Plugin: DashBoard

[rgleason]

Big Tom:

Related Post #301
http://www.cruisersforum.com/forums/...ml#post1914181

I created a combined apparent-/true wind angle dial instrument requested in FS #1687, see attached picture :

Needles: Orange =apparent wind angle, Blue =true wind angle.
Data in 4 corners:
Top: app. wind angle + speed indicated by A: XX° (A=apparent)
Bottom: true wind angle + speed indicated by T: XX° (T=true)
Needles and data have the same colours.

Is your AW-TW Instrument in the current v4.1.915 "android" branch in dashbooard? I am having trouble finding it. Is found by the title

"App & True Wind Angle" ?
If there is no nmea data running is there just one orange arrow and no 4 corner data boxes?
--The addition of "Angle" in the title threw me off because it is both Angle and Speed. Small thing but why not call it App & True Wind or A-T Wind?

==============================================

Related Post #307
http://www.cruisersforum.com/forums/...ml#post1915794

XTE display in the routing module is not really useful "sailing off the road"..XTE is always way too high and highway-display not useful. ...really need a decent steering direction, i.e. "go left" or "go right".

I set a waypoint on my GPS as a "target to go to" which creates a RMB sentence in the NMEA stream. ...I did (in my local O copy) an instrument now, reading this RMB sentence ...

Needles: Orange=Course to steer to waypoint
Data in 4 corners & middle
Top: Bearing to waypoint & Waypoint name
Middle: Course over ground
Bottom: Distance to waypoint & VMG to waypoint

Needle is dead-stable and shows the bearing to the waypoint., the further you're away, the less it changes ... Of course it's affected by your COG, but even in rough sea, that doesn't change so much ...

Advantages :
In one glimpse, how far you have to steer to either side to hit the WP.
No way to see this with XTE and highway.

When tacking upwind, ..tack when needle goes more than 45° off course ...

When to tack to make it around the WP :
True Wind Angle (on boat axis) * 2 =
"minimum angle the the needle needs to point sideways" ...
This is far from perfect but better than nothing ...

And the VMG (Velocity made good) is very useful to see which tack is preferable towards the WP.

- Needle points to the Waypoint. (needle will gradually move towards 45 degrees)
-Statement: "Compass Rose is rotating with the needle"
-"The needle shows the course to steer towards the WP, in this case we should go a bit to the left ([Needle] vertical upwards would be the perfect course)."

- Does the compass rose represent the current angle to the waypoint?
Shouldn't it be a port starboard compass going to 180 degrees? with red and green rings? - Maybe I have this wrong....

- What does the compass rose to needle 8 degrees represent? - Bearing to waypoint?

- Why does the compass and waypoint look rotated further clockwise? -Probably 13 degrees which is Course over ground?

- " in this case we should go a bit to the left (vertical upwards would be the perfect course)." - Big_Tom I am confused, do you mean "left", I would think you would steer right?

- I think this could be very good with some changes to the symbolism perhaps.

- Is there a way of getting some tick marks in there at the outer ring for True and Apparent wind direction?

- Do you have a github repos for this?

Thanks for all your ideas, skills and work.

[tom_BigSpeedy]

Sorry for not answering, I was out sailing for one and a half weeks ...

Just to give you an impression, what I'm working on ...



The red dot symbolizes the waypoint bearing, the blue arrow is the local current under your boat.
Then laylines are shown in semi-transparent red (wind from port tack in this example) and green after your tack.
So you can see if you make it around the waypoint ...

The major problem with the current calculation is to estimate your boat drift (="Leeway") based on wind speed and -angle and get it into a formula (or lookup table)...
NKE does it based on boat speed and heel angle, and as I'm in the comfortable situation to have a heel angle sensor in my system, I'm fine ...

Once you have this, it's a simple triangle calculation.

I will include the current direction/angle into the layline calculation.

I will also include TWA and AWA pointers somehow ...


This is far from being ready for testing. I didn't even verify the algorithm for the laylines yet, just included them here for demo purposes ...

Thomas

@Rick: before I upload this to my github repo, I want to test this "live" on my boat in 2 weeks.

[rgleason]

Thomas, this looks very promising. I might have to get a heel sensor (and rudder sensor) so I can use it properly. Thank you. Hope the test goes well.

I was thinking that from
True Boat Vector =Boat COG +SOG
Apparent Boat Vector = Boat HDG + STW
then calculating the resultant would be a Drift Vector =Leeway + Current

How to subtract/Determine Leeway or Current so we can separate them is one problem with this method.

Maybe your technique is better, the righting moment/heel represents a force directly sideways on the boat at 90 degrees to the the Boat Heading, which is probably proportional to the amount of leeway.

Maybe we can enter the degrees heel by hand?
How do you establish the table?
Two the boat sideways and measure forces?

[rgleason]

Tow the boat sideways and measure the force?

[tom_BigSpeedy]

Rick,

the problem is that boat-drift and current may differ after the tack, and the predicted laylines are (normally) inaccurate due to that.

Let me think about a solution to get the heel and or drift into the system other than a separate (expensive) sensor.

Thomas

[bcn]

Thomas,

as you have heading you might calculate a theoretical course/waypoints and then after some time compare to the real position as a result of heading & drift/set.

Not instanteneous and due to the movement or oszillation of the heading and the other factors some filtering/averaging should be done.

Rick's idea about the vectors in other words using position over time instead of instanteneous orientation and SOG...

Hubert

[tom_BigSpeedy]

Hi Hubert,

if you look at the laylines in the screenshot above, they're an "arc" already (5° in this example).
The goal is to set the width of the arc to the degrees of to your SOG variations over time, i.e. the more you yaw over time, the wider the arc gets.
I'll plan to do some exponential smoothing on the SOG variations.

Nevertheless : if you have the current directly from aft on your current tack,
you will only have your boat/heel drift (="Leeway"), and after the tack you may have the current from 90° ...
Or you heel/drift to port and current sets you to starboard, "nulling" your boat drift...

I don't want everyone to buy a heel sensor.
But everyone has a simple bearing compass on board, and oftentimes those have a (mechanical) heel angle display.
Maybe a very, very simple heel-polar with 4 values for 30°, 60°, 90°, 120° for 5, 10, 20 knots would do as well for heel input ?
This is something almost everyone could do in a reasonable amount of time.
Then use the formula to do a basic drift calculation from it...
The formula still has an adjustable factor, as different boats have different drifts angles at the same heel.


We have to see this current calculation as a "best guess", and it should be treated like this, but I already saw last week that it basically works while sailing in real life.

Thomas

[rgleason]

Nice!

goal is to set the width of the arc to the degrees of to your SOG variations over time, i.e. the more you yaw over time, the wider the arc gets.
I'll plan to do some exponential smoothing on the SOG variations.

Yes! Thank you.

I don't want everyone to buy a heel sensor.
But everyone has a simple bearing compass on board, and oftentimes those have a (mechanical) heel angle display.
Maybe a very, very simple heel-polar with 4 values for 30°, 60°, 90°, 120° for 5, 10, 20 knots would do as well for heel input ?
This is something almost everyone could do in a reasonable amount of time.
Then use the formula to do a basic drift calculation from it...
The formula still has an adjustable factor, as different boats have different drifts angles at the same heel.

Nice Idea Bcn; Sort of like delta dX/dY Physics calculations, ..maybe this could be done, but I am not a math wiz.

as you have heading you might calculate a theoretical course/waypoints and then after some time compare to the real position as a result of heading & drift/set.
Not instanteneous and due to the movement or oszillation of the heading and the other factors some filtering/averaging should be done.
Rick's idea about the vectors in other words using position over time instead of instanteneous orientation and SOG...

Great:

have to see this current calculation as a "best guess", and it should be treated like this, but I already saw last week that it basically works while sailing in real life.

[rgleason]

ok

he problem is that boat-drift and current may differ after the tack, and the predicted laylines are (normally) inaccurate due to that.

Let me think about a solution to get the heel and or drift into the system other than a separate (expensive) sensor.

Just thinking;
boat drift = leeway (due to affect of wind) + current <---isn't this the definition of drift?

Drift | Define Drift at Dictionary.com

Navigation. (of a ship) the component of the movement that is due to the force of wind and currents.

Nautical.

1. the flow or the speed in knots of an ocean current.
2. the distance between the end of a rope and the part in use.
3. the distance between two blocks in a tackle.
4. the difference in diameter between two parts, one of which fits within the other, as a mast and its mast hoops, or a treenail and its hole.

Leeway (due to wind) would be similar on both tacks, assuming all things are equal.
Current would be from a different angle, but same speed, therefore Drift would be affected.

Yes, these factors will affect the effective laylines.
Perhaps we need to think about there being two types of laylines.
1. Apparent Laylines based on wind + polar diagram.
2. True Laylines based on wind + polar diagram + leeway + current.

Calculating #1 is relatively straightforward I think.
Calculating #2
Option A
--Leeway -use vector addition for leeway -based on a table , big T suggests.
--Current - Calculate Drift and deduct Leeway above?
Option B
-- Use the instantaneous differential approach that BCN suggests?
-- So the laylines get adjusted based on the last X minutes between two positions?
-- The current layline would be easier to calc than the opposite tack I think, but the comparison of the two laylines is most instructive.

--Random thoughts here.

[rgleason]

Does a recorded polar diagram inherently include leeway?

[rgleason]

We tried to deal with leeway when working with patrick on polauto. --I have to go back and look at it again.

[tom_BigSpeedy]

Originally Posted by rgleason

Does a recorded polar diagram inherently include leeway?

It depends on which input you have for your TWA calculation for heading and speed, I guess ...
If you (the instruments) use SQG/COG for boat speed/course and calculate it against the AWA of the wind vane, which is calibrated against the boat's heading (bow = compass heading), then there will be an influence ...



I like yours / Huberts ideas as well. Could probably easily be integrated into the "layline arcs" ...

I'll defaulted heel/current values anyway to 0, if not available. It won't stop you from using the "instrument". You just don't see the current-arrow...

Thomas

[bcn]

Thomas,

the "expected against measured drift/set" would be a candidate for a Kalman type of adaptive filtering. But I'm not a Mathematician to lay this out as an algorithm I fear.

Kinda Death Reckoning with adaptive filter. And taking HDG, SOG, COG, drift and set as parameters even a tack should fit into the model.

Hubert

[rgleason]

Thomas I am in favor of your idea to adapt Dashboard. I would like very much to work on this this winter in whatever way I can help, no matter what format you decide, because it will be an improvement for sailors. By the way I have not been able to get laylines working in win but I believe it works in linux.

[rgleason]

Thomas, Thinking about polauto recording performance, you are correct, I believe Patrick's software is able to show drift using the inputs you indicate above.