Is there an algebraic solution for the to find the intersection of the following two functions for values of $x\geq 0$:
$$f_1(x)=1-2e^{-x/a}=f_2(x)=-1+2e^{-x/b}$$
$a$ and $b$ are positive constants.

The equation can be simplified to:
$$e^{-x/a}+e^{-x/b}=1$$

A Plot is here:

http://img194.imageshack.us/img194/8276/inversionrecovery.jpg

I am searching for the $x$-value of the intersection in the second plot (this is for an inversion recovery experiment inf magnetic resonance).

If there is no algebraic solution, can you suggest a numerical algorithm for this problem?

Thanks

  • 0
    In some special cases we can solve. For example $a=1,b=2$ ... this amounts to a quadratic equation followed by a logarithm.2011-11-24
  • 0
    On another hand if $\frac{1}{a}$ and $\frac{1}{b}$ are algebraic numbers which are linearly independent over $Q$, then $x$ has to be transcendental... So depending by what "algebraic" means, in general I would expect the answer to be no....2011-11-24
  • 0
    Let $e^{-x/{(a*b)}}=t$, then the equation can be simpified to be :$t^b+t^a=1$. The group theroy tells us that there is no general formula to solve polynomial equations whose oder is greater than 4. And I don't think we can find a general formula to solve this equation, for any a and b.(I am not professional, just my thought...) A possible solution is the numerical solution.2011-11-24
  • 0
    So far as I remember, sums of exponentials aren't that easy to solve (barring things like the hyperbolic functions). I too doubt that you'd get a symbolic solution. Be happy with Newton-Raphson.2011-11-24
  • 1
    Well, $x^6+x=1$ has no solution in radicals, although the solution is "algebraic" of course.2011-11-24

2 Answers 2