0
$\begingroup$

I have the following 2D dynamical system that I solve with NDSolve:

q[v_,x_]:=-v+c Tanh[3*v]-Sign[x];
sol = NDSolve[{x'[t] == v[t], 
   v'[t] == q[x[t],v[t]], x[0] == 0.01, 
   v[0] == 0.01}, {x[t], v[t]}, {t, 0, 100}];

The plot of the solution:

Plot[{Evaluate[x[t] /. sol], Evaluate[v[t] /. sol]}, {t, 0, 100}]

Now I wish to add uncorrelated Noise to the second equation, this is what I have so far:

A[v_]:=1/Cosh[3*v];    
proc = ItoProcess[{\[DifferentialD]x[t] == -v[t]\[DifferentialD]t, \[DifferentialD]v[
          t] == -q[x[t], v[t]] \[DifferentialD]t + 
         A[v[t]] \[DifferentialD]w[t]}, {x[t], v[t]}, {{x, v}, {0.01, 0.01}}, 
      t, w \[Distributed] WienerProcess[]]

and for the plot:

RandomFunction[proc, {0., 100., 0.01}]

I get the message:

"The specification "description" is not a random process recognizedby the system"

what is the source of the error and what is the proper way to write the 2d SDE to obtain both trajectories of x[t] and v[t]?

$\endgroup$
  • $\begingroup$ You have two typos: 1. = should be ==. 2 missing a dt. $\endgroup$ – xslittlegrass May 8 '16 at 17:29
  • $\begingroup$ @xslittlegrass, corrected, still won't go $\endgroup$ – jarhead May 9 '16 at 11:06
3
$\begingroup$

I am not an expert in dynamical systems, but in this way I have no error:

A[v_] := 1/Cosh[3*v];
q[v_, x_] := -v + c Tanh[3*v] - Sign[x];
c = 10^-4;
sol = NDSolveValue[{x'[t] == v[t], v'[t] == q[x[t], v[t]], 
    x[0] == 0.01, v[0] == 0.01}, {x[t], v[t]}, {t, 0, 100}];

Plot[{sol[[1]], sol[[2]]}, {t, 0, 100}]

As you can see, I substituted your NDsolve with NDSolveValue and changed the Plot consistently.

proc = ItoProcess[{\[DifferentialD]x[
       t] == -v[t] \[DifferentialD]t, \[DifferentialD]v[
       t] == -q[x[t], v[t]] \[DifferentialD]t + 
      A[v[t]] \[DifferentialD]w[t]}, {x[t], 
    v[t]}, {{x, v}, {0.01, 0.01}}, t, 
   w \[Distributed] WienerProcess[]];
RandomFunction[proc, {0., 100., 0.01}];
ListLinePlot@%

enter image description here

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.