1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Differential Equations - Hermite Polynomials

  1. Oct 31, 2011 #1
    1. The problem statement, all variables and given/known data
    Here is the entire problem set, but (obviously) you don't have to do it all, if you could just give me a few hints on where to even start, because I am completely lost.

    Recall that we found the solutions of the Schrodinger equations
    [itex] (x^2 - \partial_x ^2) V_n(x) = (2n+1) V_n(x) \quad \quad n = 0,1,2 [/itex]

    [itex] V_0 (x) = e^{\frac{x^2}{2}} \quad \quad V_n (x) = (x - \partial x)^n V_0(x) [/itex]

    1. Show that for each [itex] n [/itex] there is a polynomial [itex] P_n [/itex] such that [itex] V_n(x)=P_n(x)V_0(x) [/itex]

    2. Show that for any smooth function [itex] v(x) [/itex]

    [itex] (x - \partial_x) v(x) = -e^{\frac{x^2}{2}} \frac{d}{dx} \left ( e^{\frac{-x^2}{2}} v(x) \right ) [/itex]

    2. Relevant equations
    During class we went over the commutator of [itex] [B,C] [/itex] and a lot its it's identities.

    Also [itex] (x^2 + \partial_x ^2) V_n = (2n+1) V_n \rightarrow V_n := (A^*)V_0 \neq 0 [/itex]

    3. The attempt at a solution
    I have no idea where to even start on this. None of this is in the book and our professor said that each of the problems can be solved using a few "tricks" and then after that they aren't that hard.
  2. jcsd
  3. Oct 31, 2011 #2
    I think the problem set is really not that "deep" in the sense that you need to have a very good unserstanding of differential equations, Hermite Polynomials or anything else. You just need to know how to do derivatives. Take a good look at rules of differential calculus. Compute the first derivative of V0(x). What do you get? Compute the second derivative. Do you see a pattern? You just need to have a real good look at what you are doing. You could also start with exercise 2 since it's more general and then do exercise 1 after that, using the result of ex. 2.
  4. Oct 31, 2011 #3
    You could start at 2:

    [tex](x-\partial x)^2 V_0(x)=\left(x^2-2x\partial x+\partial^2 x\right) V_0(x)[/tex]

    Keep in mind that powers of operators are higher-order operators and not operators raised to a power right.

    How about 3? Then use the Binomial Theorem for n.

    For (2), can't you just use the chain-rule and work through that and show that? Remember when you have expressions like:

    [tex](x+\partial x) f(x)[/tex]

    that means the operator [itex](x+\partial x)[/itex] operating on the function x. You know that right? Same dif with the first one raised to higher powers
  5. Oct 31, 2011 #4
    Thank you all very much! I got it, I was just thinking too much into the problem!
  6. Oct 31, 2011 #5
    Think I'm wrong on that. Rathers it's the operator [itex](x-\partial x)[/itex] operating on the operator [itex](x-\partial x)[/itex] so:

    [tex](x-\partial x)^2=(x-\partial x)(x-\partial x)=x^2-x\partial x-\partial x (x)+\partial x \partial x[/tex]

    [tex]=x^2-x\partial x-1+\partial^2 x[/tex]

    and it's operating on the function f(x), not x as I stated above.

    Still gives the polynomial though.
    Last edited: Oct 31, 2011
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook