Ordinary Differential Equations/Substitution 1: Difference between revisions

From testwiki
Jump to navigation Jump to search
imported>Mrjulesd
Rejected the last text change (by 77.126.78.170) and restored revision 3589438 by 2601:445:4380:7DD0:5551:13E3:C552:DA
 
(No difference)

Latest revision as of 19:24, 12 April 2021

[[../First Order|First-Order Differential Equations]]

As we saw in a previous example, sometimes even though an equation isn't separable in its original form, it can be factored into a form where it is. Another way you can turn non-separable equations into separable ones is to use substitution methods.

General substitution procedure

All substitution methods use the same general procedure:

  1. Take a term of the equation and replace it with a variable v. The key is that the new variable must cover all instances of the variable y. Otherwise substitution would not help.
  2. Solve for dydx in terms of v and dvdx. To do this, take the equation v=f(x,y) where f is the term you replaced and take its derivative.
  3. Plug in dvdx and solve for v.
  4. Plug v into the original term replaced, and solve for y.

Constant coefficient substitution

Lets say we have an equation with a term ay + bx + c, such as

dydx=G(ay+bx+c).

where G is a function. This is non-separable. But we can sometimes solve these equations by turning the term into a function v, defining v(x,y) and finding v'(x,y,y').

v(x,y)=ay+bx+c

The trick with the derivation of v is that y is also a function of x. The derivative of v thus becomes

dvdx=adydx+b

In maxima this looks like so:

(%i1) v:a*y(x)+b*x+c;

and

(%i2) diff(v,x);

yielding

(%o1) a*(dy/dx)+b

Next, we rearrange terms and solve for y'(x,v,v'):

dydx=dvdxba

Now plug v back into the original equation, dydx=G(ay+bx+c), and get it into the form dvdx=f(v)

dvdxba=G(v)
dvdx=aG(v)+b

Solve for v, that is integrate on both sides:

dvdx=aG(v)+b
dvaG(v)+b=dx
dvaG(v)+b=dx
dvaG(v)+b=x+D

Once you have v(x), plug back into the definition of v(x) to get y(x).

y(x)=v(x)cbxa

It is highly suggested that one should not memorize this equation, and instead remember the method of solving the problem. The final equation is rather obscure and easy to forget, but if one knows the method, he/she can always solve it. It will also help if one uses other substitution methods.

Example 1

dydx=(x+y+3)2

Lets replace the quantity being raised to a power with v.

v=x+y+3

Now lets find v'.

dvdx=dydx+1

Solve for y'

dydx=dvdx1

Plug in for y and y':

dvdx1=v2
dvdx=v2+1

Now we solve for v, using the methods we learned in Separable Variables:

dvv2+1=dx
dvv2+1=dx
tan1(v)=x+C
v=tan(x+C)

Now that we have v(x), plug back in and find y(x).

y+x+3=tan(x+C)
y=tan(x+C)x3

Other methods

These are not the only possible substitution methods, just some of the more common ones. Substitution methods are a general way to simplify complex differential equations. If you ever come up with a differential equation you can't solve, you can sometimes crack it by finding a substitution and plugging in. Just look for something that simplifies the equation. Remember that between v and v' you must eliminate the y in the equation.

Example 2

2ydydx=y2+x1

This equation isn't separable, and none of the methods we previously used will quite work. Let's use a custom substitution of v = y2 + x − 1. Solve for v':

dvdx=2ydydx+1
2ydydx=dvdx1

Plug into the original equation

dvdx1=v

Solve for v

dvv+1=dx
dvv+1=dx
ln(v+1)=x+C
v=Cex1

Now plug in and get y

Cex1=y2+x1
y2=Cexx

Pretty easy after using that substitution. Keep this method in mind, you will use this for more complex equations.

Template:BookCat