Ordinary Differential Equations - Existence/Uniqueness Proof

Click For Summary

Homework Help Overview

The discussion revolves around a problem from the context of ordinary differential equations, specifically focusing on the Existence and Uniqueness Theorem. The original poster is attempting to demonstrate the convergence of a sequence defined by an integral involving a function and its previous terms.

Discussion Character

  • Exploratory, Assumption checking, Mathematical reasoning

Approaches and Questions Raised

  • The original poster explores various approaches to manipulate the integral definition of the sequence but encounters difficulties in applying the results from previous problems. They express uncertainty about how to handle absolute values and inequalities effectively.
  • Some participants suggest moving absolute values inside the integral and applying known results to progress towards the desired inequality.
  • There is a mention of a misunderstanding regarding an inequality, prompting a clarification on its validity and implications for the problem.

Discussion Status

The discussion is ongoing, with participants providing guidance on how to approach the problem. Suggestions include applying properties of integrals and inequalities to derive the necessary results. However, there is no explicit consensus on a complete solution yet.

Contextual Notes

The problem is constrained by specific results from the textbook that must be utilized, and the original poster has noted challenges in applying these results effectively. The discussion also highlights the importance of initial conditions and the definitions of the sequence involved.

Tsunoyukami
Messages
213
Reaction score
11
I'm having some difficulty with a problem from Boyce & DiPrima's Elementary Differential Equations and Boundary Value Problems, 9th Edition. The problem comes from Section 2.8: The Existence and Uniqueness Theorem and is part of a collection of problems intended to show that the sequence ##{\phi_n(t)}## converges, where

$$\phi_{n+1}(t) = \int_0^t f[s,\phi_n(s)] ds$$

with ##\phi_0(t) = 0##. The problem requires the following two results:

1) If ##\phi_{n-1}(t)## and ##\phi_n(t)## are members of the sequence ##{\phi_n(t)}## then
$$|f[t,\phi_n(t)]-f[t,\phi_{n-1}(t)| \leq K|\phi_n(t) - \phi_{n-1}(t)|$$
where ##K## is chosen to be the maximum value of ##\frac{∂f}{∂y}## in the region ##D##.

2) If ##|t|<h##, then
$$|\phi_1(t)| \leq M|t|$$
where ##M## is chosen so that ##|f(t,y)|\leq M## for ##(t,y)## in ##D##.

The problem is as follows: "Use the results of Problem 16 and part (a) of Problem 17 [Results 1 and 2 above, respectively] to show that ##|\phi_2(t) - \phi_1(t)| \leq \frac{MK|t|^2}{2}##."

(The problem and results are found in Boyce & DiPrima's Elementary Differential Equations and Boundary Value Problems, 9th Edition; Problem 2.8.17(b); pg. 120.)I've been struggling with this problem even though it seems like it should be fairly straightforward. From now on I'm going to write ##\phi_n(t) = \phi_n## for the sake of laziness. I've been playing around with a few different approaches - I tried writing out the left-hand side of my desired result explicitly using the definition of ##\phi_n## which gives me

$$|\phi_2 - \phi_1| = | \int_0^t f(s,\phi_1) ds - \int_0^t f(s,\phi_0) ds | = | \int_0^t f(s,\phi_1) - f(s,\phi_0) ds | $$

I got stuck here. I can almost apply Result 1 now, but not quite since I have my integrand inside the absolute value...and even if I could it didn't yield promising results. Then I thought of using the inequality ##|x-y| \leq | |x| - |y| |## which let me express ##|\phi_1| ## using Result 2 but I was again stuck with no way to write ##|\phi_2|##.

I noticed that the right-hand side of the first result looks very similar to the left-hand side of what I wish to show when you take ##n=2## but that didn't get me very far either:

$$|f[t,\phi_n(t)]-f[t,\phi_{n-1}(t)| \leq K|\phi_n(t) - \phi_{n-1}(t)|$$

So I could try to show

$$|f[t,\phi_n(t)]-f[t,\phi_{n-1}(t)| \leq \frac{MK^2|t|^2}{2}$$

But then I realized that this is only true if I assume that what I want to show is true and so this shouldn't be the approach I take.

Any guidance would be appreciated. I feel like this should be pretty straightforward, but I'm lost for ideas right now. Thanks in advance for any help!
 
Physics news on Phys.org
Starting with what you have:
## | \phi_2 - \phi_1 | \leq \left| \int _0^t f(s, \phi_2) - f(s, \phi_1) ds \right| ##
You can move the absolute value inside the integral to get:
## \left| \int _0^t f(s, \phi_2) - f(s, \phi_1) ds \right| \leq \int _0^t \left| f(s, \phi_2) - f(s, \phi_1) \right| ds##
Then you can apply rule 1.
Expanding that out and doing the same steps should get you to something that looks like :
## \int_0^t | \phi_1 | ds ##
Which should look like the result you need.
 
Sorry, I used the wrong subscripts on the phis inside the integral...after you apply rule 1, use the fact that ##\phi_0= 0##, and then apply rule 2.
Just to be clear...when I see ## MK|t|^2 / 2 ##, I am looking for something that looks like ##\int_0^t MK|s| ds##.
 
I got stuck here. I can almost apply Result 1 now, but not quite since I have my integrand inside the absolute value...and even if I could it didn't yield promising results. Then I thought of using the inequality |x−y|≤||x|−|y|| which let me express |ϕ1| using Result 2 but I was again stuck with no way to write |ϕ2|.

I just want to mention the inequality ##|x - y| \leq \left| |x| - |y| \right|## is not true. The correct inequality would be ##|x - y| \geq \left| |x| - |y| \right|##.

You have correctly expanded the left hand side of what you want to show:

$$| \phi_2(t) - \phi_1(t) | = \left| \int_0^t f(s, \phi_1(s)) \space ds - \int_0^t f(s, \phi_0(s)) \space ds \right| = \left| \int_0^t f(s, \phi_1(s)) - f(s, \phi_0(s)) \space ds \right|$$

As RUber has stated, you can use the property ##\left| \int_a^b f(x) \space dx \right| \leq \int_a^b |f(x)| \space dx## to write:

$$\left| \int_0^t f(s, \phi_1(s)) - f(s, \phi_0(s)) \space ds \right| \leq \int_0^t \left| f(s, \phi_1(s)) - f(s, \phi_0(s)) \right| \space ds$$

After applying rule #1 and the triangle inequality, you are going to wind up with something like:

$$K \int_0^t \left| \phi_1(s) - \phi_0(s) \right| \space ds \leq K \int_0^t |\phi_1(s)| + |\phi_0(s)| \space ds$$

Then you need to apply rule #2 in terms of ##s##. The condition ##\phi_0(t) = 0## will also be important to use.
 
Last edited:

Similar threads

Replies
6
Views
3K
  • · Replies 11 ·
Replies
11
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
8
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 2 ·
Replies
2
Views
329