Skip to content
GitLab
Explore
Sign in
Register
Primary navigation
Search or go to…
Project
L
lectures
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container Registry
Model registry
Operate
Environments
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Mathematics for Quantum Physics
lectures
Commits
4a63e594
Commit
4a63e594
authored
4 years ago
by
Scarlett Gauthier
Browse files
Options
Downloads
Patches
Plain Diff
Add page 7.
parent
7efa29ae
No related branches found
Branches containing commit
No related tags found
1 merge request
!9
DifferentialEquationsLecture2
Pipeline
#41808
passed
4 years ago
Stage: build
Stage: deploy
Changes
1
Pipelines
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
src/differential_equations_2.md
+86
-1
86 additions, 1 deletion
src/differential_equations_2.md
with
86 additions
and
1 deletion
src/differential_equations_2.md
+
86
−
1
View file @
4a63e594
...
@@ -399,13 +399,98 @@ PDE was supplied with an initial condition.
...
@@ -399,13 +399,98 @@ PDE was supplied with an initial condition.
Putting the solutions to the two ODEs together and redefining
Putting the solutions to the two ODEs together and redefining
$
\T
ilde{A}=A
\c
dot c_1$, we arrive at the solutions for theb PDE,
$
\T
ilde{A}=A
\c
dot c_1$, we arrive at the solutions for theb PDE,
$
\p
si_n(x,t) =
\T
ilde{A} e^{-i
\f
rac{
\l
ambda_n t}{
\h
bar}} sin(
\f
rac{n
\p
i x}{L}).$
$
\p
si_n(x,t) =
\T
ilde{A}
_n
e^{-i
\f
rac{
\l
ambda_n t}{
\h
bar}} sin(
\f
rac{n
\p
i x}{L}).$
Notice that there is one solution $
\p
si_{n}(x,t)$ for each natural number $n$.
Notice that there is one solution $
\p
si_{n}(x,t)$ for each natural number $n$.
These are still very special solutions. We will begin discussing next how to
These are still very special solutions. We will begin discussing next how to
obtain the general solution in our example.
obtain the general solution in our example.
## Self-adjoint differential equations: Connection to Hilbert spaces! ##
As we hinted was possible earlier, let us re-write the previous equation by
defining a linear operator, $L$, acting on the space of functions which satisfy
$
\p
hi(0)=
\p
hi(L)=0$:
$$L[
\c
dot]:=
\f
rac{-
\h
bar^2}{2m}
\f
rac{d^2}{dx^2}[
\c
dot]. $$
Then, the ODE can be writted as
$$L[
\p
hi]=
\l
ambda
\p
hi.$$
This equation looks exactly like, and turns out to be, an eigenvalue equation!
!!! info "Connecting function spaces to Hilbert spaces"
Recall that a space of functions can be transformed into a Hilbert space by
equipping it with a inner product,
$$
\l
angle f, g
\r
angle =
\i
nt^{L}_{0} dx f
*
(x) g(x) $$
Use of this inner product also has utility in demonstrating that particular
operators are
*Hermitian*
. The term hermitian is precisely defined below.
Of considerable interest is that hermition operators have a set of nice
properties including all real eigenvalues and orthonormal eigenfunctions.
The nicest type of operators for many practical purposes are hermitian
operators. In quantum physics for example, any physical operator must be
hermitian. Denote a hilbert space $
\m
athcal{H}$. An opertor
$H:
\m
athcal{H}
\m
apsto
\m
athcal{H}$ is said to be hermitian if it satisfies
$$
\l
angle f, H g
\r
angle =
\l
angle H f, g
\r
angle
\ \f
orall
\
f,
\
g
\ \e
psilon
\ \m
athcal{H}.$$
Now, we would like to investigate whether the operator we have been working with,
$L$ satisfies the criterion of being hermitian over the function space
$
\p
hi(0)=
\p
hi(L)=0$ equipped with the above defined inner product (i.e. it is a
Hilbert space). Denote this Hilbert space $
\m
athcal{H}_{0}$ and consider let
$f,
\
g
\ \e
psilon
\ \m
athcal{H}_0$ denote two functions from the Hilbert space.
Then, we can investigate
$$
\l
angle f, L g
\r
angle =
\f
rac{-
\h
bar^2}{2m}
\i
nt^{L}_{0} dx f
*
(x)
\f
rac{d^2}{dx^2}g(x).$$
As a first step, it is possible to do integration by parts in the integral,
$$
\l
angle f, L g
\r
angle =
\f
rac{+
\h
bar^2}{2m} (
\i
nt^{L}_{0} dx
\f
rac{d f
*}{dx} \frac{d g}{dx} - [f*
(x)
\f
rac{d g}{dx}]
\b
ig{|}^{L}_{0} )$$
The boundary term vansishes, due to the boundary conditions $f(0)=f(L)=0$,
which directly imply $f
*(0)=f*
(L)=0$. Now, intergrate by parts a second time
$$
\l
angle f, L g
\r
angle =
\f
rac{-
\h
bar^2}{2m} (
\i
nt^{L}_{0} dx
\f
rac{d^2 f
*}{dx^2} g(x) - [\frac{d f*
}{dx} g(x)]
\b
ig{|}^{L}_{0} ).$$
As before, the boundary term vanishes, due to the boundary conditions
$g(0)=g(L)=0$. Upon cancelling the boundary term however, the expression on
the right hand side, contained in the integral is simply
$
\l
angle L f, g
\r
angle$. Therefore,
$$
\l
angle f, L g
\r
angle=
\l
angle L f, g
\r
angle. $$
We have demonstrated that $L$ is a hermitian operator on the space
$
\m
athcal{H}_0$. As a hermitian operator, $L$ has the property that it's
eigenfunctions form an orthonormal basis for the space $
\m
athcal{H}_0$. Hence it
is possible to expand any function $f
\ \e
psilon
\ \m
athcal{H}_0$ in terms of
the eigenfunctions of $L$.
!!! info "Connection to quantum states"
Recall that q quantum state $
\k
et{
\p
hi}$ can be written in an orthonormal
basis $
\{
\k
et{u_n}
\}
$ as
$$
\k
et{
\p
hi} =
\u
nderset{n}{
\S
igma}
\b
ra{u_n}
\k
et{
\p
hi}
\k
et{u_n}.$$
In terms of hermitian operators and their eigenfunctions, the eigenfunctions
play the role of the orthonormal basis. In reference to our running example,
the 1D Schr
\"
{o}dinger equation of a free particle, the eigenfunctions
$sin(
\f
rac{n
\p
i x}{L})$ play the role of the basis functions $
\k
et{u_n}$.
To close our running example, consider the initial condition
$
\p
si(x,o) =
\p
si_{0}(x)$. Since the eigenfunctions $sin(
\f
rac{n
\p
i x}{L})$
form a basis, we can now write the general solution to the problem as
$$
\p
si(x,t) =
\o
verset{
\i
nfinity}{
\u
nderset{n}{
\S
igma}} c_n e^{-i
\f
rac{
\l
ambda_n t}{
\h
bar}} sin(
\f
rac{n
\p
i x}{L}),$$
where in the above we have defined the coefficients using the Fourier
coefficient,
$$c_n:=
\i
nt^{L}_{0} dx sin(
\f
rac{n
\p
i x}{L})
\p
si_{0}(x). $$
\ No newline at end of file
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment