[DPRG] offroad robot exercises  11 Nov 06
Subject: [DPRG] offroad robot exercises  11 Nov 06
From: David Peterson
robodave1 at tx.rr.com
Date: Thu Nov 2 01:30:21 CST 2006
Yes, that was my point, my math was "graph paper" math, standard rectangular
to polar, not taking into account changing distance meaning of degree
measure at different latitudes. What I really hoped to clarify was the
correct degree direction measure. Because my distance estimate was off, I'm
sure the degree "heading" calculation was off as well. And I had not looked
for a formula yet to calculate the appropriate direction (heading, bearing)
based on a set of GPS points. It would seem that each new heading should be
90 degrees from the last, but is there a formula for calculating the initial
direction? I don't have a tool like "dlaz" (guessing it is a Linux program)
It is interesting that the headings below are not 90 degrees from each
other. Pretty close though. Should we use the 110.99 to 197.30 to 291.63 to
15.66 degree headings? (Assuming of course I actually have a robot respond
to compass headings apprpriately).
Dave
 Original Message 
From: "David P. Anderson" <dpa at io.isem.smu.edu>
To: "Comcast email" <robodave1 at comcast.net>
Cc: <dprglist at dprg.org>
Sent: Thursday, November 02, 2006 12:50 AM
Subject: Re: [DPRG] offroad robot exercises  11 Nov 06
> Howdy
>
> Hi Dave,
>
> This is just a simple 100 foot square, so no GPS is really needed. These
> points are provided as an option for robot builders that may need them.
> You can probably do just fine with a compass and some wheel encoders,
> no GPS required. That's basically how jBot does it.
>
> Howsoever, I'm not following your math.
>
> I use a seismologists' tool called "dlaz" that returns azimuth and angular
> distance between two lat/lon points. Here's what I get for the four
> parking
> lot GPS waypoints:
>
> From point 1 to point 2 in lat/lon is:
>
> 32.78228 096.76082 to 32.78218 96.76051
>
> which is: 110.99 deg azimuth at 31.02 meters
> or 101.77 feet or 1221 inches
>
> From point 2 to point 3 in lat/lon is:
>
> 32.78218 096.76051 to 32.78191 096.76061
>
> which is: 197.30 deg azimuth at 31.42 meters
> or 103.09 feet or 1237 inches
>
> From point 3 to point 4 in lat/lon is:
>
> 32.78191 096.76061 to 32.78201 096.76091
>
> which is: 291.63 deg azimuth at 30.15 meters
> or 98.91 feet or 1187 inches
>
> From point 4 to point 1 in lat/lon is:
>
> 32.78201 096.76091 to 32.78228 096.76082
>
> which is: 15.66 deg azimuth at 31.16 meters
> or 102.23 feet 1227 inches
>
> My GPS readings were all + 3 feet so that seems like pretty good
> agreement
> with these numbers. It's not an exact square, but for the Borenstein
> UMBMark tests it doesn't have to be. I'll go over that in more detail
> at the meeting Saturday.
>
> best regards,
> dpa
>
>
>
>> A rough estimate of degree directions based on GPS coordinates would be :
>> 252.12 degrees for 0.0003257 degree measure, ~100 feet? from point 1 to
>> point 2
>> 159.68 degrees for 0.0002879 from 2 to 3
>> 71.57 degrees for 0.0003162 from 3 to 4
>> 161.57 degrees for 0.0002846 from 4 to 1
>>
>> Given the variations from 0.0003257 to 0.0002846 in decimal minute or
>> degree
>> measure from GPS readings to estimate 100 feet, can there be any surety
>> of
>> degree direction?
>>
>> This was basic rectangular to polar conversion, but I'm probably
>> neglecting
>> the changing measure of degree distance as latitude changes. And also did
>> not consider the changes in magnetic North based on position, thinking
>> it's
>> about 11 degree difference.
>>
>> I had hoped to navigate by magnitude and direction (magnetic compass)
>> readings, though do not have it finished yet. The compass can be
>> calibrated
>> to whatever is defined as North though.
>>
>> Perhaps there can be another definition of points to navigate, maybe ...
>> traffic cones? : )
>>
>> Dave
>>
>>
>>  Original Message 
>> From: "David P. Anderson" <dpa at io.isem.smu.edu>
>> To: <dprglist at dprg.org>
>> Sent: Wednesday, November 01, 2006 2:18 AM
>> Subject: [DPRG] offroad robot exercises  11 Nov 06
>>
>>
>> > Howdy
>> >
>> > For the upcoming DPRG meeting at The Science Place
>> > on 11 November 2006, I will be presenting a short
>> > program on the the Borenstein UMBMark test and
>> > how to use it to calibrate your robot.
>> >
>> > Thereafter we will take robots to one of two asphalt
>> > parking lots on the Fair grounds to run the offroad
>> robot exercises for whomever may want test out their
>> > robots.
>> >
>> > Scot Sumner suggested a large parking lot on the north
>> > side of the Fairgrounds near the Railroad Museum, which
>> > looks ideal for us, and he says is usually empty. Here
>> > are some satellite photographs and GPS coordinates for
>> > that parking lot.
>>
>> > <http://geology.heroy.smu.edu/~dpawww/robots/dprg/11nov06/>
>> >
>> > best
>> > dpa
>
>
> 
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.409 / Virus Database: 268.13.21/511  Release Date: 11/1/2006
>
More information about the DPRG mailing list
