Subject:
|
Re: Inertial guidance
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Sat, 12 Jan 2002 19:47:13 GMT
|
Original-From:
|
Steve Baker <SJBAKER1@AIRMAILnomorespam.NET>
|
Reply-To:
|
sjbaker1@=AvoidSpam=airmail.net
|
Viewed:
|
1450 times
|
| |
|
|
Trevyn Watson wrote:
> Steve Baker wrote:
>
> > You also generally need six sensors in order to measure the accellerations
> > in all six degrees of freedom - that's not gonna be nice for an RCX!
>
>
> Actually, i think you could use TWO rotation sensors because the
> rotation sensors work in both directions, and Lego generally doesnt go
> up and down (but it might if you had some sort of maze with ramps).
By "Six degrees of freedom" I mean:
1) Translation to Left/Right
2) Translation Forwards/Backwards
3) Translation Up/Down
4) Rotation left/right
5) Rotation fore/aft
6) Roll left/right.
These are often called (respectively) by the names:
Sway, Surge, Heave, Yaw, Pitch, Roll.
You need to measure them all - even if you are only interested in
your position in two-dimensions because if you go up even a gentle
slope, your pitch angle causes some of your Heave motion to affect
your 2D position along the ground and also causes your Surge sensor
to feel a small amount of the effect of gravity (because it's not
laying completely flat). All of this and more has to be accounted
for in your readings. Remember that your measurements are always relative
to the vehicle (and not to the ground) because that is where the sensor
is mounted.
It is in the nature of inertial measurement (and it's great sensitivity
to cumulative errors) that this kind of thing is important. If you
only had surge and sway sensors, then driving downhill on any *tiny* slope would
feel like a tiny forward accelleration to the surge sensor. That would
cause it to continually increase your robot's idea of it's forward speed
for as long as you stayed on the slope (even if you were stationary!),
hence your idea of your forward position would be wildly incorrect in
a very short space of time.
These sensors have to be so sensitive in order to work at all usefully
that even the amount by which your living room's floor is not perfectly
level would be enough to have your robot thinking it's dozens of meters
away after just a couple of minutes parked stationary on that very slight
slope.
Unless you have really high quality accelerometers and laser gyro's, I
don't think inertial navigation is gonna work for Lego robots.
----------------------------- Steve Baker -------------------------------
Mail : <sjbaker1@airmail.net> WorkMail: <sjbaker@link.com>
URLs : http://www.sjbaker.org
http://plib.sf.net http://tuxaqfh.sf.net http://tuxkart.sf.net
http://prettypoly.sf.net http://freeglut.sf.net
http://toobular.sf.net http://lodestone.sf.net
|
|
Message has 1 Reply:
 | | Re: Inertial guidance
|
| Steve, I see your point. I forgot the sensor also responds to gravity. I suspect that the best we can do is a very rough demo of principle, not a practical navigation system. Thanks for your insight! That is what makes this forum so useful. - pete. (...) (22 years ago, 13-Jan-02, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Inertial guidance
|
| (...) Actually, i think you could use TWO rotation sensors because the rotation sensors work in both directions, and Lego generally doesnt go up and down (but it might if you had some sort of maze with ramps). (22 years ago, 12-Jan-02, to lugnet.robotics)
|
9 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
