 |
WiiLi.org
a new revolution
|
| View previous topic :: View next topic |
| Author |
Message |
TiagoTiago
Joined: 20 Jan 2007
Posts: 710
Location: Brasil
Digg It |
 Posted: Tue May 01, 2007 8:39 pm Post subject: |
|
|
the point I'm tryign to get at, isn't the techinique I sugested better than trying to guess based on just one trio of accelerometers?
_________________
please put the scripts on the wiki so they dont get lost as new stuff is posted!
phpBB doesnt like me,somtimes it will forget to warn me about new replies to threads I asked it to,if you see a thread I should have responded, could please email me? |
|
| Back to top |
|
 |
slash
Joined: 31 Mar 2007
Posts: 10
Digg It |
| Posted: Sun Oct 14, 2007 5:38 am Post subject: |
|
|
wait, thats a way to not need a sensor bar? (i is confused)
_________________
~/Slash/~ |
|
| Back to top |
|
|
Classiclll
Joined: 10 Feb 2007
Posts: 94
Location: Japan
Digg It |
| Posted: Thu Oct 18, 2007 2:32 pm Post subject: Re: a way to possibly get rotational acceleration |
|
|
| TiagoTiago wrote: | shouldn't it be possible to read rotational acceleration by combining a second trio of accelerometers in offset and comparing the acceleration on the same axis of them?
putting two wiimotes or a wiimote and a nunchuck taped together (or in some other way fixed together) |
i'm thinking about your idea, too.
there may be some theoritical and sensing limitation.
i'm trying to examine this limitation.
_________________
Classiclll |
|
| Back to top |
|
|
TiagoTiago
Joined: 20 Jan 2007
Posts: 710
Location: Brasil
Digg It |
| Posted: Thu Oct 18, 2007 6:00 pm Post subject: |
|
|
coolies! keep me update plz 
_________________
please put the scripts on the wiki so they dont get lost as new stuff is posted!
phpBB doesnt like me,somtimes it will forget to warn me about new replies to threads I asked it to,if you see a thread I should have responded, could please email me? |
|
| Back to top |
|
|
Classiclll
Joined: 10 Feb 2007
Posts: 94
Location: Japan
Digg It |
| Posted: Wed Nov 14, 2007 1:42 pm Post subject: |
|
|
hi TiagoTiago
this is a only preparation for starting some discussion.
any comments are welcome.
| Code: | <the problem definition>
let
* sFt : sensed force vector (at time t), in terms of G's
* mFt : moving force vector
* tFt : tangent force vector of self-rotation
* cFt : centrifugal force vector of self-rotation
* gF : gravitational force vector (always toward the Earth)
* dt : time duration from last sensing (at time t)
then
in general
sFt = mFt + tFt + cFt +gFt
now we want to devide 4 forces from the sensed force vector sFt
once done, we ca obtain
Velocity at t, wVt <- sum(mFt*dt)
Location at t, wLt <- sum(wVt*dt)
and
Rotational Velocity at t, wRt <- sum(tFt*dt)
Attitude at t, wAt <- sum(wRt*dt)
* attitude is consist of two unit vector, toBottom and toFront.
<in case of stable assumption>
we can suppose mFt=0, cFt<<0, tFt<<0 under the assumption.
so we obtain sFt==tFt+gF(t-1), gFt == sFt, and
attitude at t, wAt.rotate(axis,angle)
where axis is a unit vector of cross product of sFt and toBottom
with my 3D vector class "Loc", wSt.cross(toBottom).unit()
or simply
roll of wii at t, roll == asin(sFt.x)
pitch of wii at t, pitch == asin(sFt.y) |
satble analisys code is followed
http://www.geocities.jp/classiclll_newweb/AttitudeAnalysis1/applet/index.html
[edit] modified the statements of stable assumption. (nov.18)
_________________
Classiclll
Last edited by Classiclll on Sun Nov 18, 2007 12:59 pm; edited 5 times in total |
|
| Back to top |
|
|
TiagoTiago
Joined: 20 Jan 2007
Posts: 710
Location: Brasil
Digg It |
| Posted: Wed Nov 14, 2007 6:25 pm Post subject: |
|
|
| slash wrote: | | wait, thats a way to not need a sensor bar? (i is confused) |
with the proper algorithm and if the measurements are accurate enough yes, this could help with acquiring some of the data that usually is acquired with the help of a sensor bar
the idea is that two different points in a rotating object will accelerate differently, knowing how to interpret that difference should allow to get a better reading of how the wiimote is being rotated than what we currently have
_________________
please put the scripts on the wiki so they dont get lost as new stuff is posted!
phpBB doesnt like me,somtimes it will forget to warn me about new replies to threads I asked it to,if you see a thread I should have responded, could please email me? |
|
| Back to top |
|
|
Classiclll
Joined: 10 Feb 2007
Posts: 94
Location: Japan
Digg It |
| Posted: Sun Nov 18, 2007 4:07 am Post subject: |
|
|
| Code: | <modeling the TiagoTiago's idea>
now we have a couple of accelerometers, two remotes, or remote w/t nunchuck.
let
* xXtn : xXt is a previously-defined vectors, and n(=0,1) for distinction
* rod : virtual rod between our accelerometers, rod.norm() is rod length.
and supposing
1. both accelerometers are on the same coordinate.
2. sensing are done simultaneously.
<using TiagoTiago's model>
by subtracting sFt1 from sFt0 we can obtain the self-rotatonal force vector of the rod.
let
rFt = sFt0-sFt1, then rFt == (tFt0+tFt1) + (cFt0 +cFt1)
(please examine above, by drawing force vectors on 2D space = paper.)
now if we makeup the situation of 1. and 2., and if we measure the rod vector value,
we can distinct the tFt0+tFt1 to obtain the self-rotation of wiiremote, by projection of rFt to rod.
using self-rotation vector, in theory, we can get the gFt without stable assumption.
and we can distinct the mFt, the target acceleration. |
next, i will try to establish 1. and estimate rod's value by some calibration, and to get the yaw without the IRSenser information.
_________________
Classiclll |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
