Ma316-001-F10 further explanation, §3.2

further explanation

existence and uniqueness

Consider two functions p(t) and q(t):

In[122]:=

sec3_2_work_18.gif

Out[122]=

sec3_2_work_19.gif

We know that the differential equation
    y '' + p(t)  y ' + q(t) y = 0,
with initial conditions
    sec3_2_work_20.gif,  sec3_2_work_21.gif

In[145]:=

sec3_2_work_22.gif

Out[145]=

sec3_2_work_23.gif

Has a unique solution:

In[125]:=

sec3_2_work_24.gif

Out[125]=

sec3_2_work_25.gif

...as long as p(t) and q(t) are continuous

In[150]:=

sec3_2_work_26.gif

sec3_2_work_27.gif

Out[150]=

sec3_2_work_28.gif

superposition of two (linearly independent) solutions can satisfy any initial condition

Now, suppose that we have two solutions, sec3_2_work_29.gif  and sec3_2_work_30.gif, of
    y '' + p(t) y ' + q(t) y = 0

In[177]:=

sec3_2_work_31.gif

Out[177]=

sec3_2_work_32.gif

Then, if their Wronskian is non-zero at sec3_2_work_33.gif, given an initial condition there:

In[179]:=

sec3_2_work_34.gif

Out[179]=

sec3_2_work_35.gif

we can find a linear combination of sec3_2_work_36.gif and sec3_2_work_37.gif,  sec3_2_work_38.gif that solves the initial value problem
    y '' + p(t) y ' + q(t) y = 0,  sec3_2_work_39.gif,  sec3_2_work_40.gif:

In[202]:=

sec3_2_work_41.gif

Out[202]=

sec3_2_work_42.gif

superposition of two (linearly independent) solutions gives the general solution

Now suppose that we are given any  solution φ(t) of the differential equation
    y '' + p(t) y ' + q(t) y = 0.
We want to show that we can write φ(t) as a linear combination of sec3_2_work_43.gif and sec3_2_work_44.gif, that is, that we can find a sec3_2_work_45.gif and sec3_2_work_46.gif so that
    sec3_2_work_47.gif
So, suppose that we have such a φ  and that we know that at some point sec3_2_work_48.gif the Wronskian of sec3_2_work_49.gif and sec3_2_work_50.gif is non-zero.

In[210]:=

sec3_2_work_51.gif

Out[210]=

sec3_2_work_52.gif

Then, at sec3_2_work_53.gif, we know that φ(t)  satisfies the initial value problem
    sec3_2_work_54.gif
That is, a solution to the differential equation with the initial conditions:

In[249]:=

sec3_2_work_55.gif

Out[249]=

sec3_2_work_56.gif

That is, we're solving the differential equation with these initial conditions:

In[251]:=

sec3_2_work_57.gif

Out[251]=

sec3_2_work_58.gif

But, by the work in the previous section, we know that we can find a sec3_2_work_59.gif and a sec3_2_work_60.gif such that
    sec3_2_work_61.gif
satisfies the initial value problem
    sec3_2_work_62.gif.
Thus:

In[246]:=

sec3_2_work_63.gif

Out[246]=

sec3_2_work_64.gif

And the uniqueness theorem (the first section above) says that this solution is unique—that is, it must be the same function as the φ(t) that we started with:

In[248]:=

sec3_2_work_65.gif

Out[248]=

sec3_2_work_66.gif

Thus, because we started off with the assumption that φ(t)  was any  solution to the differential equation, we've shown that the linear combination
    sec3_2_work_67.gif
is the general solution  to the differential equation!

Super cool!

Spikey Created with Wolfram Mathematica 7.0