Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I try to solve for the specific function $f(x) = \frac{2-2a}{x-1} \int_0^{x-1} f(y) dy + af(x-1)$

It looks similar to the function used to find the Renyi's parking constant because it came out from a simple generalization of that problem.

The skill I have gained in my differential class can't even solve $f(x) = f'(x-1)$

I'm not looking for anyone to solve it. I just want to know the techniques for solving DE where functions and it's derivatives are evaluated at different points.(What's the terminology for this kind of DE?)

share|cite|improve this question
Delay differential equation. – kennytm Sep 4 '10 at 17:50
Can't $f(y + 1) = f'(y)$ be solved by just substituting $y = e^{ry}$ so you get $e^r = r$ (You need a special function to solve that). So then you appeal to some uniqueness and existence theorem. So the function that maps $y(t)$ to $y(t + 1)$ must be continuous and so must its derivative be. – Jonas Teuwen Sep 4 '10 at 17:54
Differential-difference equation. (Google it.) – whuber Sep 4 '10 at 18:51
The first one, naturally is a delay integro-differential equation (DIDE). ;) – J. M. Sep 4 '10 at 22:07
whuber: My understanding of differential-difference equations is that they look something like this formula for the derivative of a Bessel function: $\frac{\mathrm{d}}{\mathrm{d}x}C_n(x)=\frac12(C_{n-1}(x)-C_{n+1}(x))$, and it can be shown that only the two solutions to the Bessel DE are the solutions to this differential-difference equation. – J. M. Sep 4 '10 at 22:43

Let $u=x-1$ ,

Then $f(u+1)=\dfrac{2(1-a)}{u}\int_0^uf(y)~dy+af(u)$





Let $f(u)=\int_Ce^{us}K(s)~ds$ ,

Then $u\int_Cse^{(u+1)s}K(s)~ds+\int_Ce^{(u+1)s}K(s)~ds-au\int_Cse^{us}K(s)~ds+(a-2)\int_Ce^{us}K(s)~ds=0$






$\therefore s(e^s-a)K'(s)+(se^s-2(a-1))K(s)=0$




$\ln K(s)=-\ln(e^s-a)+\int_k^s\dfrac{2(a-1)}{r(e^r-a)}dr+c_1$


$\therefore f(u)=\int_C\dfrac{ce^{us+\int_k^s\dfrac{2(a-1)}{r(e^r-a)}dr}}{e^s-a}ds$


But since the above procedure in fact suitable for any complex number $s$,

$\therefore f_n(x)=\int_{a_n}^{b_n}\dfrac{c_ne^{x(p_n+q_ni)t+\int_{k_n}^{(p_n+q_ni)t}\dfrac{2(a-1)}{r(e^r-a)}dr}}{e^{(p_n+q_ni)t}-a}d((p_n+q_ni)t)$

For some $x$-independent real number choices of $a_n$ , $b_n$ , $p_n$ , $q_n$ and $k_n$ such that:

$\displaystyle\lim_{t\to a_n}(p_n+q_ni)te^{x(p_n+q_ni)t+\int_{k_n}^{(p_n+q_ni)t}\dfrac{2(a-1)}{r(e^r-a)}dr}=\lim_{t\to b_n}(p_n+q_ni)te^{x(p_n+q_ni)t+\int_{k_n}^{(p_n+q_ni)t}\dfrac{2(a-1)}{r(e^r-a)}dr}$

$\int_{a_n}^{b_n}\dfrac{e^{x(p_n+q_ni)t+\int_{k_n}^{(p_n+q_ni)t}\dfrac{2(a-1)}{r(e^r-a)}dr}}{e^{(p_n+q_ni)t}-a}d((p_n+q_ni)t)$ converges

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.