Anyone used MatLAB before?

[simms]

I have an equation
dx/dt = A*B exp(D*t)/E.

How can I put this into MatLAB to get the ODE for x?

 

[RaynorWolfcastle]

matlab is a numberical computation package, the symbolic stuff is handled through a Maple library which I've never used. The matlab helpfile is very good though, so check it out in there.

 

[Cattlegod]

just use mathematica

 

[Heisenberg]

I'd try to use Maple first to do it. It's much simpler to use than Mathematica.

 

[simms]

Originally posted by: Heisenberg
I'd try to use Maple first to do it. It's much simpler to use than Mathematica.

Alright - assuming I went back to school to use Mathematica or Maple, what syntax would I use? I've never really used anything other than MatLAB and Simulink.

 

[RaynorWolfcastle]

Originally posted by: Heisenberg
I'd try to use Maple first to do it. It's much simpler to use than Mathematica.

like I said, matlab's symbolics are handled by a maple library so if all he has is matlab, he can still use that to solve the problem. If I remember my Maple correctly, it will look something like:

dsolve(diff(x(t),t)=A*B*exp(D*t)/E, x(t))

looking at that again, it looks like a pretty straightforward DE though, wouldn't the solution just be x(t) = A*B*exp(D*t)/(D*E)+k, where k is a constant of integration?

 

[Heisenberg]

Originally posted by: RaynorWolfcastle

Originally posted by: Heisenberg
I'd try to use Maple first to do it. It's much simpler to use than Mathematica.

like I said, matlab's symbolics are handled by a maple library so if all he has is matlab, he can still use that to solve the problem. If I remember my Maple correctly, it will look something like:

dsolve(diff(x(t),t)=A*B*exp(D*t)/E, x(t))

looking at that again, it looks like a pretty straightforward DE though, wouldn't the solution just be x(t) = A*B*exp(D*t)/(D*E)+k, where k is a constant of integration?

Yeah, I think that's the right syntax. I haven't use Maple for ODE stuff for a while. I was reading the equation with E in the exponential, but if E is out front then your solution is right. If E's in the exponential, it'd be x(t) = A*B*(E/D)*exp(D*t/E) + k.

 

[simms]

Originally posted by: RaynorWolfcastle

Originally posted by: Heisenberg
I'd try to use Maple first to do it. It's much simpler to use than Mathematica.

like I said, matlab's symbolics are handled by a maple library so if all he has is matlab, he can still use that to solve the problem. If I remember my Maple correctly, it will look something like:

dsolve(diff(x(t),t)=A*B*exp(D*t)/E, x(t))

looking at that again, it looks like a pretty straightforward DE though, wouldn't the solution just be x(t) = A*B*exp(D*t)/(D*E)+k, where k is a constant of integration?

That's what I was thinking, it doesn't seem to be an ODE. However, I have the following equations:

B*Cl - 0 = Vb dCb/dt (Caps are variables, lower cases are subscripts) .... (eq 2)

And

Cl = C1 exp (-B*t/Vl) .... (eq 3)

So it's a simple substitution to get
B*(C1 exp (-B*t/Vl)) - 0 = Vb dCb/dt

Then rearrange.. but I just get dCb/dt = a general form of what I posted.

I have B = 1 L/min, Vb = 50 L, Vl = 2 L, C1 = 0.5 g/L, but no idea of an initial condition.. and it says "Substitute equation 3 into equation 2 and solve the ODE by Standard Euler or Runge-Kutta methods. You can use Mathematica for this purpose."

Heisenberg, thanks for your input. To clarify, E = Vb which is outside of the exp.

 

[jagec]

here's a copy/paste of a simple pair of matlab files I wrote LONG ago.

% filename rhs1.m
function ydot=rhs1(t,y); %compute RHS of eqns for t,y
ydot(1) = y(2);
ydot(2) = 1+ y(2)^2;

ydot = ydot';

% filename simple.m
y1_0=0;
y2_0=0;
t0=0;
tend=pi/4;
tspan=[t0 tend];
[t,y]=ode45(@rhs1,tspan,[y1_0 y2_0])
plot(t,y)

The second one's the driver for the first. Note that it's solving TWO diff eq's.

 

[Heisenberg]

Originally posted by: simms
I have B = 1 L/min, Vb = 50 L, Vl = 2 L, C1 = 0.5 g/L, but no idea of an initial condition.. and it says "Substitute equation 3 into equation 2 and solve the ODE by Standard Euler or Runge-Kutta methods. You can use Mathematica for this purpose." .

Both the Euler and Runga-Kutta methods are numerical methods for solving DE's. If they want you to solve it numerically rather than analytically then you'll probably have to use Matlab or Mathematica.

 

[Random Variable]

No. But I have Maple 10 on my computer.

 

[simms]

Originally posted by: Heisenberg

Originally posted by: simms
I have B = 1 L/min, Vb = 50 L, Vl = 2 L, C1 = 0.5 g/L, but no idea of an initial condition.. and it says "Substitute equation 3 into equation 2 and solve the ODE by Standard Euler or Runge-Kutta methods. You can use Mathematica for this purpose." .

Both the Euler and Runga-Kutta methods are numerical methods for solving DE's. If they want you to solve it numerically rather than analytically then you'll probably have to use Matlab or Mathematica.

I am using Matlab. Did you mean maple?

 

[Heisenberg]

Originally posted by: simms

Originally posted by: Heisenberg

Originally posted by: simms
I have B = 1 L/min, Vb = 50 L, Vl = 2 L, C1 = 0.5 g/L, but no idea of an initial condition.. and it says "Substitute equation 3 into equation 2 and solve the ODE by Standard Euler or Runge-Kutta methods. You can use Mathematica for this purpose." .

Both the Euler and Runga-Kutta methods are numerical methods for solving DE's. If they want you to solve it numerically rather than analytically then you'll probably have to use Matlab or Mathematica.

I am using Matlab. Did you mean maple?

No, I meant Matlab. I don't think Maple can solve DE's numerically, but I could be wrong.

 

[hypn0tik]

Do you really need Matlab for that differential equation?

Edit: I'm assuming A, B, D and E are constants?

 

[simms]

Originally posted by: hypn0tik
Do you really need Matlab for that differential equation?

Edit: I'm assuming A, B, D and E are constants?

Yes, A B D E are constants.

It has to be done via some kind of computer algorithm, so Matlab has to be used (or Mathematica), but I've never solved an ODE or even a DE in MatLAB.