0
$\begingroup$

I have the following equation and want to find $f(x)$

$f(x)=x $ + $\int_{0}^{1} (xy^2 + yx^2)$ $f(y)dy$

When i tried to get a solution from wolfram alpha, it gave me an answer but says it is solving a Fredholm Integral Equation. I am a high schooler and have no idea what that means. Is there any simpler method to solve the above problem for $f(x)$?
Also if it is convenient can one please explain what a fredholm integral equation is and how to solve it?

Edit: Can someone help me?

$\endgroup$
  • $\begingroup$ If you look at the RHS, you can see that it has the form $ax^2+bx$. Try plugging in $f\left(x\right)=ax^2+bx$ and see what happens. $\endgroup$ – Joshua Tilley Jul 24 '18 at 12:54
1
$\begingroup$

This is not a topic you'd understand unless you know some graduate level analysis, so don't feel bad!

This is a special type of equation called a "Fredholm" Integral Equation. These arise in many areas of engineering and physics, and are a very interesting mathematical topic unto themselves.

Generally speaking, a Fredholm integral is written in the following form:

$u(x)=f(x)+\lambda \int_{a}^{b}K(x,y)u(y)dy$. The problem is you are given a function $f(x)$, and another function $K(x,y)$, and your goal is to solve for $u(x)$.

Now let's first discuss $K(x,y)$. This is something called the "Kernel", and the solution of the Integral Equation depends heavily on this Kernel. Now the Kernel notation $K(x,y)$ may lead you to think this a function of two variables (not encountered until calculus III). It's not necessarily wrong to think of it this way, but essentially, all you are doing is fixing an $x$ (treating it like a constant), which comes from your interval $[a,b]$ and you integrate with respect to $y$.

Remember integrating with respect to $y$ means you have a function of the variable $y$ and you want to find the anti-derivative/area under the curve. Example: $\int y^2 dy=y^3/3+C$.

Your Kernel is $K(x,y)=xy^2+x^2y$, and has a special property. It is a "separable" Kernel, which means it can be written as a sum and product of functions $g_n(x), h_n(y)$. When a Kernel is separable, it can be solved using linear algebra methods (also a slightly more advanced course than calculus).

I will edit my answer with the full solution included if you wish.

One final point: sometimes, Integral equations have no solutions, it depends on $\lambda$, and also on the "energy" of the Kernel. It turns out that if $\lambda$ obeys a certain inequality, then the solution to the Integral Equation not only exists, it is the only one, sometimes there are none.

There a lot of other possibilities, for example, if your Kernel is not separable. The Kernel function $K(x,y)$ doesn't even need to be continuous. The solution depends a lot on what kind of function you are working with. You also can have your function $u(y)=tan(e^y)$ which is a non-linear function. When you throw in non-linear terms, the equations become harder to solve in general, and we need to use computational methods to get an approximate solution (which is yet another topic unto itself).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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