Newsgroups: comp.robotics
Path: brunix!cat.cis.Brown.EDU!agate!howland.reston.ans.net!gatech!udel!MathWorks.Com!yeshua.marcam.com!charnel.ecst.csuchico.edu!csusac!csus.edu!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: 0.1 degree heading sensor
Message-ID: <nagleCtMoLx.EIM@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <CtJ1E7.5sK.3@cs.cmu.edu> <315n46$iv7@usenet.rpi.edu>
Date: Thu, 28 Jul 1994 02:08:21 GMT
Lines: 17

mantoj@cortez.its.rpi.edu (Jason S. Mantor) writes:
>	I worked on a simmilar problem as an undergrduate research project.
>They had a small automobile chassis with an onboard computer and did not know
>how to determine their effective turning radius from the position of their
>steering servo.  It was designed to tow a small trailer so it had an optical
>encoder mounted at the pivot to measure the angle the trailer followed at.
>	I just wrote a program that drove the car in circles at different
>steering angles and recorded the trailer's tracking angles.  If you know the
>geometry of the trailer and the car you can use trig (some of the equations
>need to be solved iteratively !) to calculate the vehicle's turning radius.

      ETAK, in their early car navigation systems, did that with just two
wheel sensors, a magnetic compass, and a map.  Cold calibration took about
fifty miles of uncontrolled driving, after which the system had figured out 
the wheelbase and tire sizes.

					John Nagle
