ODE45 specific time point settings

Question

Eran Sandman 2020-6-10

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/546032-ode45-specific-time-point-settings

评论： Star Strider 2020-6-16

采纳的回答： Star Strider

在 MATLAB Online 中打开

Hello!

I'm trying to stimulate BIG model in different situations (Beta-cells , insulin , glucose).

The following function is very simple. I'm able to change the parameter 's' at t>100.

%%% BIG Model %%%
function yprime = big(t,y) 
s=1;
delta=1;
q=1;
if t>100
  s=0.2;
end
    
yprime = zeros(3,1);
yprime(1) = m - s*y(2)*y(1);%dG/d
yprime(2) = (q * y(3) * (y(1))^2) - delta * y(2); %dI/dt
yprime(3) = y(3)*0.01*(y(1)-5);%dB/dt

The second thing that I need to do is to change m=2 at t=90, t=110, t=200 (rest of the time m=1)

The problem is that the ODE is choosing the time points by itself and it skips these specific time point. How can I overcome this?

This is the code that i'm using to solve this.

clear all
clc
[t,y] = ode45(@(t,y) big(t,y) , [0:1:400] ,[5,1/5,1/125]);
subplot(1,3,1)
plot(t,y(:,1))
title('Glucose')
xlabel('Time')
subplot(1,3,2)
plot(t,y(:,2))
title('Insulin')
xlabel('Time')
subplot(1,3,3)
plot(t,y(:,3))
title('Beta-Cells')
xlabel('Time')

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2020-6-12

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/546032-ode45-specific-time-point-settings#answer_450378

在 MATLAB Online 中打开

I verified that this version works, however I do not see any significant change in the integrated results.

These added lines:

tcnd = (((t>=90) & (t<91)) | ((t>=110) & (t<111)) | ((t>=200) & (t<201)));
m = tcnd*2 + (~tcnd)*1;

‘toggle’ ‘m’ between ‘1’ and ‘2’ at the appropriate times. An additional fprintf call allows verification.

The (Slightly Revised) Code —

function yprime = big(t,y) 
s=1;
delta=1;
q=1;
if t>100
  s=0.2;
end
tcnd = (((t>=90) & (t<91)) | ((t>=110) & (t<111)) | ((t>=200) & (t<201)));
m = tcnd*2 + (~tcnd)*1;
% fprintf('t = %f\t\tm = %d\n',t,m)                   % Un-Comment To See dm’ At Each Time
yprime = zeros(3,1);
yprime(1) = m - s*y(2)*y(1);%dG/d
yprime(2) = (q * y(3) * (y(1))^2) - delta * y(2); %dI/dt
yprime(3) = y(3)*0.01*(y(1)-5);%dB/dt
end

The Plot —

The only discernable difference are the ‘spikes’.

6 个评论
显示 4更早的评论隐藏 4更早的评论

Eran Sandman 2020-6-16

Evenutally i changed the conditions of 't' to nearby points which likley the intergration goes through them. So it worked well.

Star Strider 2020-6-16

I’m happy you got it sorted.

I still don’t understand what the problem is with ‘s’. In spite of my experience with metabolic models (specifically glucose regulatory models), I don’t recognise the one you’re using.

请先登录，再进行评论。

Answer 2

darova 2020-6-12

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/546032-ode45-specific-time-point-settings#answer_450116

在 MATLAB Online 中打开

Try tolerance

function yprime = big(t,y) 
m = 1;
if any( abs([90 110 200]-t)<0.5 )
    m = 2;
end

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 3

Boxn Hen 2020-6-12

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/546032-ode45-specific-time-point-settings#answer_450150

在 MATLAB Online 中打开

Do you change m at one point like t=90 ?Ode steps small size ,if the tspan is [0,1] and you set m=1 when t=0,while m=2 in other time,it will be meaningless.What I experience is effective if you change m for a period.

function yprime = big2(t,y) 
s=1;
delta=1;
q=1;
m=1;
if t>100
  s=0.2;
end
 if t>90
     m=2;
 end
yprime = zeros(3,1);
yprime(1) = m - s*y(2)*y(1);%dG/d
yprime(2) = (q * y(3) * (y(1))^2) - delta * y(2); %dI/dt
yprime(3) = y(3)*0.01*(y(1)-5);%dB/dt
end