Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness The Lecture Contains: Reduction of nth Order IVPs Existence & Uniqueness Well-posedness Sufficient Conditions for Wellposedness
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness
Reduction of n th Order IVPs
An order initial value problem (IVP) can be expressed as
(1.4)
This is equivalent to the following system of Let , then
first order IVPs:
and
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness
Thus, in almost all cases, it will suffice to develop and analyze numerical methods for IVPs that are written as first order vector systems in the explicit form (1.5) where
The initial combination would be specified as
Before proceeding with the numerical solution of the differential equation, we must ask ourselves about the existence and uniqueness of the solution of the given problem. This is stated by the following theorem:
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness Existence & Uniqueness Theorem: If
is a differential equation such that such that (1.6) For all and all (this is called the Lipschitz condition and is called the Lipschitz such that is continuous in the region
, and if there exists a constant
constant), then there exists a unique continuously differentiable function
and
, the initial condition.
An example that illustrates the fact that an initial value problem may have no solution is the equation
The family of solutions is given at the singular point
. Since all solution pass through .
, initial values cannot be
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness
Remark: The continuity of
in both t and y guarantees the existence of a solution and the
Lipschitz condition guarantees uniqueness of the solution and is not needed for existence. For example, the IVP
has two solutions given by and The function does not satisfy a Lipschitz condition at .
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness Well-posedness
In addition to existence and uniqueness, we will want to know something about the stability of solutions of the IVP. In particular, we will usually be interested in the sensitivity of the solution to small changes in the data. Perturbations arise naturally in numerical computation due to discretization and round off errors. A formal study of sensitivity would lead us to the following notion of a well posed problem.
Definition: The IVP
(1.7) is well-posed if there exist positive constants K and such that, for any , the perturbed IVP
(1.8)
satisfies (1.9) whenever for .
Remark: By well-posedness, we mean that small perturbations in the stated problem will lead to small changes in the solutions.
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness Sufficient Conditions for Wellposedness
We shall now give a sufficient condition for wellposedness of an IVP and this is given in the following theorem:
Theorem: If
satisfies a Lipschitz condition on
then
is well-posed on
with respect to all initial data.
Proof: We can show this by considering the perturbed problem
(1.10)
where
and .
are the small perturbations. Let
be the norm
defined as
Let
. Here
is the changed solution and
is the true solution. Now
subtracting (1.7) from (1.10), we get
and
[Link] [Link] AM]
Objectives_template
Module 1: Introduction Lecture 2: Existence, Uniqueness, and Wellposedness
Therefore,
By integration, we find
Consequently, the largest change in the solution of the perturbed problem is bounded by
Where
is independent of
[Link] [Link] AM]
