How do I use BVP5c (or 4c) iteratively to solve BVP for different values of a known parameter

16 views (last 30 days)
I have managed to solve a system of 4 (first-order autonomous) ODE's subject to 7 boundary conditions and for fixed values of my known parameter g. However, now I want to vary g and plot the different curves. Originally I specified the value of g inside my ODE and ODE_bc functions (see below) - e.g. I may write g=1; just below where I specified my unknown parameter r. However, now it seems that I need to specify g outside my functions in order to iterate over different values of g (using a for loop) and using the code below I get the error:
Error using bvp5c
Too many input arguments.
Error in Iterated_ODE (line 4)
sol1 = bvp5c(@simp_ODE,@simp_bc,solinit1,options,g);
I've tried taking the 4th argument g out entirely for the simp_ODE and simp_bc and then for the 5th argument in BVP5c (above) but with no success.
How can I correct this problem?
(see current code below)
options = bvpset('stats','on');
g = 1;
solinit1 = bvpinit(linspace(0,1,20),@ODE_init,[0.75 -0.7 1.47]); %initial guess
sol1 = bvp5c(@simp_ODE,@simp_bc,solinit1,options,g);
sol = bvp5c(@ODE,@ODE_bc,sol1,options,g);
plot(sol.y(1,:),sol.y(3,:)); %plots y(3)=theta vs y(1)=x $
axis equal
hold on
for i=1:10
g = g+0.5; %Increment known/prescribed parameter
sol = bvp5c(@ODE,@ODE_bc,sol,options); %use previous solution as inital guess
plot(sol.y(1,:),sol.y(3,:)); %plots y(3)=theta vs y(1)=x $
axis equal
hold on
end
function G = ODE_init(s)
G = [s %guess x(s) = s%
0
0
0];
end
function dydx = simp_ODE(x,y,array, g) %first-order simplified ODE using cos(x) = 1+O(x), sin(x)=x+O(x^3)%
b = array(1);
n=array(2);
r = array(3);
dydx = [x
y(2)
y(3)
n*y(2)-g*(x+y(1)*y(2))/r]; % b,n,r are unknown parameters to solve for%
end
%tried to specify unknown parameters using array and g is a known parameter%
function res = simp_bc(ya,yb,array,g) %bc's to simplified ODE using a 'first order approximation%
b = array(1) ;
n = array(2);
r = array(3);
res = [ya(1) %x(0)=0%
ya(2) %y(0) = 0%
ya(3) %theta(0) = 0%
yb(4) %psi(1) = theta'(1) =0 in non-dim bc%
yb(3)+b-pi/3 %Wetting angle relation at boundary theta(1)+beta = pi/6%
1 - r*sin(b) %geometric condition%
yb(2) - r*(n +cos(b))]; %internal force y(1) -r*n-r*cos(b) == 0%
end
function dyds = ODE(s,y,array,g)
b = array(1);
n = array(2);
r = array(3);
% g = 1;
dyds = [cos(y(3));
sin(y(3));
y(4);
n*sin(y(3))-g*(y(1)+y(2)*sin(y(3)))/r];
end
function res = ODE_bc(ya,yb,array,g)
b = array(1) ;
n = array(2);
r = array(3);
res = [ya(1) %x(0)=0%
ya(2) %y(0) = 0%
ya(3) %theta(0) = 0%
yb(4) %psi(1) = theta'(1) =0 in non-dim bc%
yb(3)+b-pi/3 %Wetting angle relation at boundary theta(1)+beta = pi/6%
yb(1) - r*sin(b) %geometric condition%
yb(2) - r*(n +cos(b))]; %internal force bc y(1) -r*n-r*cos(b) == 0%
end

Accepted Answer

darova
darova on 7 Mar 2020
Edited: darova on 7 Mar 2020
sol = bvp5c(@(x,y)ODE(x,y,ar,g),@(a,b)ODE_bc(a,b,ar,g),sol,options); %use previous solution as inital guess
  13 Comments
ANDREW CLEMENT
ANDREW CLEMENT on 7 Mar 2020
Actually that final suggestion gave me an idea. I'm not sure if you meant to omit the @ symbols before the functions but it gave me no error on that line so I tried in all the lines using BVP5c and it worked! Thank you very much for your help

Sign in to comment.

More Answers (0)

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!