Go to Post Don’t choose to aim for the stars because they mark the peak of our abilities, don’t choose to aim for the stars because it is easy, choose to aim for the stars because it shows the depth of our potential; choose to aim for the stars because it is hard! - Ken Leung [more]
Home
Go Back   Chief Delphi > Technical > Programming
CD-Media   CD-Spy  
portal register members calendar search Today's Posts Mark Forums Read FAQ rules

 
Closed Thread
Thread Tools Rate Thread Display Modes
  #61   Spotlight this post!  
Unread 08-01-2009, 13:34
Unsung FIRST Hero
Mike Betts Mike Betts is offline
Electrical Engineer
no team
Team Role: Engineer
 
Join Date: Dec 2001
Rookie Year: 1995
Location: Homosassa, FL
Posts: 1,442
Mike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Tom Bottiglieri View Post
Has anyone thought about using an FIR filter on joystick inputs? It would work for acceleration from a stop in either direction (F/R), but I am stuck on how change directions or stop without considerable lag.

Thoughts?
Tom,

The FIR filter should work just as well when changing directions. Indeed, you want to back off the throttle from the one direction before increasing in the other direction. Of course, you may want to try using the new Jaguars for this. Also, something other than constant coefficients for the FIR would be interesting...

JMHO.

Mike
__________________
Mike Betts

Alumnus, Team 3518, Panthrobots, 2011
Alumnus, Team 177, Bobcat Robotics, 1995 - 2010
LRI, Connecticut Regional, 2007-2010
LRI, WPI Regional, 2009 - 2010
RI, South Florida Regional, 2012 - 2013

As easy as 355/113...
  #62   Spotlight this post!  
Unread 08-01-2009, 13:47
Tom Bottiglieri Tom Bottiglieri is offline
Registered User
FRC #0254 (The Cheesy Poofs)
Team Role: Engineer
 
Join Date: Jan 2004
Rookie Year: 2003
Location: San Francisco, CA
Posts: 3,187
Tom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond reputeTom Bottiglieri has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Mike Betts View Post
Tom,

The FIR filter should work just as well when changing directions. Indeed, you want to back off the throttle from the one direction before increasing in the other direction. Of course, you may want to try using the new Jaguars for this. Also, something other than constant coefficients for the FIR would be interesting...

JMHO.

Mike
It would be interesting to see an analysis on real output speed vs allowed acceleration on the rover wheel/CIM setup. Physics says max acceleration should always be the same, but I tend to remember CIMs being kind of sloshy in their top end. This may break a linear filtering system and limit your max acceleration.

This is all based on the non linearity of PWM duty cycle to torque in the CIM. Anyone with knowledge on this, care to share?

On a side note: I rather like the "Middle Income" situtation Paul suggested. I just don't think input filtering is enough to be super useful.
  #63   Spotlight this post!  
Unread 08-01-2009, 14:16
Unsung FIRST Hero
Mike Betts Mike Betts is offline
Electrical Engineer
no team
Team Role: Engineer
 
Join Date: Dec 2001
Rookie Year: 1995
Location: Homosassa, FL
Posts: 1,442
Mike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond reputeMike Betts has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Tom Bottiglieri View Post
...This is all based on the non linearity of PWM duty cycle to torque in the CIM. Anyone with knowledge on this, care to share?...
Tom,

Check this out and you will see why I suggested the Jags... However, I have not been able to do any testing myself yet...

Regards,

Mike
__________________
Mike Betts

Alumnus, Team 3518, Panthrobots, 2011
Alumnus, Team 177, Bobcat Robotics, 1995 - 2010
LRI, Connecticut Regional, 2007-2010
LRI, WPI Regional, 2009 - 2010
RI, South Florida Regional, 2012 - 2013

As easy as 355/113...
  #64   Spotlight this post!  
Unread 09-01-2009, 09:53
Jared Russell's Avatar
Jared Russell Jared Russell is offline
Taking a year (mostly) off
FRC #0254 (The Cheesy Poofs), FRC #0341 (Miss Daisy)
Team Role: Engineer
 
