[Octave-bug-tracker] [bug #62185] [Octave Forge] (Control) 'impulse' throws error for systems with nonzero feedthrough

[Torsten Lilge]

Update of bug #62185 (project octave):

        Operating System:       Microsoft Windows => Any                    
                 Summary: Function 'impulse' throws error for a certain simple
'ss' system => [Octave Forge] (Control) 'impulse' throws error for systems
with nonzero feedthrough

    _______________________________________________________

Follow-up Comment #1:

Thanks for the report.

The system from force to acceleration is as transfer function


Transfer function 'ans' from input 'Force' to output ...

                -s^2 - 1.57e-15 s - 3.14e-15
 Acceleration:  ----------------------------
                       s^2 + 1 s + 10


with a feedthrough of -1. With an impulse as input function, which is infinite
at t = 0 in continuous time, the output value y(t=0) is also infinite at t =
0. That's why it does not make much sense to compute the (not realizable)
impulse response.

The documentation of matlab says that it just ignores this "discontinuity" for
systems with nonzero feedthrough and returns C*B at t = 0 instead (the part of
y(0) without D*u(0)).

One might discuss whether to not compute the impulse response of such a system
(as Octave does) or to show an incomplete response.


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?62185>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/