Scribd
Upload
125 views4 pages

Solving Systems of ODE's With Matlab's Ode45 0.1 Previous Encounters With Ode45

This document discusses using Matlab's ode45 function to solve systems of first-order ordinary differential equations (ODEs). It provides an example of a two-equation system and shows how to define the yprime function to return a column vector of derivatives. The document also explains how to pass parameter values to yprime and use contour plots to visualize 3D data from ODE solutions. The homework asks the reader to write a script that solves two second-order ODEs by rewriting them as systems of first-order equations and producing phase portraits and solution plots for different initial conditions.

Uploaded by

wordsaucer
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
125 views4 pages

Solving Systems of ODE's With Matlab's Ode45 0.1 Previous Encounters With Ode45

This document discusses using Matlab's ode45 function to solve systems of first-order ordinary differential equations (ODEs). It provides an example of a two-equation system and shows how to define the yprime function to return a column vector of derivatives. The document also explains how to pass parameter values to yprime and use contour plots to visualize 3D data from ODE solutions. The homework asks the reader to write a script that solves two second-order ODEs by rewriting them as systems of first-order equations and producing phase portraits and solution plots for different initial conditions.

Uploaded by

wordsaucer
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Solving Systems of ODE’s with Matlab’s ode45

0.1 Previous encounters with ode45


Earlier in the semester we saw how to solve a linear, first-order ODE of the form

y 0 (t) = f (t, y). (1)


We called the Matlab function ode45 using the syntax
[t y]=ode45(@yprime,[t0 tf],y0).
The LHS tells Matlab to store the output from ode45 in two vectors, t and y.
The arguments on the RHS are:

@yprime : handle for Matlab function returning the value of (1)


[t0 tf] : Beginning and ending time for the desired solution
y0 : Initial condition, i.e. y(t0)
Previously, we were able to run ode45 and then use the data stored in t and y
to produce a plot of the numerical solution using the plot command.

0.2 Systems of first-order ODE’s


You’ve been learning in lecture how a linear ODE of order greater than one can
be rewritten as a system of first-order linear equations. In general, if you’re
given any system of first-order ODE’s, whether they’re linear or not, Matlab’s
ODE solvers can at least attempt to solve them numerically. Take, for example,
the following system:
y10 = y2 (2)
y20 = − 51 y2 − sin(y1 )
How do we model this system with Matlab? If you stop to think about it, the
above system is very similar to (1), except that it has two equations instead of
one. We might rewrite it as follows to emphasize the similarity:
y10 (t, y1 , y2 )=f1 (t, y1 , y2 ) (3)
y20 (t, y1 , y2 )=f2 (t, y1 , y2 )
Just as before, we’ll have to write a Matlab function that will take in t and
y and return the value of the derivative, except now y is going to be a vector
containing the two values y1 and y2 . Remember that if y is a row or column
vector, we can access its ith element using the syntax y(i).

function dydt=yprime(t,y)

dydt(1,1)=y(2);
dydt(2,1)=-0.2*y(2)-sin(y(1));

1
Question: What kind of vector does the function yprime output?
Answer: A column vector! This wasn’t arbitrary: Matlab’s ODE’s solvers,
including ode45, want the function that evaluates the derivative to return a
column vector (this does appear to be arbitrary!). If you start getting red error
messages, this might be one place to check!

EXAMPLE: Suppose we want a solution for (2) with t ∈[0,40] and initial
conditions y1 (0) = 0 and y2 (0) = 3. We’re going to call ode45 in the same way
as in the 1-dimensional case, except that now our y0 is a vector of the form [y1
y2]. The command will look like:

[t y]=ode45(@yprime,[0 40],[0 3]);

Now you can plot t vs. y to see the solutions, or y1 vs. y2 to see the phase
portrait. Try this with a few different choices of initial conditions.

0.3 Solving systems of ODE’s with one or more


free parameters
In Lab 1, we tried to improve our model of oscillatory data by adding terms to
our differential equation. What if we wanted to have our yprime function be
able to take in parameters such as these, in addition to t and y? Fortunately,
the ODE solvers in Matlab accomodate this with ease. Suppose we took the
system (2) and let the coefficient of y2 in the second equation be a free parame-
ter. Then we can redefine yprime to be a function of three variables as follows:

function dydt=yprime(t,y,a)

dydt(1,1)=y(2);
dydt(2,1)=a*y(2)-sin(y(1));

If the derivative function called by ode45 has additional parameters, ode45


is called as follows:

[t y]=ode45(@yprime,[t0 tf],[y1 y2],[],a)

The seemingly random [] in the list of arguments for yprime is an empty place-
holder where certain options can be input to ode45. If you’re curious about that,
you can query help ode45 to see more information. Otherwise, just remember
to include it when passing in additional arguments to yprime.

Note: There is no limit to how many free parameters ode45 will pass on to
yprime. You can simply redefine yprime to handle more inputs and add those

2
inputs to the end of the list of ode45 arguments to your heart’s desire. Try gen-
erating solutions for a few different values of a to see the effect of the parameter
on the system.

0.4 Contour plots


NOTE: You won’t need contour plots for this week’s homework. However, you
might find it handy when answering the energy equation questions in Lab 2!

The Matlab contour function will automatically create contour plots for a 3-
dimensional set of data. Recall from calculus that the contour of a multivariable
function, say f (x, y), is a set of pairs (x, y) for which f (x, y) = c, where c is
constant.

Suppose we want to plot contours of the function f (x, y) = x2 + y 2 . First


we need to create a grid of data values (x, y) using the meshgrid function:

[x,y]=meshgrid(0:.1:1,0:.1:1);

To create a plot of 20 different contours of f (x, y), type:

contour(x,y,x.^2+y.^2,20)

As always, you can type help contour to see the documentation.

0.5 Homework
Write an m-file that solves the following two second-order, nonlinear ODE’s:

1. y 00 + sin y = 0

2. y 00 + 1
10 y + sin y = 2 cos t

In order to use the techniques discussed in this worksheet, you’ll have to re-
write both equations as a system of first-order ODE’s: you may want to ask a
neighbor or refer to your textbook if you’ve forgotten how to do this.

What your script should do: Produce two plots per system. One plot
should contain a phase portrait (y1 vs. y2 ) for several different initial condi-
tions. The other should show solutions (y(t) vs. t) corresponding to different
inital conditions. They should be clearly labeled, more or less like the ones
shown below (but don’t worry about having captions).

3
Figure 1: Some solution trajectories Figure 2: A phase portrait for sev-
for the sample system. eral initial conditions.

You might also like