Join Date: Nov 2002
Rookie Year: 2001
Location: San Francisco, CA
Posts: 3,078
Jared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Warren Boudreau View Post
First, I want to thank everyone for such a wonderful open flow of information on this topic. This is such a tremendous paradigm shift from previous years that teams could fall into a "Cold War" mentality and not share this information with others.
This is something I would like to see end once and for all in FIRST.
  #65   Spotlight this post!  
Unread 09-01-2009, 10:16
Paul Copioli's Avatar Unsung FIRST Hero Woodie Flowers Award
Paul Copioli Paul Copioli is offline
President, VEX Robotics, Inc.
FRC #3310 (Black Hawk Robotics)
Team Role: Engineer
 
Join Date: Jun 2001
Rookie Year: 2000
Location: Rockwall, TX
Posts: 1,391
Paul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond reputePaul Copioli has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

All,

Warren is absolutely correct in that my analyssi did not include the dead weight of the trailer. The robot still has to accelerate the trailer, but you also get some tongue weight transfer. The actual equation break down is like this:

F = (Mrobot + Mtrailer) * accel

Ff = (Mrobot + Mtongue) * g * Mu

Setting both sides equal to each other and simplifying:

accel = (Mrobot + Mtongue)/ (Mrobot + Mtrailer) * g * Mu


Let's assume tongue weight is negligible (it is actually really small). The equation breaks down to:

accel = Mrobot * g * Mu / (Mrobot + Mtrailer)

Our estimation of a trailer weight of 40 lbs and a robot weight of 150 lbs (120 + bumpers + battery) puts us at 150 / 190 so the accel is:

accel = 0.79 * g * Mu

Warren's estimate was pretty darn good at 75%. If your robot mass is less, then your ratio gets smaller and you will have a lower available acceleration.

This analysis assumes all of your robot wheels are driving. If you do not have all of your wheels driving, then you will have to decrease your acceleration even more.
  #66   Spotlight this post!  
Unread 09-01-2009, 10:24
cgredalertcc cgredalertcc is offline
Registered User
AKA: Charlie Baxter
FRC #1747 (Harrison Boiler Robotics)
Team Role: College Student
 
Join Date: Dec 2006
Rookie Year: 2005
Location: greenwood indiana
Posts: 69
cgredalertcc is a name known to allcgredalertcc is a name known to allcgredalertcc is a name known to allcgredalertcc is a name known to allcgredalertcc is a name known to allcgredalertcc is a name known to all
Send a message via AIM to cgredalertcc
Re: Implementing Traction Control for an advantage in the 2009 game

In terms of effective application of power, wouldn't using all 4 motors and independently driving each wheel be better for an electronic traction control system? I say this because with a system like that you would more easily be able to isolate the slipping wheel and you would not have to take power away from a wheel that is not slipping, whereas in a one motor per two wheel design if you detect either wheel as slipping your traction control system would take away power from one wheel that is slipping, but possibly also a wheel that is not slipping.

My second thought:

Its been mentioned several times that in the automotive world cars use the anti lock braking system to perform traction control duties. Couldn't a bicycle disk break be used with a pneumatic cylinder driving it to exhibit the same result? Although the system would use a good deal of air, I would think the added driveability would definitely be worth it this year.

My final thought:

If the problem we face is that Traction=Normal Force * Coefficient of friction, and we have now way to increase our coefficient of friction why not increase the normal force? Last year 1741 used a very small vacuum to great effect in capturing balls. Why not apply the same idea to a robot? The area of the robot (28 x 38) is 1064 square inches if even 1/2 psi vacuum is applied to that area that is over 500 pounds of downward force. I know Bill has said that he doesn't like this idea, but he didn't like the idea of launchers last year, and look how many teams did that.

p.s. I know that last part isn't really traction control but its an idea I had that I wanted to throw out there. I guess the thought was who needs traction control if you aren't lacking traction.
  #67   Spotlight this post!  
Unread 09-01-2009, 10:34
IKE's Avatar
IKE IKE is offline
Not so Custom User Title
AKA: Isaac Rife
no team (N/A)
Team Role: Mechanical
 
Join Date: Jan 2008
Rookie Year: 2003
Location: Michigan
Posts: 2,149
IKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Back on a speed sensor (should this kick over to a different thread?), has anyone tried using the Ultrasonics in order to get Doppler effect? I plan on testing this this weekend, but if someone has already got some results to get us headed in the right direction, it would be great.

My thought is that if you use the Ultrasonic distance measuring sensors at an angle, they should measure different distance values for different speeds (doppler effect). What I hope my tests will show is whether or not there is enough range and resolution to get a decent speed measurement.

Addition:
Found this paper describing the physics and moth and testing results from someone elses experimentation.
http://cigr-ejournal.tamu.edu/submis...2001%20007.pdf

Last edited by IKE : 09-01-2009 at 11:46. Reason: Found a cool paper.
  #68   Spotlight this post!  
Unread 09-01-2009, 14:54
Adam Y.'s Avatar
Adam Y. Adam Y. is offline
Adam Y.
no team (?????)
 
Join Date: Mar 2002
Location: Long Island
Posts: 1,979
Adam Y. is a splendid one to beholdAdam Y. is a splendid one to beholdAdam Y. is a splendid one to beholdAdam Y. is a splendid one to beholdAdam Y. is a splendid one to beholdAdam Y. is a splendid one to beholdAdam Y. is a splendid one to behold
Send a message via AIM to Adam Y.
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Abwehr View Post
I think you meant first derivative of wheel velocity (or second derivative of wheel position). This is your acceleration.

Here's the rub - encoders are digital sensors, with limited resolution. The "derivative" operator is continuous. You can approximate the derivative with the "difference" (i.e. accel = speed - last_speed), but this is only ever an approximation.

Luckily, the straight difference isn't the only way to approximate the derivative. For example:

accel = -last_last_speed + 2*last_speed - speed;

This is a smoother derivative approximation, but it is now time-delayed (since it is centered around the time of last_speed). This is the tradeoff of filtering - with smoothness comes time delay.
As Adam Savage, Jamie Hyneman, and countless others have said,"Close enough." Mathematically you will never be able to create a differentiators using analog or digital components. It does not actually stop you from implementing systems like PID controllers. You just have to realize that the differentiators only acts like a differentiators at certain frequencies. I wonder how Labview implements analog filters because it does have the capability to use the Laplace transform but its an analog construct.
Quote:
This may break a linear filtering system and limit your max acceleration.
Going much farther would require the expertise of someone with his PHD. At the undergraduate level you really only learn how to implement linear control systems.
__________________
If either a public officer or any one else saw a person attempting to cross a bridge which had been ascertained to be unsafe, and there were no time to warn him of his danger, they might seize him and turn him back without any real infringement of his liberty; for liberty consists in doing what one desires, and he does not desire to fall into the river. -Mill

Last edited by Adam Y. : 09-01-2009 at 15:06.
  #69   Spotlight this post!  
Unread 09-01-2009, 16:31
Lesman's Avatar
Lesman Lesman is offline
Registered User
FRC #1014 (Bad Robot)
Team Role: Engineer
 
Join Date: Feb 2008
Rookie Year: 2006
Location: Dublin
Posts: 35
Lesman will become famous soon enough
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Qbranch View Post
The coefficients of friction (preliminarily) for static is 0.06 and dynamic is 0.05. If you calculate out how long it would take for a 120lb robot to accelerate to 10ft/sec, you'll find the difference in time to be about 0.1sec. Not much difference.
Not quite:
Max acceleration = CoF * g (ft/s^2)
dynamic: .05 * 32.174 = 1.6087
static: .06 * 32.174 = 1.93044
thus the 0-10ft/s times are:
dynamic: 10/1.6087 = 6.2162
static: 10/1.9304 = 5.1802
difference: 1.0359

That's just a bit over a second, a fairly significant difference. Not to mention the fact that preventing slip will improve handling.
  #70   Spotlight this post!  
Unread 09-01-2009, 17:39
IKE's Avatar
IKE IKE is offline
Not so Custom User Title
AKA: Isaac Rife
no team (N/A)
Team Role: Mechanical
 
Join Date: Jan 2008
Rookie Year: 2003
Location: Michigan
Posts: 2,149
IKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond reputeIKE has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Also at the end of 5 seconds one will be about 4 feet ahead of the other. Of course this is 24 vs 20 feetwhich means you have traveled about half way across the field.
  #71   Spotlight this post!  
Unread 09-01-2009, 18:08
Dominicano0519's Avatar
Dominicano0519 Dominicano0519 is offline
My team is not competing this year
AKA: Demolition Man
FRC #0381 (Tornadoes)
Team Role: College Student
 
Join Date: Feb 2006
Rookie Year: 2006
Location: Trenton
Posts: 165
Dominicano0519 will become famous soon enough
Send a message via MSN to Dominicano0519 Send a message via Yahoo to Dominicano0519
Re: Implementing Traction Control for an advantage in the 2009 game

what might help would be to reverse one of the cims to make them both spin clockwise ( or counterclockwise i forget which way) any way electric motors favor one direction the other direction give less accl.

just a thought

also can someone correct me if im wrong ,because its been a while for me , isn't it just the torque that makes the wheels slip; and if so couldn't you simple create a drivetrain like a 4wd with a cim on each wheel and put a high gear ratio on it.
__________________
Favorite Phrases:
1. I like your robot, no-homo
2. "It depends how you define alone" ( Bill Clinton when being questioned about Monica Lewinsky)

Last edited by Dominicano0519 : 09-01-2009 at 18:16.
  #72   Spotlight this post!  
Unread 09-01-2009, 18:14
jgraber jgraber is offline
Registered User
no team
 
Join Date: Nov 2006
Location: Dallas
Posts: 38
jgraber is just really nicejgraber is just really nicejgraber is just really nicejgraber is just really nice
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Paul Copioli View Post
All,
Warren is absolutely correct in that my analysis did not include the dead weight of the trailer. The robot still has to accelerate the trailer, ...

accel = 0.79 * g * Mu

Warren's estimate was pretty darn good at 75%. If your robot mass is less, then your ratio gets smaller and you will have a lower available acceleration.
Therefore, max weight overpowered robots have most acceleration.
I get max acceleration of about 1.5 fpss, and top speed of 10fps,
which is top wheel speed of about [overstrike]32rpm [/Overstrike] 380rpm

Here are two opposite approaches to gear design to aid traction control:
- use little to no reduction on motors so that the peak motor torque is near the friction x torque on wheel. Highcurrent on slow speed motors.
Note Low rotational momentum of gear train, and high current deltas.
Jaguar speed controllers are reported to be better at low speed control.
I don't like this approach....

- use huge reduction (5000/32?) on motors so that the fuller range of motor speeds can be applied and more torque to drive wheel at exactly right speed.
No reason for wheels to ever go > 380rpm.
Note higher rotational momentum of gear train, but lower current deltas.

Last edited by jgraber : 09-01-2009 at 18:39. Reason: math error
  #73   Spotlight this post!  
Unread 09-01-2009, 19:28
Alan Anderson's Avatar
Alan Anderson Alan Anderson is offline
Software Architect
FRC #0045 (TechnoKats)
Team Role: Mentor
 
Join Date: Feb 2004
Rookie Year: 2004
Location: Kokomo, Indiana
Posts: 9,113
Alan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond reputeAlan Anderson has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by Dominicano0519 View Post
what might help would be to reverse one of the cims to make them both spin clockwise ( or counterclockwise i forget which way) any way electric motors favor one direction the other direction give less accl.
CIM motors have no significant directional bias.
  #74   Spotlight this post!  
Unread 09-01-2009, 20:36
eatbuckshot eatbuckshot is offline
Registered User
FRC #0611
 
Join Date: Jan 2009
Location: Northern VA
Posts: 7
eatbuckshot is an unknown quantity at this point
Re: Implementing Traction Control for an advantage in the 2009 game

I finally have something to say and registered, it seems my OCD for computers ' and its peripherals' performance have a use!
Anyways, about mice: I was very ecstatic after I saw the first ever precise and quantified mouse benchmark published on esreality.com http://www.esreality.com/?a=longpost&id=1265679

