3
$\begingroup$

$f(x)$ is a differentiable function satisfying the relation $f(x)=x^2+\int_0^x e^{-t} f(x-t) dt$, then $\sum\limits_{k=1}^9f(k)$ is equal to

I first differentiated the given relation on both sides w.r.t. $x$ and obtained $f'(x)=2x+e^{-x} f(0).$ Substituting $0$ in the given relation, we have $f(0) = 0$. By substituting the value we get $f'(x)=2x$. Now integrating this on both sides would result in $f(x)=x^2+c$; again for $x=0$ we get $c = 0$. Thus $f(x)=x^2$. Turns out the answer is $$f(x)=x^2+\frac{x^3}{3}.$$ Could someone please explain where I went wrong?

$\endgroup$
5
  • $\begingroup$ We really can't know where you went wrong unless you edit to include all your work. $\endgroup$
    – amWhy
    Commented Jun 24 at 17:30
  • $\begingroup$ Tutorial of MathJax and please do not use images. $\endgroup$
    – Bowei Tang
    Commented Jun 24 at 17:30
  • $\begingroup$ Okay,sorry i am new to Math stackexchange $\endgroup$
    – Cnidarian
    Commented Jun 24 at 17:33
  • $\begingroup$ Thanks for editing! $\endgroup$
    – Cnidarian
    Commented Jun 24 at 17:40
  • $\begingroup$ I integrated it on both sides $\endgroup$
    – Cnidarian
    Commented Jun 24 at 17:45

3 Answers 3

4
$\begingroup$

Let $v=x-t,$ $$\int_{0}^{x}e^{-t}f(x-t)dt=e^{-x}\int_{0}^{x}e^{v}f(v)dv$$

Differentiate the equation on both sides w.r.t $x$ gives:$$f'(x)=2x+f(x)-e^{-x}\int_{0}^{x}e^{v}f(v)dv=2x+f(x)-f(x)+x^2$$ $$f'(x)=2x+x^2$$ Then solve the ODE with initial condition $f(0)=0$.

$\endgroup$
7
  • $\begingroup$ Thanks,but i do have a solution(which is the one you have answered).But could you please tell me where i went wrong ? $\endgroup$
    – Cnidarian
    Commented Jun 24 at 17:49
  • $\begingroup$ @Electro $f'(x)=2x$ is not true. The derivative of $\int_0^x e^{-t} f(x-t)dt$ is not $0$ . Check my steps. $\endgroup$
    – Bowei Tang
    Commented Jun 24 at 17:50
  • $\begingroup$ but why am i not allowed to differentiate it directly,i.e without using king's rule ? $\endgroup$
    – Cnidarian
    Commented Jun 24 at 18:00
  • $\begingroup$ @Electro If you calculate it by using the definition, you'll know this kind of function cannot be differentiated directly. $\endgroup$
    – Bowei Tang
    Commented Jun 24 at 18:02
  • 1
    $\begingroup$ @Electro Feel free to accept if my answer helps. $\endgroup$
    – Bowei Tang
    Commented Jun 24 at 18:33
2
$\begingroup$

Before differentiating use the IVth property of definite integration that $$\int_{0}^{a} g(t) dt=\int_{0}^{a} g(a-t) dt.$$ Then $$\int_{0}^{x} e^{-t} f(x-t) dt=e^{-x}\int_{0}^{x} e^{t} f(t) dt.$$ Next you get $$f(x)=x^2+e^{-x}\int_{0}^{x}e^{t} f(t) dt.~~~~~~(*)$$ Note that $f(0)=0$.Now differentiate w.r.t.x $$f'(x)=2x-e^{-x}\int_0^{x} e^{t}f(t) dt+e^{-x} e^xf(x)$$ Next, use (*) to write $$f'(x)=2x-f(x)-x^2+f(x)\implies f'(x)=2x+x^2\implies f(x)=x^2+x^3/3+C$$ $f(0)=0$ gives $C=0$.

$\endgroup$
2
$\begingroup$

Here is an alternative solution using the Laplace transform. Taking the transform of both sides of the equation, we get $$ \mathcal{L}\{f(x)\}=\mathcal{L}\{x^2\}+\mathcal{L}\{e^{-x}\}\mathcal{L}\{f(x)\} =\frac{2}{s^3}+\frac{1}{s+1}\,\mathcal{L}\{f(x)\} $$ $$ \implies\mathcal{L}\{f(x)\}=\frac{2}{s^3}+\frac{2}{s^4} \implies f(x)=x^2+\frac{x^3}{3}. $$

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .