# My Way: (was Thumbs up? (was Surfs up and Monitoring systems))

[From Rick Marken (2012.04.30.1500)]

Rupert Young (2012.04.29 21.30 BST) --

Hi Rupert

Here's my approach. I just have two control systems, one controlling
the speed of the boat and the other controlling the bearing. The
system controlling bearing does it by varying the rudder and the
system controlling speed does it by varying the angle of the sail
(relative to the carlin -- center line-- of the boat). These outputs
both affect the states of the environmental variables being controlled
(boat direction and boat speed) as do the two disturbance, wind
direction and wind speed.

Right not the physics of the situation is completely make believe;
that's the hard part. But once I have that done I should have two
control systems that will sail the boat in a fixed direction at a
fixed speed (and into the wind). Hierarchical control will come in if
I can figure out how to design control systems that can do tacking.

See what you think. By the way, you run mine by hitting F9. You'll see
that both control systems keep their perceptions matching their
references in the fact of randomly varying changes in wind direction
and speed.

Best

Ricky the sailor man

Sailing Sim.xls (17 KB)

···

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

Content-Type: application/vnd.ms-excel; name="Sailing Sim.xls"
Content-Disposition: attachment; filename="Sailing Sim.xls"
X-Attachment-Id: f_h1o26hck1

Hi, Rick --

I got nothing but Error 523 everywhere:
"The calculation procedure does not converge'".

Bill

[From Rick Marken (2012.04.30.1830)]

Hi, Rick --

I ï¿½got nothing but Error 523 everywhere:
"The calculation procedure does not converge'".

It may be a calculation option thing. I presume you are not using
Excel; maybe Open Office? Anyway, if they have an "options" menu for
"Calculation" check the "Iteration" selection (turn Iteration on) and
set Max Iterations to 100 and Max size of an iteration to .001 (that's
what I have it set to in Excel). If that doesn't work (or if
"Iteration" is already checked) then I have no idea what's going on.
But the model is in the very early stages. As I said, the physics part
is not correct (or even close) yet. I probably won't have that working
for some time, given my math/physics skills. But I'll get it
eventually. Then we can see if we can get your spreadsheet program to
run it properly.

Best

Rick

···

On Mon, Apr 30, 2012 at 3:31 PM, Bill Powers <powers_w@frontier.net> wrote:
--
Richard S. Marken PhD
rsmarken@gmail.com

[From Rick Marken (2012.05.01.2245)]

Bill Powers said:

Hi, Rick –

The options were set as you described. Here is what I see on
starting the program.

Boy, that is ugly. Has anyone else tried it? It works for me but maybe it only works in Excel.

BP: It takes me forever to figure out what spreadsheet equations
mean.

I know. Me too. I probably shouldn’t have posted it yet; I plan to add graphics and stuff that will help people see what’s going on. It’s in a very primitive state still. I posted it only to illustrate the general architecture I would use to model sailing; just have systems that control speed and direction of the boat; those are the main things I control when I sail. I sail in order to get from point A to point B at a nice clip.

I put the model in spreadsheet form because Rupert’s model was also a spreadsheet so I presumed he could compare his to mine pretty easily. I also like spreadsheet modeling because everyone with a spreadsheet should have no problem running it (though you have proved that that’s not true so that reason for using a spreadsheet is no longer in force).

I could write the program using visual basic (now available again in the latest versions of Excel), which makes it much easier to see what the program is doing. So I may transfer some of the code to visual basic. But it will take me a while to finish the program so I wouldn’t waste any mroe time trying to understand this program. If I ever get a good version of the sailing model working I’ll explain how it works it in detail.

Best

Rick

···

Richard S. Marken PhD
rsmarken@gmail.com

This is due to circular references. In all the models I've been sending out I got this error, but only in OO, in excel it works. I have yet to work out how to resolve it in OO.

···

On 30/04/2012 23:31, Bill Powers wrote:

Hi, Rick --

I got nothing but Error 523 everywhere:
"The calculation procedure does not converge'".

