Existence and Uniqueness Theorem for 1st Order IVPs

From ProofWiki
Jump to navigation Jump to search


It has been suggested that this page be renamed.
In particular: Ought to be renamed, this covers pretty much all IVPs
To discuss this page in more detail, feel free to use the talk page.


Theorem

Let $x' = \map f {t, x}$, $\map x {t_0} = x_0$ be an explicit ODE of dimension $n$.

Let there exist an open ball $V = \sqbrk {t_0 - \ell_0, t_0 + \ell_0} \times \map {\overline B} {x_0, \epsilon}$ of $\tuple {t_0, x_0}$ in phase space $\R \times \R^n$ such that $f$ is Lipschitz continuous on $V$.


This article, or a section of it, needs explaining.
In particular: Notation needs to be explained: $\sqbrk {t_0 - \ell_0, t_0 + \ell_0}$ looks as though it should be an interval (and so needs to be written in Wirth interval notation $\closedint {t_0 - \ell_0} {t_0 + \ell_0}$ so as to abide by house style rules), and $\tuple {t_0, x_0}$ is probably an ordered pair. It's not clear enough. The immediate confusion arises because as $\closedint {t_0 - \ell_0} {t_0 + \ell_0}$ is a closed interval it is counter-intuitive for it to be one of the factors of an open ball expressed as a Cartesian product.
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.


Then there exists $\ell < \ell_0$ such that there exists a unique solution $\map x t$ defined for $t \in \closedint {t_0 - \ell_0} {t_0 + \ell_0}$.


This article, or a section of it, needs explaining.
In particular: what is an "IVP"?
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.


Proof

For $0 < \ell < \ell_0$, let $\XX = \map \CC {\closedint {t_0 - \ell_0} {t_0 + \ell_0}; \R^n}$ endowed with the sup norm be the Banach Space of Continuous Functions on Compact Space $\closedint {t_0 - \ell_0} {t_0 + \ell_0} \to \R^n$.

By Fixed Point Formulation of Explicit ODE it is sufficient to find a fixed point of the map $T: \XX \to \XX$ defined by:

$\ds \map {\paren {T x} } t = x_0 + \int_{t_0}^t \map f {s, \map x s} \rd s$


This article, or a section of it, needs explaining.
In particular: Notation not clear: what actually does $T x$ mean? And is $x$ that fixed point posited?
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.


We also have Closed Subset of Complete Metric Space is Complete.

Therefore the Banach Fixed-Point Theorem it is sufficient to find a non-empty subset $\YY \subseteq \XX$ such that:

$\YY$ is closed in $\XX$
$T \YY \subseteq \YY$
$T$ is a contraction on $\YY$


This article, or a section of it, needs explaining.
In particular: Notation not clear: what does $T \YY$ mean?
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.


First note that $V$ is closed and bounded, hence compact by the Heine-Borel Theorem.


The validity of the material on this page is questionable.
In particular: For Heine-Borel Theorem to apply, $V$ needs to be demonstrated to be not only bounded, but also Definition:Totally Bounded Metric Space.
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by resolving the issues.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Questionable}} from the code.
If you would welcome a second opinion as to whether your work is correct, add a call to {{Proofread}} the page.


Therefore since $f$ is continuous, by the extreme value theorem, the maximum $\ds m = \sup_{\tuple {t, x} \mathop \in V} \size {\map f {t, x} }$ exists and is finite.

Let $\kappa$ be the Lipschitz constant of $f$.

Let:

$\YY = \set {y \in \XX: \norm {\map y t - x_0} \le m \size {t - t_0}, t \in \closedint {t_0 - \ell_0} {t_0 + \ell_0} }$

be the cone in $\XX$ centred at $\tuple {t_0, x_0}$.


This article, or a section of it, needs explaining.
In particular: Is this a specialised use of the term "cone" different from the 3-d geometrical construct? If so, it needs to be defined and linked to -- if not, it needs to be explained why it is a cone.
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.


Clearly $\YY$ is closed in $\XX$.

Also for $y \in \YY$ we have:

\(\ds \size {\map {\paren {T y} } t - x_0}\) \(\le\) \(\ds \size {\int_{t_0}^t \map f {s, \map y s} \rd s}\)
\(\ds \) \(\le\) \(\ds m \int_{t_0}^t \rd s\) Absolute Value of Definite Integral
\(\ds \) \(=\) \(\ds m \set {t - t_0}\)

Therefore $T \YY \subseteq \YY$.

Finally we must show that $T$ is a contraction on $\YY$ (we will find that this restricts our choice of $\ell$).

Let $y_1, y_2 \in \YY$.

We have:

\(\ds \size {\map {\paren {T y_1} } t - \map {\paren {T y_2} } t}\) \(\le\) \(\ds \size {\int_{t_0}^t \map f {s, \map {y_1} s} - \map f {s, \map {y_2} s} \rd s}\)
\(\ds \) \(\le\) \(\ds \int_{t_0}^t \size {\map f {s, \map {y_1} s} - \map f {s, \map {y_2} s} } \rd s\) Absolute Value of Definite Integral
\(\ds \) \(\le\) \(\ds \kappa \int_{t_0}^t \size {\map {y_1} t - \map {y_2} t} \rd s\) Lipschitz Condition
\(\ds \) \(\le\) \(\ds \kappa \size {t - t_0} \norm {y_1 - y_2}_\sup\) Estimation Lemma for Contour Integrals


This article, or a section of it, needs explaining.
In particular: Specific meaning of $\norm {y_1 - y_2}_\sup$ needs to be established.
You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it.
To discuss this page in more detail, feel free to use the talk page.
When this work has been completed, you may remove this instance of {{Explain}} from the code.



Taking the supremum over $t$ we have:

$\norm {T y_1 - T y_2}_\sup \le \kappa \ell \norm {y_1 - y_2}_\sup$

for all $y_1, y_2 \in \YY$.


Therefore choosing $\ell < \kappa^{-1}$, $T$ is a contraction on $\YY$ as required.

This completes the proof.

$\blacksquare$

Retrieved from "https://proofwiki.org/w/index.php?title=Existence_and_Uniqueness_Theorem_for_1st_Order_IVPs&oldid=639461"