Expedition Expedition Navigation Software

You are not logged in. Would you like to login or register?



11/02/2017 5:07 pm  #11


Re: Damping - smoothing and averaging mechanisms

Thanks Nick. The inside GNSS article from 2016 was an interesting read. I also read a European GNSS Agency report (the GNSS User Technology Report - Issue 1) from 2016. Both suggest that better GPS systems (eg. using the Doppler effect to determine velocity or predictive filters to improce accuracy) exist and are on the radar for the average user, but I see nothing to suggest that they are available in any of the instruments I use today.

An aside: In an attempt to find out what the manufacuters say about their specs I looked up the GPS chip used in my phone (a Galaxy S7 Edge). Eventually I found it's a BCM4751 ... and as for the specs, they're not really available ... not unless I sign an NDA, I suspect. My phone is now a year old and that chip is apparently not recommended for new designs. So, bring on the BCM4774. This puppy does sensor-fusion on-chip and can do some amazing stuff if configured correctly! These technologies are moving forward very quickly at the moment.

Getting back to yachts, from what I can see - and again it's based on the publicly available documentation - it's only the highest end products that do any sensor fusion and intelligent filtering at all. I suspect the GPS in a plotter (even a high-end model) provides nothing much beyond the basics.

So finally, it's back to Expedition. Nick: You obviously understand GPS systems, damping, filters, resolution, error, noise, etc. but unfortunately most of the people who use and configure Expedition and the instrument systems on their vessels do not. When I started this thread I was simply looking for a clear and unambiguous definition of the filter types availabe. The one comment I found "The standard damping weights to the latest. There is a box-car option." was not entirely clear. I think most people (me included) would appreciate a little more guidance than what is currently available in the Expedition help system.

 

11/05/2017 2:50 pm  #12


Re: Damping - smoothing and averaging mechanisms

The sensor data, and processing between actual physical device and any displayed value is far from simple.

Raw position, masthead/paddlewheel speed, heading, measured wind angles all can have offset errors due to poor installation, and deviations due to on board and atmospheric phenomena.  Scale errors can occur due to non linear flows.
Assuming you correct out those errors, the interactions due to heel, and acceleration of pitch, roll are a next.

Some sensors do insitu processing and send out nmea strings, some feed raw signals to a processor, which then become accessible to Expedition.

Leave your plotter powered up for a week, and you may find your boat is more active due to gps position noise than you would expect. Mine apparently has a 100m radius even though the dock lines constrain to 0.5. That does not manifest in SOG that I have ever noted, though having the ZGS-100 on stern rail and paddlewheel & compass near mast base give interesting differences during turns.


Semiconductor Development Engineer, Retired Viticulturalist, Occasional sailor
"Lioness" 1962 Hinckley Bermuda 40 Custom Yawl
 

11/06/2017 12:26 pm  #13


Re: Damping - smoothing and averaging mechanisms

I support what Sheldon says.  An exercise everyone should do is to run your GPS (s) for (say) and hour each while moored firmly in the pen, then load the logfile into Expedition and look at the "track".  It might just surprise you, and certainly give insight into how GPSs work.  Especially good if you have more than one installed and can swap among them.

 

11/12/2017 5:18 pm  #14


Re: Damping - smoothing and averaging mechanisms

I started this thread asking Nick for a "definitive explanation of how damping is implemented in Expedition" but I am still looking for definitive answers on damping, smoothing, averaging, and filtering ... little of which seems to be available in the Expedition help systems.

Sheldon and Duncan: We all agree that GPS data can be a bit noisy. The question is "How should we deal with this?" That is "Is it a problem in Expedition or is it something we should be dealing with before the data is presented to Expedition?" My view is that the primary filtering should be happening before Expedition but that Expedition should be able to help us visualize the data with a view to applying more appropriate filtering etc at the source.

     Thread Starter
 

Board footera

 

Powered by Boardhost. Create a Free Forum

Zapfware NMEARemote by Zapfware. Remote access to Expedition Channels from iPad and iPhone devices