Control: Inside and Out

[From Rick Marken (960622.1040)]

Me:

The control system itself doesn't know the output or feedback function.
All it [the control system] knows is the perceptual signal. I'll send you a
perceptual signal and you send me back the disturbance.

Bruce Gregory (960622.1105 EDT) --

Sorry for being so dense, but I _still_ cannot see what is the issue here

It happens to everyone when PCT is involved;-)

When I was learning to fly in the summer, the plane would undergo
variations in altitude as the result of passing through rising and falling
parcels of air. These altitude excursions were reflected in the altimeter,
which I identify as a "perceptual signal".

Correct. They are the correlate of one perceptual signal in one control
system in you.

By watching the way I was manipulating the controls,

Now you are talking from the point of view of another control system in you;
the control system that's controlling the perception of altitude is not
perceiving it's own outputs (manipulation of the controls). Another control
system, controlling a different perception (of the relationship between
control stick position and altimeter reading, for example) is involved here.

my flight instructor inferred that I was contributing to these [altitude]
oscillations by excessive efforts to control the altitude

The instructor must base this conclusion on another perception -- a perception
that another control system in you _could_ be controlling; something like
the perception of the phase relationship between two perceptions --of
altitude and stick movement.

He instructed me to release the controls

The control system in you that is controlling altitude is not the one that
controls whether or not you have released the controls. That is done by
the system, in you, that is controlling the perception of grip on the
control stick. Control systems don't control their outputs; only their
inputs.

In fact, the plane did a better job of flying itself than I did in
"helping" it.

Yes. You have a control dynamics problem (your altitude control gain is
too high, probably). Very common when learning to control.

Question: when I was not manipulating the controls, did not the altimeter
readings perceptual signal) contain information about the disturbances?

A system outside of the one controlling altitude could say this; this would be
the system that can perceive _both_ the altimeter and the fact that the output
of the altitude control system (your grip) has released the stick. Even
then, however, the altimeter readings contain "information" only about the
net effect of all disturbances to (contributions to the variance in)
altitude; rising and falling air parcels, variations in attack angle, speed,
etc. But the control system controlling (now not controlling) altitude still
only perceives altitude. Other, higher order (relationship or logic level)
systems (in you or in your instructor) are perceiving the altitude variable
as a _result of_ (or informative about) the effect other variables.

Question: When I resumed manipulating the controls, did not the altimeter
readings contain information about the sum of the disturbances and my
efforts to control for them?

Again, that is a perception that could only be had by a system other than
the one controlling altitude.

The altitude control system (in you) is just one of many control systems
controlling many variables -- some simple (like altitude, force, etc) and
some complex (like the state of relationships, such as "covarying",
"independent", "informative", between other variables). The variable
that a control system controls does not inform that system about anything;
it is just _controlled_ by that system. From the point of view of the
system controlling altitude, the variable being controlled is just a
variable that is kept matching the only other input to the control
system -- the reference signal. But from the point of view of other
systems (in you) the altitude variable might be perceived as "informative".

"Informativeness" is a _perception_ that, as you showed in you example,
can itself be controlled. The degree to which the altitude signal is
perceived as informative about uncontrolled influences on ("disturbances"
to) altitude can be controlled; that is, a system other than the one
controlling altitude can control the degree to which the altitude variable
is perceived as informative about disturbances; as you noted, this
"informativeness controlling" system can control how informative the
altitude signal is about disturbances by changing the reference for
the system that is controlling the grip on the altitude control stick.

What Martin (and Hans, I see) can't seem to do is see control just from
the point of view on one control system; they seem to see control only
from the point of view of a person looking at the control system diagram --
which contains "information" about all the variables involved in a control
loop. If you "look at" control from the point of view of the control system
itself, all you see is what the control system sees -- the changing state
of the perceptual variable. The altitude control system sees only the varying
altitude measure; that is the only _perceptual_ input to the system. There
is no perceptual representation in this system of the output function (the
system has no idea how error is transformed into the muscle tensions); it
has no perceptual representation of the feedback function (the complex
physical effects that lead from muscle tensions to forces applied to the
stick to changes in wing shape to changes in forces on the wings to
changes in altitude that are measured by the altimeter). All the control
system knows is the state of the altimeter; all this other stuff about
"information in the perception", "output functions" and what not are other,
higher order control systems talking -- control systems that are _not_ the
altitude control system but are able to "see" this system from a higher
level. The perceptual inputs to these higher order systems include
variables that are _not_ perceived by the altitude control system --
such as measures of "disturbanecs" like wind speed and measures of
the output of the altitude control system itself (whether or not it is
gripping the control stick).

Does this help?

Best

Rick