Confusion, Control -Reply

[Hans Blom, 960118]

(Rick Marken (960118.0800))

Remember that my controller was able to remain in control when no
observations were forthcoming

I remember that your controller maintained its output while input
(observations) were cut off. I also remember that the controlled
variable remained in the reference state as long as no disturbances
were added during the non-observation period (or the observation
period, for that matter).

That is what I meant: the system showed behavior matching the
behavior "prescribed" by the reference, even when no perception
regarding the _effect_ of the behavior was forthcoming.

But I don't remember it remaining in _control_ (keeping a variable
in a reference state and protecting it from disturbances) when no
observations (perceptions) were forthcoming.

How would you define "control" if no observations are forthcoming
(that disturbances cannot be protected from under that condition is
obvious)? In the PCT-sense, control is having reference and percept-
ion matched. In other words, PCT demands the perception being there
-- and being processed -- at all times. My demo stepped outside this
paradigm.

So you, operating from the PCT paradigm, call it "no control". I,
given my engineering background, call it control. But can we agree
somewhere? That "pseudo-control", if you want to call it that, works
fine as long as there are no significant disturbances?

Greetings,

Hans Blom

PS: The prime goal of science is, I think, to discover knowledge that
no reasonable person can disagree with.