Examples of Feedforward Control?

CSGnet Archive CSG2010
21

[From Bruce Gregory (2010.01.04.1630 GMT)]

That's a possibility of course. However, I have heard the following answer to your question. "It's because of the tilt of the Earth's axis. In June, it's winter in Australia, because Australia is tilted farther away from the Sun." As far a analogies go, there is nothing wrong with that one. As a model, however, it has problems. The student has heard that seasons are caused by the tilt of the earth's axis of rotation, so he comes up with an answer that incorporates that story. If you press further, you typically find that there is no model behind the answer. You have to understand the scale of the system to realize why the differing distances to the sun of the northern and southern hemispheres are unlikely to explain the seasons. (They do seem to play a role in long-term climate change, however.)

···

[From Richard Kennaway (2010.01.04.1345 GMT)]

[From Bruce Gregory (2010.01.04.1408 GMT)]
I find science education filled with faulty uses of the term model. The idea that students have a faulty model of some physical phenomenon suggests that they somehow look inside their heads, manipulate a representation of the world, and then report on the result of their manipulations. It has always seemed much simpler to me to think that students are arguing by analogy ("the earth must be closer to the sun in summer because when you are closer to a source of heat you feel warmer"). Models are pretty sophisticated ways to look at the world; I doubt that they come naturally to anyone.

What's the difference between an analogy and a model? A mathematical model consists of a set of mathematical variables that correspond to some physical quantities, and mathematical relationships between them that match the way the physical quantities behave. Someone imagining a long elliptical orbit for the Earth with summer at the closest point to the Sun seems to me to be applying a model, even if they aren't doing mathematical calculations. Just a wrong model. ("So why is it winter in Australia at the same time as summer in the US?")

--
Richard Kennaway, jrk@cmp.uea.ac.uk, Richard Kennaway
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.

22

[From Richard Kennaway (2010.01.04.0928 GMT)]

Happy New Year!

[From Bill Powers (2010.01.03.-0925 MST)]
I hope somebody is digging into the history of feedforward. I expect
to find that it's a product of ignorance about how control systems
really work, coupled with someone's experience that he could usually
work out how things work without having to crack a textbook. That's a
kind of hubris with which I am embarassingly familiar, though with
decreasing frequency in my later years.

JRK: Did it originate in amplifier design? I was googling for feedforward and found this: Theory of Feed-Forward Audio Amplifiers - A survey,
a design from 1921 by Harold Black, whose name I'm sure you'll know. (For those who don't know, the inventor of the negative feedback amplifier. Amplifiers and control systems have a lot in common.) Note that (from Wikipedia: Harold Stephen Black - Wikipedia) this was six years before he came up with the negative feedback amplifier.

BP: Bruce Abbott also came up with some equally surprising ancient references to feedforward in the engineering literature. As I said to him, the implication is that we should be searching for its first use in psychlogy, the place where it is most likely to have lost its original meaning.

JRK: I remember a series of articles that appeared there analysing in detail the design of a particular commercially manufactured audio amplifier, about which there was a controversy: was it a feedback or a feedforward design? The authors mathematically modelled the whole thing, and pointed out that if you arranged the equations like this, it looked like feedforward, and if you arranged them like that, it looked like feedback. Mathematically equivalent descriptions.

This would be a sure sign to me that it doesn't matter what you call it.

"Fast and coordinated arm movements cannot be executed under pure feedback control because biological feedback loops are both too slow and have small gains."

That's another one I'd like to track down. The justification given most often for "feedback is too slow" is that the delay in real control systems makes them unstable if the gain is high. Demo 3-1 in LCS3 takes care of that.

The "small gains" myth comes from using the wrong model. In Demo 4-1, Trackanalyze, the gains and delays I measure in one series of three runs are

Difficulty Steady-state Gain Delay

      1 150 117 msec
      3 88.5 133 msec
      5 9.2 117 msec

... which shows gain falling off with frequency as it should, and being quite high at low frequencies (= low difficulty). Gain here is the measured integration gain divided by the measured damping factor. Note the measured delay in milliseconds.

It proposes a feedforward scheme using an inverse model to compute the required muscular outputs to execute a desired trajectory, with the parameters of the inverse model tuned by feedback.

All I can say is that I wish I had that kind of blind faith in an omnipotent brain. No, I don't.

Best,

Bill P.

···

At 09:30 AM 1/4/2010 +0000, you wrote:

23

[From Bill Powers (2010.01.04.1005 MST)]

Bruce Gregory 92010.01.04 12:02 GMT --

BP: I was a bit surprised to find how pleased I felt to see your name again. I hope all is well with you.

BG: I have always wondered how a system "using an inverse model to compute the required muscular outputs" might be expected to evolve.

BP: Why, that's easy. Natural selection simply removes all the organisms that don't compute the right outputs. That leaves only the ones who do.

Best,

Bill P.

24

[From Bill Powers (2010.01.04.1023 MST)]

Fred Nickols (2010.01.04.0828 MST)--

FN: Hmm. I don't see either of those examples as "feedforward." It seems to me that the punched cards and the model served as reference conditions or standards for the loom and the cutting tool respectively. The loom and the cutting tool simply did what they were told to do so to speak. Operating conditions were no doubt standardized and so (a) no variance from the "program" occurred and (b) no disturbances existed, hence no need for feedback from the process. A "program" would work. If that's "feedforward" then it seems to me that "feedforward" is basically a matter of issuing commands to a compliant system that can do nothing except follow orders.