In short, All mouses ARE NOT created equal, even if they are optical, laser, wireless, wired, high DPI. Many generic mouses(in my opinion) have a very low what esreality terms "malfunction speed", a maximum threshold speed when the mouse begins to report very inaccurate changes in position causing the cursor to seemingly go crazy or unresponsive. Luckily there are some very good mice made years ago that are still even one of the best today.
this benchmark was done in 2006
http://www.esreality.com/?a=longpost&id=1265679&page=21
This page shows all the mice and their perfect control and malfunction speeds on a graph.The best are the mx500 and mx300 capable of performing at 4 m/s if the usb polling rate is overclocked to 1000hz from 125hz; i'm not sure if the robot controller can do that.. I also have an mx518 that may be able to perform at those speeds according to this post http://forums.logitech.com/logitech/...17117#M17 117
Quote:
Originally Posted by nickmagus View Post
I looked extensively in to optical mice last year however they do not work at high enough speeds to be feasible (even for this challenge) if you find one fast enough please inform me but the best i could get was ~5ft/s before it losses all idea of whats going on.

...
Maybe the mx500, mx300, mx518, a4tech x7


Quote:
Originally Posted by jgraber View Post
If a laser mouse uses a real laser, you should search the rules for the word laser.

I was pleased to see that my idea of using an optical mouse had already been thought of, and even tried, and found to be not effective at speeds > 5f/s. Now I don't have to spend the time to try it myself.

Quote:
Originally Posted by Don Rotolo View Post
... One post mentioned using an optical mouse - my MS optical mouse seems pretty reliable at 1/16" off the desk surface, but it would probably get quickly dirty on a FIRST field. ...
hmm what do you mean by getting dirty? how?
  #75   Spotlight this post!  
Unread 09-01-2009, 23:09
DonRotolo's Avatar
DonRotolo DonRotolo is offline
Back to humble
FRC #0832
Team Role: Mentor
 
Join Date: Jan 2005
Rookie Year: 2005
Location: Atlanta GA
Posts: 7,007
DonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond reputeDonRotolo has a reputation beyond repute
Re: Implementing Traction Control for an advantage in the 2009 game

Quote:
Originally Posted by cgredalertcc View Post
In terms of effective application of power, wouldn't using all 4 motors and independently driving each wheel be better for an electronic traction control system?
Yes, absolutely.
Quote:
Originally Posted by cgredalertcc View Post
Couldn't a bicycle disk break be used with a pneumatic cylinder driving it to exhibit the same result?
I don't think [neumatic cylinders can react fast enough. Even a slow automotive system operates at 15 Hz, far above what a pneumatic system might achieve. Not only are you limited by the valves, air needs to flow and is compressible; pressuretakes time to reach a certain level - this is different from hydraulic systems.
Quote:
Originally Posted by cgredalertcc View Post
Last year 1741 used a very small vacuum to great effect in capturing balls. Why not apply the same idea to a robot?
read Bill's Blog from January 6th.
Quote:
Originally Posted by eatbuckshot View Post
hmm what do you mean by getting dirty? how?
Think dust bunnies: bits of hair, thread, and other crud; the object touching the floor but not rolling will act something like a broom, and stuff on the floor will clog up the optical sensor. Maybe not enough in one round, but it should be considered.
__________________

I am N2IRZ - What's your callsign?
Closed Thread


Thread Tools
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
The 2009 Control System Q&A Thread crake FRC Control System 59 11-01-2009 10:43
Drive train for 2009 game hihihiflcl81pig General Forum 4 27-12-2008 02:17
Buying the 2009 control system BornaE FRC Control System 9 16-10-2008 17:16
The Access Points on the 2009 Control System Shadow503 Rumor Mill 10 28-04-2008 23:22
UNFAIR ADVANTAGE for CDI and new control system archiver 2000 6 23-06-2002 22:13


All times are GMT -5. The time now is 19:22.

The Chief Delphi Forums are sponsored by Innovation First International, Inc.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi