Living Control Systems III

[Philip 5/5/2014]

May we resume discussion of LCS III? I just got the book and I’m reading it and there’s some things I need to ask. I’ll soon be finishing my own book and I don’t want to leave anything important unaddressed (by the way, you guys are going to LOVE the applications I found for PCT)!

My first question is about what’s on page 25. What is the explanation of the anomalous correlation between mouse movements and the controlled variable being so low and the correlation between the mouse movement and the uncontrolled variables being so much higher? A correlation of 0.10 seems very strange because it’s not just small compared to the others, it’s anomalously small. Bill says it’s not usually as small as in figure 2-2, but what was Bill doing in that run that made it so small? I understand that if one we’re to look at the instantaneous direction of mouse movement and compare it to the instantaneous direction of aspect change in any given instant of time, then there might be a low correlation between the direction the mouse is moving in a single instant and the direction it should ideally be moving in at that instant - i.e. because the disturbance itself is unpredictable and whatnot, and we don’t predict our actions but only control the perception and so we’re technically allowed to perform sweeping error corrections over the time average. But then again, we’d be talking about two different things: instantaneous movement direction and instantaneous position. Am I shooting into the dark here?

Also, could you please explain to me the essential difference between “position vs disturbance”, “disturbance vs mouse”, and “position vs mouse”?

On a side note, looking at the position graph, there seems to be a relation between the mouse position and object aspect traces, in that it vaguely looks like the mouse position is a harmonic function about the object aspect (which is approximately a “straight line”). Am I just imagining things?

Phil

.

[From Adam Matic 2014.05.06.17.00]

···

Philip 5/5/2014

My first question is about what’s on page 25. What is the explanation of the anomalous correlation between mouse movements and the controlled variable being so low and the correlation between the mouse movement and the uncontrolled variables being so much higher? A correlation of 0.10 seems very strange because it’s not just small compared to the others, it’s anomalously small.

Hi Philip,

It is not anomalous to have a correlation between a controlled variable and output variable in the range of 0.0 - 0.2. It is expected. The better the control, the smaller the correlation. If you practice controlling one aspect of the shape a few times, you can get the correlation as low as 0.01 or 0.00.

Also, could you please explain to me the essential difference between “position vs disturbance”, “disturbance vs mouse”, and “position vs mouse”?

Figure 2.2 on page 23, shows three graphs. Black line is an aspect of the object - first graph shows position, second orientation and third shape. The green line is disturbance, and there are three disturbances, different for each aspect of the object. As you can see when you compare the graphs, the output (red = mouse movement) is the same in all three graphs. It is almost an inverse of the first disturbance (green). Position of the object (black) shown on the first graph is almost constant.

Position vs disturbance is the correlation between black and green line, position vs mouse is correlation between object position (black) and mouse movement (red), and so on.

The low correlation between position of the object and mouse movements comes from the fact that in this case - position of the object was a controlled variable - it was kept constant against disturbances using mouse movements. There are usually low correlations between constant and a varying quantities. At the same time, mouse movements correlate negatively (-0.99) with the disturbance variable, since they are almost exactly opposite.

Hope this helps,

Adam

[David Goldstein (2013.12.31.14:05)]

Dear CSGnet friends:

Rick and I will start going through LCS III on Monday, January 20. He will be posting a study guide for all of us. I just thought I would post the following quotes and comments as a way to start.

Quotes,
and Comments about Living Control Systems III: The Fact of Control

David
M. Goldstein, Ph.D.

About the Author: iv to v. Isn’t the quote “I’ve
always been pretty much a blue-collar type, by training and by preference�
interesting. I wouldn’t have expected this from a theoretician.

Forward: xi to xiii: “…that negative feedbackk
control is the ingredient that has been missing from the life sciences for as
long as they have existed…

“…They contrrol. They never produce any behavior
except for the purpose of making what they are experiencing become more like
what they intend or want to experience, and then keeping it that way even in a
changing world.�

“If they plan, they plan perceptions, not actions.
If they respond to stimuli, they do so in order to prevent those stimuli from
affecting variables they have under control.�

“To me, this theory is beautiful, ot because it is mine
but because it is true.�

Chapter 1: Concepts of Control

Page 1: “To control something is to act to bring it
to a specified condition, and then maintain it close to the that condition even
if unpredictable external forces and changes in environmental properties tend
to alter it.�

MCT:
ANALYZE-COMPUTE-ACT CONTROL

Input
leads to actions which leads to the desired result.

The engineers in the Apollo 13 mission had to figure out how to reduce the CO2
level with something that the astronauts could build; they came up with a
system that worked.

Page 3 “…get the flavor of all thee little problems
that arise in trying to build a system that, all by itself, can analyze the
environment, compute how to achieve a prescribed result, and then act on the
environment in exactly the calculated way.�

Problem: controlling the speed of a car. Bill goes
through all the forces acting on the car. The problem is that the net force
acting on the car changes as the car moves over the terrain.

PERCEIVE-COMPARE-ACT
CONTROL

Page 7: “Negative feedback control as used in PCT
can be summed up very quickly. It involves continuously perceiving the current
state of whatever is to be controlled, continuously comparing that perception
with the intended state, and continuously acting to reduce the difference
between perceived and intended states , keeping the difference as close to zero
as possible.�

Page 9: “A negative feedback controller, or
perceive-compare-act controller, or PCT controller doesn’t have to know what is
causing the speed to change.�

“Most of what has been said in the literature of the
life sciences about limitations of negative feedback control is simple false.�

SIMULATION
AND MODELING

The two ways of calculating the time to fill a
bucket is discussed.

Simulation is advocated over analytical solutions.

Readers may want to download a free version of
VenSim:

http://www.shodor.org/tutorials/vensim/pre.php

and follow a simulation of how a temperature control system works
in Vensim:

http://plaidcircuitry.com/computational_thinking/vensim/index.html

SOME
PHILOSOPHY OF SCIENCE

Page 17: “…the way that collections of materials
behave depends as much on how they are organized as on what the materials are.�

“You can’t cause a mouse to jump in the air by
itself if it doesn’t want to.â€?

David

···

Yeah, ok

···

Sent from my iPhone

On Dec 31, 2013, at 2:10 PM, “D GOLDSTEIN” davidmg@VERIZON.NET wrote:

[David Goldstein (2013.12.31.14:05)]

Dear CSGnet friends:

Rick and I will start going through LCS III on Monday, January 20. He will be posting a study guide for all of us. I just thought I would post the following quotes and comments as a way to start.

Quotes,
and Comments about Living Control Systems III: The Fact of Control

David
M. Goldstein, Ph.D.

About the Author: iv to v. Isn’t the quote “I’ve
always been pretty much a blue-collar type, by training and by preference�
interesting. I wouldn’t have expected this from a theoretician.

Forward: xi to xiii: “…that negative feedback
control is the ingredient that has been missing from the life sciences for as
long as they have existed…

“…They control. They never produce any behavior
except for the purpose of making what they are experiencing become more like
what they intend or want to experience, and then keeping it that way even in a
changing world.�

“If they plan, they plan perceptions, not actions.
If they respond to stimuli, they do so in order to prevent those stimuli from
affecting variables they have under control.�

“To me, this theory is beautiful, ot because it is mine
but because it is true.�

Chapter 1: Concepts of Control

Page 1: “To control something is to act to bring it
to a specified condition, and then maintain it close to the that condition even
if unpredictable external forces and changes in environmental properties tend
to alter it.�

MCT:
ANALYZE-COMPUTE-ACT CONTROL

Input
leads to actions which leads to the desired result.

The engineers in the Apollo 13 mission had to figure out how to reduce the CO2
level with something that the astronauts could build; they came up with a
system that worked.

Page 3 “…get the flavor off all the little problems
that arise in trying to build a system that, all by itself, can analyze the
environment, compute how to achieve a prescribed result, and then act on the
environment in exactly the calculated way.�

Problem: controlling the speed of a car. Bill goes
through all the forces acting on the car. The problem is that the net force
acting on the car changes as the car moves over the terrain.

PERCEIVE-COMPARE-ACT
CONTROL

Page 7: “Negative feedback control as used in PCT
can be summed up very quickly. It involves continuously perceiving the current
state of whatever is to be controlled, continuously comparing that perception
with the intended state, and continuously acting to reduce the difference
between perceived and intended states , keeping the difference as close to zero
as possible.�

Page 9: “A negative feedback controller, or
perceive-compare-act controller, or PCT controller doesn’t have to know what is
causing the speed to change.�

“Most of what has been said in the literature of the
life sciences about limitations of negative feedback control is simple false.�

SIMULATION
AND MODELING

The two ways of calculating the time to fill a
bucket is discussed.

Simulation is advocated over analytical solutions.

Readers may want to download a free version of
VenSim:

http://www.shodor.org/tutorials/vensim/pre.php

and follow a simulation of how a temperature control system works
in Vensim:

http://plaidcircuitry.com/computational_thinking/vensim/index.html

SOME
PHILOSOPHY OF SCIENCE

Page 17: “…thee way that collections of materials
behave depends as much on how they are organized as on what the materials are.�

“You can’t cause a mouse to jump in the air by
itself if it doesn’t want to.�

David

David, sorry about the last message . Was responding to someone else. Happy Holidays.

John

···

Sent from my iPhone

On Dec 31, 2013, at 2:10 PM, “D GOLDSTEIN” davidmg@VERIZON.NET wrote:

[David Goldstein (2013.12.31.14:05)]

Dear CSGnet friends:

Rick and I will start going through LCS III on Monday, January 20. He will be posting a study guide for all of us. I just thought I would post the following quotes and comments as a way to start.

Quotes,
and Comments about Living Control Systems III: The Fact of Control

David
M. Goldstein, Ph.D.

About the Author: iv to v. Isn’t the quote “I’ve
always been pretty much a blue-collar type, by training and by preference�
interesting. I wouldn’t have expected this from a theoretician.

Forward: xi to xiii: “…that negative feedback
control is the ingredient that has been missing from the life sciences for as
long as they have existed…

“…They control. They never produce any behavior
except for the purpose of making what they are experiencing become more like
what they intend or want to experience, and then keeping it that way even in a
changing world.�

“If they plan, they plan perceptions, not actions.
If they respond to stimuli, they do so in order to prevent those stimuli from
affecting variables they have under control.�

“To me, this theory is beautiful, ot because it is mine
but because it is true.�

Chapter 1: Concepts of Control

Page 1: “To control something is to act to bring it
to a specified condition, and then maintain it close to the that condition even
if unpredictable external forces and changes in environmental properties tend
to alter it.�

MCT:
ANALYZE-COMPUTE-ACT CONTROL

Input
leads to actions which leads to the desired result.

The engineers in the Apollo 13 mission had to figure out how to reduce the CO2
level with something that the astronauts could build; they came up with a
system that worked.

Page 3 “…get the flavor of all thhe little problems
that arise in trying to build a system that, all by itself, can analyze the
environment, compute how to achieve a prescribed result, and then act on the
environment in exactly the calculated way.�

Problem: controlling the speed of a car. Bill goes
through all the forces acting on the car. The problem is that the net force
acting on the car changes as the car moves over the terrain.

PERCEIVE-COMPARE-ACT
CONTROL

Page 7: “Negative feedback control as used in PCT
can be summed up very quickly. It involves continuously perceiving the current
state of whatever is to be controlled, continuously comparing that perception
with the intended state, and continuously acting to reduce the difference
between perceived and intended states , keeping the difference as close to zero
as possible.�

Page 9: “A negative feedback controller, or
perceive-compare-act controller, or PCT controller doesn’t have to know what is
causing the speed to change.�

“Most of what has been said in the literature of the
life sciences about limitations of negative feedback control is simple false.�

SIMULATION
AND MODELING

The two ways of calculating the time to fill a
bucket is discussed.

Simulation is advocated over analytical solutions.

Readers may want to download a free version of
VenSim:

http://www.shodor.org/tutorials/vensim/pre.php

and follow a simulation of how a temperature control system works
in Vensim:

http://plaidcircuitry.com/computational_thinking/vensim/index.html

SOME
PHILOSOPHY OF SCIENCE

Page 17: “…the way thhat collections of materials
behave depends as much on how they are organized as on what the materials are.�

“You can’t cause a mouse to jump in the air by
itself if it doesn’t want to.�

David