Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Logistic equation problem

  1. May 20, 2005 #1
    I just want to check my answer for this question:

    The population P(t) of a city suburb is governed by the initial value problem

    P' = P((10^-2) - (10^-6)P) P(0) = 2000

    Time measured in months.
    a) What is the limiting value of the population at t approaches infinity?
    b) How long before the population is one-half of this limiting value?

    a) I used the formula for the general solution
    birth-rate/death-rate = 10^-2/10^-6 = 10^4

    But I'm not sure if that is what I was supposed to do...

    b) Solving the DE
    P' = P((10^-2) - (10^-6)P)

    [tex]\int \frac{dP}{P(10^{-2}-10^{-6})} = \int dt[/tex]

    Using partial fractions, the left can be seperated

    [tex]\int \frac{10^2}{P} + \frac{10^{-4}}{10^{-2}-10^{-6}P} = \int dt[/tex]

    [tex]10^2 * ln(P-ln(10^{-2}-10^{-6}P) = t + c[/tex]

    Using the initial value P(0) = 2000, C = 250,000

    Half of the limiting value = (10^4)/2 = 5000

    Substituting into the equation, t = 139 months (3sf)
  2. jcsd
  3. May 20, 2005 #2


    User Avatar
    Science Advisor
    Homework Helper

    Check your work. Looks like you have lost some terms. You are missing a P in the equation before partial fractions, but your partial fraction expansion looks OK. Your parentheses need some work in

    [tex]10^2 * ln(P-ln(10^{-2}-10^{-6}P) = t + c[/tex]

    Consequently your C does not look right.

    Part a) looks OK. P = [tex]10^4 [/tex] is where the derivative goes to zero, so that is the limiting value of P.
  4. May 20, 2005 #3


    User Avatar
    Science Advisor
    Homework Helper

    So, what up Shan? What about when you correct your formula as Dan said, how about solving it "explicitly" for P(t). You can do that? Don't want to? I don't know. However if you did, it would clearly show what the limiting value is as t goes to infinity. Why not try and report it here. :smile:

    Edit: Plot too but suppose I shouldn't push it.
    Last edited: May 20, 2005
  5. May 20, 2005 #4
    lol sorry, sorry, it was a typo. It's supposed to be

    [tex]\int \frac{dP}{P(10^{-2}-10^{-6}P)} = \int dt[/tex]

    Another typo... V_V

    [tex]10^2 * (lnP-ln(10^{-2}-10^{-6}P)) = t + c[/tex]

    so when I rearrange...

    [tex]lnP-ln(10^{-2}-10^{-6}P)) = \frac{t + c}{10^2}[/tex]

    [tex]ln\frac{P}{10^{-2}-10^{-6}P} = \frac{t + c}{10^2}[/tex]

    [tex]\frac{P}{10^{-2}-10^{-6}P} = e^{\frac{t + c}{10^2}} = Ce^{\frac{t}{100}}[/tex]

    One thing I'm not sure about, can I solve for c using the second equation or the third one? Or does it not matter? Because I used the third equation so that I end up with

    [tex]\frac{2000}{10^{-2}-10^{-6}*2000} = Ce^0[/tex]


    And c ends up being 250,000

    lol, it's the second case. When I rearrange the equation up the top (presuming that it's right), I got

    [tex]P = Ce^{\frac{t}{100}} * (10^{-2}-10^{-6}P)[/tex]

    And I don't know how to get rid of the P on the right... (yes, my algebra is bad lol)
  6. May 20, 2005 #5


    User Avatar
    Science Advisor
    Homework Helper

    Dude, you gotta' get those parenthesis right:

    [tex]ln(P)-ln(10^{-2}-10^{-6}P) = \frac{t + c}{10^2}[/tex]

    We're good here. Actually leaving as:

    [tex]10^2\{ln(P)-ln(10^{-2}-10^{-6}P)\} = t + c[/tex]

    makes it easy to calcualate c but leave it as 'c' until you need it.

    Also, those quantities for logarithms, aren't they just:


    So we have:


    So you take exponentials of both sides and get:


    This is where you're having problems right, isolating P.

    Suppose I have:


    Well, multiplying by 1-x gives me:


    soooooo . . . ax=k-kx
    You got it right?
  7. May 22, 2005 #6
    Sorry for the late reply. So are you saying that

    P = (10^4 * e^(1/100 (t+c))) / (10^6 + e^(1/100 (t+c)))
  8. May 22, 2005 #7


    User Avatar
    Science Advisor
    Homework Helper

    That's it. However, if you would divide the numerator and denominator by the numerator, then the limit as t goes to infinity is clear.
  9. May 22, 2005 #8
    Um, is it 0? After I simplify, I get

    P = 1 / (10^6 + e^(1/100 (t+c))) * (10^4 * e^(1/100 (t+c)))

    Or did I mess up somewhere?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook