Proving Hermite polynomials satisfy Hermite's equation

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
1 replies · 4K views
weezy
Messages
92
Reaction score
5
My book (by Mary L Boas) introduces first the Hermite differential equation for Hermite functions:
$$y_n'' - x^2y_n=-(2n+1)y_n$$ and we find solutions like $$y_n=e^{x^2/2}D^n e^{-x^2}$$ where ##D^n=\frac{d^n}{dx^n}##

Now she says that multiplying ##y_n## by ##(-1)^ne^{x^2/2}## gives us what is known as Hermite polynomials: $$H_n(x)=(-1)^ne^{x^2}D^n e^{-x^2}$$which satisfies another equation: $$y''-2xy'+2ny=0 $$
So far so good until I try to prove if ##H_n(x)## does indeed satisfy the equation above. In her problem set she asks to check if ##e^{-x^2/2}H_n(x)## satisfies the Hermite polynomial equation and here I don't understand why is the extra factor ##e^{-x^2/2}## there?

Also, when I try to just check if ##H_n(x)## satisfies the equation I fail to get past this step which is:
$$H_n'(x)=(-1)^ne^{x^2}[D^{n+1}e^{-x^2}+2x \cdot D^ne^{-x^2}] $$ I don't know if this can be further simplified because when I try to calculate ##H_n''(x)## it gets even more complicated and I am unable to prove that ##H_n(x)## is a solution to the equation. My first thought was to find out what ##D^n e^{f(x)}## and I thought if ##f(x)=x^2## I would get a finite polynomial but that didn't happen.

All my questions can be summarised as follows:
-
What is the purpose of ##e^{-x^2/2}## factor?
-
How can I show ##H_n(x)## is a solution to the equation ##y''-2xy'+2ny=0 ##
-
What is the relation between the differential equation of Hermite functions to that of Hermite polynomials? Are the two equations/solutions equivalent in some way?

-
 
weezy said:
All I did was make a substitution: ##H_n(x) = e^{x^2} u, u=D^ne^{-x^2}##
From that substitution, I'm curious how you got other terms with an "##n##" in them to cancel out the "##2ny##" term in the differential equation.

The substitution ##H_n(x) = (-1)^n e^{x^2/2} u ## with ##u = e^{x^2/2} D^n e^{-x^2} ## would allow us to express ##u''## as a function of ##u## since ##y = u## satisfies the first differential equation you gave: ## y''- x^2 y = -(2n+1)y ## , which implies ## u'' = x^2 u - (2n+1)u##.

I don't understand why is the extra factor ##e^{-x^2/2}## there?

A simple minded way to motivate a factor is that we want the Hermite polynomials to be polynomials in x. The result of ## D^n e^{-x^2} ## is going to have the factor ##e^{-x^2}## in it, which needs to be "canceled out" in order to produce a polynomial.