That's basically what it means. The problem is with explaining how it CAN just follow orders -- that's not easy to arrange. In the Jacquard Loom, the linkages were open loop: if the machine didn't follow orders exactly, it did not change its behavior to correct the error. So a lot of attention had to be paid to making sure there were no disturbances, that the machine itself wouldn't wear out too fast, that the linkages had no significant play or backlash, that the yarn or string was fed accurately, and so on down a long list. I suspect that it took a lot of someone's time to make sure it kept working right.

Best,

Bill P.

25

[From Bruce Gregory (2010.01.04.1741 GMT)]

I am indeed well. I checked out the CSGnet archives and was pleased to see that you had finally come come to your senses and adopted my way of thinking about the issues :wink: On a more serious note, I found that one of the powerful ideas behind PCT (negative feedback control goes without saying) is that human beings are agents and agents have goals. That insight turns out to be incredibly powerful and is, of course, totally ignored in S-R thinking. It is easy to remain clueless if you do not look for the goals and are absorbed instead by the story. People tell stories because they have goals and not the other way around. The goals, of course, are revealed by the behavior involved in exercising control.

One thing I discovered about myself was that whenever I feel anxious or discouraged control is almost always involved. I find that this recognition alone is very liberating.

I hope all is well with you, too.

Bruce

···

On Jan 4, 2010, at 12:08 PM, Bill Powers wrote:

[From Bill Powers (2010.01.04.1005 MST)]

Bruce Gregory 92010.01.04 12:02 GMT --

BP: I was a bit surprised to find how pleased I felt to see your name again. I hope all is well with you.

BG: I have always wondered how a system "using an inverse model to compute the required muscular outputs" might be expected to evolve.

BP: Why, that's easy. Natural selection simply removes all the organisms that don't compute the right outputs. That leaves only the ones who do.

Best,

Bill P.

26

[From Fred Nickols (2010.01.03.1240 MST)]

[From Bruce Gregory (2010.01.04.1741 GMT)]

On a more serious note, I found that one of the powerful ideas
behind PCT (negative feedback control goes without saying) is that human beings
are agents and agents have goals. That insight turns out to be incredibly
powerful and is, of course, totally ignored in S-R thinking. It is easy to
remain clueless if you do not look for the goals and are absorbed instead by the
story. People tell stories because they have goals and not the other way around.
The goals, of course, are revealed by the behavior involved in exercising

control.

The much talked about shift to knowledge work is in fact a shift from prefigured work routines to configured responses to the situation at hand. Instead of doing what others have figured out, many if not most people have to figure out what to do. In essence, they are expected to achieve reasonably stable results in the face of varying circumstances. (Where have I heard that before?) For management, the main implication of this shift is that employees must be dealt with as agents who are expected to act on their employer's behalf instead of as instruments of managerial will. The shift is from compliance to contribution. Management has yet to make it.

···

--
Regards,

Fred Nickols
Managing Partner
Distance Consulting, LLC
nickols@att.net
www.nickols.us

"Assistance at A Distance"

27

[From Bill Powers (2010.01.04.1305 MST)]

Bruce Gregory (2010.01.04.1741 GMT) –

BG: I am indeed well. I checked
out the CSGnet archives and was pleased to see that you had finally come
come to your senses and adopted my way of thinking about the issues
:wink:

That doesn’t work on me any more. I’m still glad you’re back.

On a more serious note, I
found that one of the powerful ideas behind PCT (negative feedback
control goes without saying) is that human beings are agents and agents
have goals. That insight turns out to be incredibly powerful and is, of
course, totally ignored in S-R thinking. It is easy to remain clueless if
you do not look for the goals and are absorbed instead by the story.
People tell stories because they have goals and not the other way around.
The goals, of course, are revealed by the behavior involved in exercising
control.

I don’t know if you’ve been following journals like Science and
Nature. Something else has gradually crept onto the scene, in the
form of new disciplines distinguished from the old ones by inserting
“neuro-” or “brain-based” in front of the old names,
like “neuro-psychology” and “brain-based learning.”
In those two distinguished journals it is showing up in the form of
reports on the causes of various conditions and behavior patterns. It’s
basically S-R psychology backed up by brain scans and very little
else.

I’m sure there are some valid observations and objectives lurking in the
background of such papers, but it’s discouraging to see that we still
have so much work left to do, after all these years. At least people have
stopped telling me that I can stop beating the horse because S-R
psychology is dead.

One thing I discovered about
myself was that whenever I feel anxious or discouraged control is almost
always involved. I find that this recognition alone is very
liberating.

I’ll be interested in your comments on the MOL paper I sent you.

Best,

Bill P.

28

[From Mike Acree (2010.01.04.1241
PST)]

Bill Powers (2010.01.04.1305 MST)–

At least people
have stopped telling me that I can stop beating the horse because S-R
psychology is dead.
Arthur Koestler said of S-R
psychology over 40 years ago (in The Ghost
in the Machine, 1967) that “there has never been a dead horse
with such a vicious kick.” He studied the phenomenon more generally in
an essay on the SPCDH.

Mike

29

[From Rick Marken (2010.01.04.2220)]

Bill Powers (2010.01.04.1305 MST)--

At least people have stopped telling me that I can stop beating the horse
because S-R psychology is dead.

If, instead of "S-R Psychology" we call it what it is -- "Psychology
based on the open-loop causal models of behavior" -- then it is clear
that all Psychologies -- S-R, cognitive, neurophysiological, etc --
have always been S-R in this sense, as evidenced by the fact that all
research in psychology is based on an open-loop causal model, the
general linear model of statistics. This is the point of my
"Revolution" paper (available at
APA PsycNet).

No, the name "S-R" may be dead but the causal model that it implies
lives on in even the most sophisticated cognitive models of behavior.
It's the open loop causal model of behavior that isn't dead; indeed,
it is still very much alive and fighting to keep my papers out of the
journals;-)

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com