Hal on IT

940207.1014 Hal PEPINSKY@IUBACS to Bill P. and others on IT.

Rather than trying to unravel error functions wouldn't it be simpler
to compare search strategies--as in sequences of choices, and
foundations of choices, of successive reference signals, simply to
compare whether disturbance error became greater or lesser from signal
to signal? l&p hal

<Bill Leach 07 Feb 1994 20:26:24


(Hal, I'm sorta "hanging" this one on yours if'n ya don't mind. Though a
bit different than what you are asking about, I think it is related).

Does PCT view changes in level for control to be a significant act? By
that I guess I am asking how or if PCT is able to qualitatively deal with
the following different scenerios:

1. Control "suddenly" requires a much larger control action but
    maintains control.

Is this always something that takes place within a single control loop?
Or is it likely something that "wakes up" addition control loops?

2. Control loop suddenly looses control.

In this case I don't have any problem suggesting that another control
loop has to become actively involved. It is still possible that no
"cognetive" control loop is required however, yes?


We have discussed error some. Error as in RMS or Peak errors in
experiments. I am somewhat curious about how PCT view what most
"controller people" call "deadband"?