Julian Edward                departmental webpage   



Professor, Ph.D. from Massachussetts Institute of Technology, 1989.
Field: Analysis
Phone number: (305) 348-3050
Office: DM 431C
E-mail: edwardj@fiu.edu

Here are some of my areas of research:
I am currently working on control problems, inverse problems and PDE's. Other stuff I have done:

Control Theory and Inverse Problems I have worked with my departmental colleague
Louis Tebou in the area of Control Theory. More recently, I have been working with Sergei Avdonin at Univ. Alaska at Fairbanks. The basic sort of problem I have been looking at is the following.
Given a vibrating system, can one devise a feedback mechanism that will bring the system
to rest in a given period of time? This field has important applications in Engineering,
where one is trying to control the vibrations of jet engines, bridges, buildings, etc.

Voting Theory.  I have worked with Keith Dougherty from University of Georgia's
Political Science department on some mathematical aspects of voting theory. The mathematical
problems arise here because minor changes in an assembly's voting rules can change the
likelihood that a  measure passes. To give a simple example, if a proposal requires unanimous consent, then
it is much less likely to be approved then if it required a majority of voters to vote in favour
of the proposal. Nonetheless, both Unanimity Rule and Majority Rule are common voting rules
throughout the world. Which of these rules is better? Experts would argue that it depends on
the context. For instance, Unanimity Rule would be unworkable in the UN general assembly, but
it might be appropriate in the US Supreme court. The reason that a given voting rule is appropriate
in one  setting but not in another is due to certain properties that can be quantified, using
either Game Theory or Probability Theory. Keith Dougherty and I have used probabilistic methods
to compare various common voting rules. Hopefully this research might help create assemblies that
better represent the preferences of their constituents. We still haven't figured out a  way to get Democrats
elected, though!

Eigenvalues and Resonances. One of the problems I have studied in the past pertains to the 
existence of eigenvalues and resonances for the Laplacian on perturbed
cylinders with Neumann boundary conditions.
Such perturbed cylinders are known as waveguides,
and the Laplace operator on waveguides is used to model the
vibrations of certain physical systems (such as a car's
exhaust pipe). The existence of $L^2$ eigenvalues and resonances 
are an obstruction to the efficient dissipation of the energy to 
infinity for such vibrating systems.

For the unperturbed cylinder, it is well known that
there are no $L^2$ eigenvalues or resonances. However, for some local
perturbations it has been shown that eigenvalues can exist,
and for asymptotic perturbations the eigenvalues can
even have finite accumulation points. Nevertheless, for generic
perturbations, eigenvalues do not exist.

Earlier this century I proved that an infinite set of resonances near the real axis will exist
if the cylinder has a "stable, periodic billiard ball trajectory". For a picture of
a stable, periodic billiard ball trajectory of length d, see this picture. Never let
your car's exhaust pipe get bent into this shape!

 Recently, I have begun collaborating with colleagues Laura Decarli, Steve Hudson, and Mark Leckband on some solvability problems involving partial differential operators modelled on the Schrodinger operator; see my publication list. This work is a bit more abstract, but does have application to eigenvalue problems among other things.


Some of my papers are available as pdf,  ps or wpd files in this  link: