Subject:
|
Re: localization (was re: GPS)
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 12 Jan 2006 04:34:07 GMT
|
Original-From:
|
steve <sjbaker1@airmail.net/NoSpam/>
|
Viewed:
|
2110 times
|
| |
|
|
Roger Glassey wrote:
> Hi John,
> A localization scheme I have used with the RCX is uses the angles between light
> beacons in known locations, measured with a rotating light sensor. Two angles
> (3 beacons) are enough to calculate, with a bit of trig, the robot coordinates
> and heading. The robot needs to know its position and heading accurately
> enough to correctly determine which beacon is represented by a peak in light
> intensity.
> The accuracy of the scheme depends on the angle sensor accuracy and the geometry
> of the robot position relative to the beacons. I localize the robot to within
> about 8 inches, with beacons at the corners of an 8 x 16 foot rectangle.
> The new light sensor that can distinguish colors with a stepper motor should
> make the problem easier.
There are lots of interesting varients on that theme:
1) A lateral thinking way to do that is to reverse the situation. Put a
rotating laser pointer on top of the robot and hook up light sensors
in the corners of the room to another (stationary) RCX. The second
RCX can tell when each light sensor is hit by the light - and can
send that to the robot which knows where the laser is pointing when
it gets the signal. The advantages are twofold:
a) The problem of knowing which beacon you are seeing is eliminated.
b) The robot is simpler - although the stationary equipment is more
complex.
2) You can put a light sensor on the robot - looking vertically upwards
at the point of an inverted conical mirror. This allows it to 'see'
in all directions. Now you can put rotating laser pointers on just
two beacons - with a second stationary RCX that knows where the two
pointers are facing. The stationary RCS sweeps first one laser,
then the other slowly across the room - and continually broadcasts
IR messages to tell the robot where the lasers are pointing. The
robot will therefore know which laser is at what angle when it
lights up the light sensor. Simple trig gets you the answer.
a) You only need two static beacons - not three.
b) The robot only needs to use one sensor port and no motors,
so it's small and simple.
c) Many robots can pick up the signal at the same time.
|
|
Message is in Reply To:
 | | Re: localization (was re: GPS)
|
| Hi John, A localization scheme I have used with the RCX is uses the angles between light beacons in known locations, measured with a rotating light sensor. Two angles (3 beacons) are enough to calculate, with a bit of trig, the robot coordinates and (...) (19 years ago, 12-Jan-06, to lugnet.robotics)
|
12 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
