[From Rick Marken (960711.1200)]
Me:
The word "disturbance" refers to an environmental variable that influences
the state of a controlled variable (also in the environment). For example,
the controlled variable may be the distance between two chairs. Disturbances
that influence this variable include the positions of the two chairs (d1 and
d2). So the controlled variable is d1-d2. The perceptual representation of
this variable, p, is a function of d1-d2. So p = f(d1-d2).
Hans Blom (960710d) --
This explanation does not accord with the standard PCT control diagram.
There, a disturbance is drawn as an influence (if _I_ may use that word as
wellon a perception that is not caused by an action.
My explanation is in complete "accord with standard PCT". The controlled
variable (the distance between the chairs) is a function of disturbances
(d1 and d2) that influence the controlled variable (and, hence, the
perception of that variable) independently of any action by the control
system. I didn't show any influence of action by the control system because
we were talking about the difference between disturbance and noise; the
control system's influence on a controlled variable is irrelevant, it seemed
to me, to the fact that disturbances -- even variable disturbances -- are
_not_ a noise source with respect to perceptual variables; they are simply
aspects of the real world on which perception is presumably based
But, just to show you how my example relates to the standard control
diagram, here is a control system that controls the (perceived) distance
between two chairs.
r
>
p-----C-----e
> >
> >
q<----------o
^
d1-->|<--d2
The controlled variable (q) is the sensory representation of the distance
between the two chairs (q = d1-d2). d1 and d2 are disturbances to q because
any change in d1 or d2 influences the value of q. The output variable
might be the angular position of the person (control system) relative to the
two chairs. So as the subject moves around the chairs (o varies) the sensory
and perceptual representation of q vary as well. So the subject can control p
by moving relative to the chairs. Or o could be a direct influence on a
chair so that d1 or d2 changes as a result of variations in o. Again, the
system controls p by varying o. An influence of o on d1 or d2 does not
change the fact that d1 and d2 can also change independently of any influence
on q by the system; for example, the wind might blow the chairs so that d1 or
d2 change independently of what the system does to the chairs.
Now it's time for you to answer a question. You had said to Bill:
But _all_ people who drive a car occasionally take their eyes off the road
without causing an immediate accident.
and Bill replied [ Bill Powers (960710.1430 MDT)]
They do this not by controlling a model but simply by holding the steering
wheel in a fixed position. You are dodging my example, though. Can't you
explain why a person can't thread a needle or drive a car blindfolded?
I would like to see your explanation of this myself. Or do the people out in
your part of the world work like their thermostats: through model based
calculation of output?
Best
Rick