Bill

--

Regards,
Rupert

[From Richard Kennaway (2012.05.02.1020 BST)]

[From Rick Marken (2012.05.01.2245)]

Bill Powers said:

Hi, Rick --

The options were set as you described. Here is what I see on starting the program.

Boy, that is ugly. Has anyone else tried it? It works for me but maybe it only works in Excel.

In Excel 2011, on a Mac, I got the message "This workbook contains defined names that conflict with cell references." After which, it opened, and as far as I can tell, hitting F9 did what it should. In Excel 2003 on a PC, it worked with no errors or warnings.

···

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

[From Rupert Young 2012.05.02 12.30 BST]

[From Rick Marken (2012.04.30.1500)]

Hi Rupert

Here's my approach. I just have two control systems, one controlling
the speed of the boat and the other controlling the bearing. The
system controlling bearing does it by varying the rudder and the
system controlling speed does it by varying the angle of the sail
(relative to the carlin -- center line-- of the boat). These outputs
both affect the states of the environmental variables being controlled
(boat direction and boat speed) as do the two disturbance, wind
direction and wind speed.

Right not the physics of the situation is completely make believe;
that's the hard part. But once I have that done I should have two
control systems that will sail the boat in a fixed direction at a
fixed speed (and into the wind). Hierarchical control will come in if
I can figure out how to design control systems that can do tacking.

See what you think. By the way, you run mine by hitting F9. You'll see
that both control systems keep their perceptions matching their
references in the fact of randomly varying changes in wind direction
and speed.

That looks pretty good; it works for me in excel but not in OO.

I did notice though that the environment values for direction and speed made big jumps between iterations when the reference was changed. But, as you say, that probably just needs some work on modelling the physical environment. Also if the wind is set to zero the speed still converges to its reference (I'd set it to 40), so I guess you need to model deceleration of the boat when the wind is absent.

It looks like it will be an interesting project and I look forward to the next version.

Our goals are different though I think. I wasn't particularly trying to model sailing but have been using sailing as an example of a situation where an "uncontrolled" perception is essential for a control system to achieve its higher level aims. I think the architecture of my last model, where the wind control unit was above the boat motion unit, makes a profound point (for me anyway) about how control systems are arranged in situations which, on the face of it, looks like stimulus/response behaviour as well as giving some insight on the relation between evolution and the hierarchy of control systems. So I would be interested in any comments you (or anyone else) may have on this. Noone has responded to this as yet so maybe people, didn't follow the point I was trying to make, or thought it was nonsense, or obvious, or naive, or wrong, or not interesting. If required I can go into more detail.

···

--

Regards,
Rupert

Rupert, check out this video by Iain McGilchrist and tell me if you see a connection between Iain's observations and yours regarding "a situation where an 'uncontrolled' perception is essential for a control system to achieve its higher level aims."

The Divided Brain and the Making of the Western World

Cheers,

Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Rupert Young <rupert@MOONSIT.CO.UK> 5/2/2012 7:17 AM >>>

[From Rupert Young 2012.05.02 12.30 BST]

[From Rick Marken (2012.04.30.1500)]

Hi Rupert

Here's my approach. I just have two control systems, one controlling
the speed of the boat and the other controlling the bearing. The
system controlling bearing does it by varying the rudder and the
system controlling speed does it by varying the angle of the sail
(relative to the carlin -- center line-- of the boat). These outputs
both affect the states of the environmental variables being controlled
(boat direction and boat speed) as do the two disturbance, wind
direction and wind speed.

Right not the physics of the situation is completely make believe;
that's the hard part. But once I have that done I should have two
control systems that will sail the boat in a fixed direction at a
fixed speed (and into the wind). Hierarchical control will come in if
I can figure out how to design control systems that can do tacking.

See what you think. By the way, you run mine by hitting F9. You'll see
that both control systems keep their perceptions matching their
references in the fact of randomly varying changes in wind direction
and speed.

That looks pretty good; it works for me in excel but not in OO.

I did notice though that the environment values for direction and speed
made big jumps between iterations when the reference was changed. But,
as you say, that probably just needs some work on modelling the physical
environment. Also if the wind is set to zero the speed still converges
to its reference (I'd set it to 40), so I guess you need to model
deceleration of the boat when the wind is absent.

It looks like it will be an interesting project and I look forward to
the next version.

Our goals are different though I think. I wasn't particularly trying to
model sailing but have been using sailing as an example of a situation
where an "uncontrolled" perception is essential for a control system to
achieve its higher level aims. I think the architecture of my last
model, where the wind control unit was above the boat motion unit, makes
a profound point (for me anyway) about how control systems are arranged
in situations which, on the face of it, looks like stimulus/response
behaviour as well as giving some insight on the relation between
evolution and the hierarchy of control systems. So I would be interested
in any comments you (or anyone else) may have on this. Noone has
responded to this as yet so maybe people, didn't follow the point I was
trying to make, or thought it was nonsense, or obvious, or naive, or
wrong, or not interesting. If required I can go into more detail.

···

--

Regards,
Rupert

[From Rupert Young 2012.05.03 15.10 BST]

···

On 02/05/2012 18:34, Chad Green wrote:

Rupert, check out this video by Iain McGilchrist and tell me if you see a connection between Iain's observations and yours regarding "a situation where an 'uncontrolled' perception is essential for a control system to achieve its higher level aims."

The Divided Brain and the Making of the Western World

Interesting video, though I didn't see anything that was particularly relevant to the points I was raising; perhaps I missed something.

--

Regards,
Rupert

No worries. Doesn't anyone else see the connection (i.e., pattern)?

Let me put it another way.

In the journal article "Behavior, Purpose and Teleology" (Rosenblueth, Wiener, and Bigelow, 1943) (http://pespmc1.vub.ac.be/Books/Wiener-teleology.pdf), the authors subdivide active behavior into two classes: purposelessness (randomness) and purposefulness.

On a related note, on April 30, Bill said:

"First, to a control system there is no such thing as noise. There are just variables that vary, sometimes too fast to keep up with and sometimes in ways that allow maintaining good control. The control system doesn't know or care about waveforms so either everything looks random or nothing does. Noise is just a classification by an observer of waveforms that he can't explain or predict in detail."

Is he saying the same thing as the authors above, namely, that behavior is either random (purposeless) or purposeful? Note that he prefaced the paragraph above with this statement: "A couple of random comments which may or may not correlate with anything."

Does this mean he was not purposeful in the comments that followed? Is it that black and white? Likewise, in an e-mail to you on 4/13 Bill ended the message with:

"All these random-seeming thoughts are things to keep in mind as you try solve the sailboat problem. I assume that when you ask how you would arrange things,. you're not just asking for the answer. You're asking how you would find the answer if I got hit by lightning before I could do it for you. I'm trying to tell you how. In mentioning S-R theory, I'm also trying to tell you why we want to find a control-system answer. There's no single right answer, because there are usually many ways to solve the same problem. All you need is one that works under all conditions."

So was Bill likewise purposeless in that message?

I didn't expect to go in this direction in this e-mail but here we are.

Cheers,

Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Rupert Young <rupert@MOONSIT.CO.UK> 5/3/2012 10:06 AM >>>

[From Rupert Young 2012.05.03 15.10 BST]

···

On 02/05/2012 18:34, Chad Green wrote:

Rupert, check out this video by Iain McGilchrist and tell me if you see a connection between Iain's observations and yours regarding "a situation where an 'uncontrolled' perception is essential for a control system to achieve its higher level aims."

The Divided Brain and the Making of the Western World

Interesting video, though I didn't see anything that was particularly relevant to the points I was raising; perhaps I missed something.

--

Regards,
Rupert

[From Rupert Young 2012.05.05 10.00 BST]

···

On 03/05/2012 18:36, Chad Green wrote:

Does this mean he was not purposeful in the comments that followed? Is it that black and white?

No. Yes.

--

Regards,
Rupert

Rupert, you could have just as easily answered: It depends.

Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Rupert Young <rupert@MOONSIT.CO.UK> 5/5/2012 5:01 AM >>>

[From Rupert Young 2012.05.05 10.00 BST]

···

On 03/05/2012 18:36, Chad Green wrote:

Does this mean he was not purposeful in the comments that followed? Is it that black and white?

No. Yes.

--

Regards,
Rupert

OK, here's a relevant article on the supposed "unknown forces" in
control theory:

Researchers unlock mystery of how 'handedness' arises
http://phys.org/news/2012-05-life-scientists-mystery-handedness.html

"We discovered that just two physical ingredients�entropy and

particle shape�are enough to cause chirality to appear spontaneouslly
in dense systems."

Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Rupert Young <rupert@MOONSIT.CO.UK> 5/5/2012 5:01 AM >>>

[From Rupert Young 2012.05.05 10.00 BST]

···

On 03/05/2012 18:36, Chad Green wrote:

Does this mean he was not purposeful in the comments that followed?

Is it that black and white?

No. Yes.

--

Regards,
Rupert

systems))

Hi, Rick –

The options were set as you described. Here is what I see on
starting the program.

Note on the bottom that it says “Error: calculation does not
converge”. So something must not be initializing correctly.

It takes me forever to figure out what spreadsheet equations
mean.

Let’s see. “Rudder angle”, I assume, applies to D5:

RudderAngle = RudderAngle+\$A\$4*(\$B\$4*(D3-D4)-RudderAngle)

A\$4\$ looks like a slowing factor – yes, there’s a label above saying
slowing, and to the right is gain, which seems to be B\$4\$. So I
have

RudderAngle = RudderAngle + Slowing*(gain*(D3 - D4) -
RudderAngle.

Now what are D3 and D4? Apparently D3 is BearingReference (Br) and D4 is
BearingPerception (Bp) so

BearingError = Br - Bp

RudderAngle = RudderAngle + Slowing*(gain*(BearingError) -
RudderAngle).

Rudder angle is set to Gain*BearingError, approaching that value after
many iterations. If Gain is 20 and BearingError is 10, the rudder will
turn to an angle of 200 (degrees? doesn’t seem likely)

Ah, then we have BoatDirection in D6, and it is equal to

BoatDirection = RudderAngle+E5+D8+E8

What is E5? To the right is a label saying Sail Angle.

BoatDirection = RudderAngle+SailAngle+D8+E8

D8 is wind direction, The formula is D18*500

E8 is wind amplitude. E8 = 500*E18

So we have

BoatDirection = RudderAngle+SailAngle+500WindDirection +
500
WindAmplitude

How come the boat direction doesn’t just keep changing as long as the
rudder angle isn’t zero?

Now, what are D18 and E18?

We have some smoothing or slowing:

\$B\$11 is 100

WindDirection = WindDirection + (D17 - WindDirection)/100

D17 = D17 + (D16 - D17)/100

D16 = D16 + (D15 - D16)/100

D15 = D15 + (D14 - D15)/100

\$B\$14 is 0.5

D14 = 0.5 - D23

D23 = D22/36523

D22 = MOD((3 * (D22 + 1); 36523) What the heck is
this?<--------------------------???

what is 36523?

Sorry, I’m stumped. Is the D22, D23 stuff a function? I don’t see any
place where D22 gets set to a value.

Over to you.

Best,

Bill P.

(Attachment a9fd45.jpg is missing)

···

To: “Control Systems Group Network (CSGnet)”
CSGNET@LISTSERV.ILLINOIS.EDU, CSGNET@LISTSERV.ILLINOIS.EDU

From: Bill Powers powers_w@frontier.net

Subject: Re: My Way: (was Thumbs up? (was Surfs up and Monitoring