ode45 array size error

Hello everyone,

I have a lenghtly code but pretty much i'm only having 1 problem.

I have the following line of code:

[t,xdot] =ode45(@thetardynamics,[(k-1)*h k*h],X(:,k),options);

and whenever i run my script i get the following error:

Index in position 2 exceeds array bounds.

Error in problemSolver_4 (line 22) [t,xdot] =ode45(@thetardynamics,[(k-1)*h k*h],X(:,k),options);% [t,y] = ode45(odefun,tspan,y0,options)

I know it has to do with the size of the matrix but where exactly is it index 2? Does it mean at the output xdot or is inside the funcion that i have named as thetardynamics.

Thabks for the help

[madhan ravi]

Upload your function thetardynamics and odefun

[Kevin Brinneman]

Sure!, here's thetardynaics:

function [t,xdot]= thetardynamics(tspan,x,u)

m1=10; m2=3; g=9.81; r=1;

xdot(1) = x(3);

xdot(2) = x(4);

xdot(3) = (-2*m2*x(3)*x(2)*x(4)-g*cos(x(1))*(m1*r+m2*x(2))+ue(1))/(m1*r^2+m2*x(2)^2);

xdot(4) = x(3)^2*x(2)-g*sin(x(1))+((ue(2)))/(m2);

xdot= xdot';

end

[Stephan]

We need to know more about tspan and how ue is passed to the function. Provide the whole code please - i suspect there is more than one issue.

[Kevin Brinneman]

Sure, here it is. It is a model predictive controller. unconstrained. Thanks for taking the time.

Answer 1

[Stephan]

Hi,

lot of bugs - find the partially fixed version atacched. The ode45 part now runs - but there is a line:

du=[eye(p) zeros(2,2*(np-1))]*DU;

where you did not specify or define p - in this i can not help you, since i dont know what is meant with p.

To fix the code there were some assumptions needed:

[t, xdot] = ode45(@(t,x)thetardynamics(t,x,u),tspan,x0,options);

--> I guess you want tspan and x0 to pass to ode, otherwise you would not have defined them

In:

function [xdot]= thetardynamics(~,x,u) 
m1=10; m2=3; g=9.81; r=1;
xdot(1) = x(3);
xdot(2) = x(4);
xdot(3) = (-2*m2*x(3)*x(2)*x(4)-g*cos(x(1))*(m1*r+m2*x(2))+u(1))/(m1*r^2+m2*x(2)^2);
xdot(4) = x(3)^2*x(2)-g*sin(x(1))+((u(2)))/(m2);
xdot= xdot'; 
end

i replaced ue by u, because i did not find any ue in your code, but u is there.

Please check this, define p and i think your code is fixed. The obviously errors are not part of this list of fixes done.

However if you look at te results of ode45, you made one step into the right direction:

plot(t,xdot(:,1),t,xdot(:,2),t,xdot(:,3),t,xdot(:,4))

gives:

Best regards

Stephan

[Kevin Brinneman]

Yes!, everything is as you defined it. variable is p=2 but i changed it since it was a constant throughout the code. I will check it out and plot the results, many thanks for taking the time to review my code!