newbie: solving system of 1st order ODEs with desolve
Subject: newbie: solving system of 1st order ODEs with desolve
From: Krisztian Magori
Date: Thu, 6 Aug 2009 13:53:43 -0400
Dear Members,
I'm a newbie at Maxima, and although I enjoy using it, I'm stuck with a
problem. I'm trying to solve a system of 1st order ODEs with desolve:
eqn_1: 'diff(L(t),t)=(bL*dM*N(t)-mL*L(t))*(1-L(t)/K)-bM*L(t);
eqn_2: 'diff(N(t),t)=bM*L(t)-dM*mM*N(t)-(1-dM)*mdM*N(t);
Using
atsolve(L(t),t=0,0);
atsolve(N(t),t=0,1);
desolve([eqn_1,eqn_2],[L(t),N(t)]);
I get a solution that has laplace and ilt terms in it, and I don't know how
to expand/remove these terms. I attempted to follow the Example at
http://maxima.sourceforge.net/docs/tutorial/en/gaertner-tutorial-revision/Pages/ODE0002.htmon
solving a system of first order ODEs by Laplace transforming the
equations, solving for the laplace transforms with linsolve, and then
inverse laplace transforming those solutions. However, the solutions ended
up having a number of different laplace terms left in them again. I suspect
that the issue might be that these equations are non-linear, because I
noticed that the solution of linsolve for the laplace transforms has other
laplace terms in it as well. Therefore, I tried to look at the contrib_ode
package, but I didn't find a way to run it with a system of two equations.
Is there a better way to arrive at a closed form solution for L(t) and N(t)?
All suggestions are greatly appreciated.
--
Krisztian Magori
postdoctoral researcher
Drake Lab
Odum School of Ecology
University of Georgia
www.twitter.com/BiteOfAMosquito