Re: Anchor alarm
- From: Paul Cooper <a.paul.r.cooper@xxxxxxxxxxxx>
- Date: Thu, 15 Mar 2007 11:50:52 GMT
On 14 Mar 2007 05:42:08 -0700, "toad" <toad_oftoadhall@xxxxxxxxxxx>
wrote:
On 14 Mar, 12:23, Paul Cooper <a.paul.r.coo...@xxxxxxxxxxxx> wrote:
On 14 Mar 2007 04:18:01 -0700, "toad" <toad_oftoadh...@xxxxxxxxxxx>
wrote:
Think about it - your automotive GPS works fine at high speed in rain
and under wet trees and even in towns with tall buildings and that's
far more accuracy than you'd even need on a boat.
Your automotive GPS is constrained by things like roads. My own, for
example, will "snap" to the nearest road unless I am something like
100m away from a road, no matter that I am in fact in a car-park or on
my drive when it snaps not to the nearest road, but to the nearest
road that runs parallel to my drive!
Your GPS isn't doing that. Your mapping software is doing that. As you
know you can snapping to road offf and the results will still be
execllent.
Of course; that is exactly the point I was making. I note that the GPS
software quite happily snaps to roads more than 5 m away from my
actual position.
To be pedantic, all GPS positions are outputs from software. The
assumptions the software makes differ from application to application.
In every case, the aim of the software is to make use of as much
information as possible to refine (if possible) the position.
It can also apply forms of
filtering that are inapplicable for a nearly stationary object such as
a boat; you can use the history of a car's motion to predict where it
ought to be pretty well.
Karman filter.
So, the automotive GPS has a lot of chances
to reduce the basic GPS error which are not applicable to a boat at
anchor. You can't rely on a better than 10m horizontal error position.
You would do better to simply average the position within a
time-window, which has the potential to substantially reduce the
error.
Are you saying saying determing a position is easier in a moving car
with potential obstuctions to the view of the sky than in a stationary
boat with a clear view of the sky? If you're not what is your point?
Yes, I am. The problem is much more constrained by the additional
information that can be used to get a position for a car than it is
for a freely moving boat where the only constraint that can be applied
is that it is on water. In a car any obstructions to the signal will
be transitory; a good fix will come along very soon after a bad one,
and (as you suggest) Kalman filtering will help greatly. In a boat,
the rigging or noisy electrical equipment will cause constant
obstructions or signal degradation. Of course, a boat keeping a
specific course can use some of the optimizations, but not all.
By the way, I work in a group which uses GPS for extremely accurate
position fixing (mm accuracy at best). Unconstrained GPS provides a
position with a 95% chance of being within 10 metres of the correct
position. I would expect some change in peformance related to time of
day because of electrical activity in the upper atmosphere; one of the
sensitivities of GPS is to the Total Electron Content of the
atmosphere between the observer and the satellite, and this has a
diurnal component.
Paul
.
- Follow-Ups:
- Re: Anchor alarm
- From: toad
- Re: Anchor alarm
- From: Keith
- Re: Anchor alarm
- References:
- Anchor alarm
- From: Len
- Re: Anchor alarm
- From: Wilbur Hubbard
- Re: Anchor alarm
- From: TonyB
- Re: Anchor alarm
- From: Alisdair
- Re: Anchor alarm
- From: toad
- Re: Anchor alarm
- From: Paul Cooper
- Re: Anchor alarm
- From: toad
- Anchor alarm
- Prev by Date: Free WAP Inshore Forecast?
- Next by Date: Re: "To finish first, first you must finish"
- Previous by thread: Re: Anchor alarm
- Next by thread: Re: Anchor alarm
- Index(es):
Relevant Pages
|