Ask A Scientist© Engineering Archive

Index Key:   ENG017
Author:      eb
Subject:     Triangulation
Text:        I am working on a autonomous robot and I need to figure out a way 
to triangulate its position remotely.  The robot has an on-board data 
acquisition and signal processing capabilities (d/a a/d, and other assorted 
goodies).  The environment is a 100 ft by 100 ft lunar surface (simulated) and 
we need an accuracy of about 3 cm.  Internal mapping is a possibility, but we 
are looking for a way to establish a external reference.  One idea was to set 
up three radio beacons at different frequencies and measure the power 
received.  We could take the FFT of this and measures its strength, hoping the 
power will drop off as a function of 1/d^2.  We experimented with this idea 
using a RF generator and a spectrum analyzer, but there was not a constant 
radiation pattern from our antenna (no 1/d^2 relationship was observed).  The 
room was full of other electronics that may have disturbed our experiment.  
Are we going in the wrong direction?  I have looked into many triangulation 
schemes, but ours seemed the simplest.  Before we invest the money to perform 
the experiment outside of the room that was full of electronics I wanted a 
professional opinion.  I am well aware of the many variables in this situation 
(antenna directivity, atmospheric losses, VSWR problems, etc.).

Response #:  1 of 1
Author:      dipper
Text:        You are right, there are many ways to do this.  The way you are 
doing this may not be that far off the track.  Common Radio telemetry 
nowadays, though, is done by triangulation, not by relative field strength.  
With a single or multiple receivers, you can detect the angular change of the 
strongest portion of the NDB (Non-Directional Beacon) and by simple 
calculations determine its current position and the incremental vector of


movement.  My suggestion, is to use a directional receiving antenna connected 
to an angular shaft encoder.  Rotating the antenna and noting peak signal 
strength at the various frequencies can give you position very well.  This is 
the basic principal of LORAN devices used today.  If you close the position 
loop, you can control the motion on the device.  To be very, very accurate, 
however, I suggest using an inertial guidance device internal to your unit, 
and use the LORAN type radio-telemetry for reference setting and periodic 
checking of position.  I hope this helps.  Without knowing all of the other 
parameters, I may be way off base.