New History Plugin for OpenCPN

[seandepagnier]

I wanted a quick easy way to analyse my sailboat's autosteering abilities. For this I need a log of gps positions and and a way to easily plot them. With the patch I have, this has become reality (with a few clicks) from opencpn, but I believe more improvements could be made, and there are many more features to add (like calculating the optimal tack angle etc...)

For a demo I used my solar powered electric trolling motor to move around the anchorage and obtained the following screenshots

1. course with respect to time
2. speed with respect to time
3. latitude with respect to longitude
4. screenshot showing interface


As well as plotting ability, it is possible to store the data to a text file, and also plot the trail (breadcrumb) plus show a more accurate predictor which is based on history rather than current position, speed and course.

What other improvements could be made? I will post the patch shortly (still doing improvements)

[gjorgensen]

Wow, that is really cool!

Since you asked for suggestions, how about programmed heading versus actual heading with respect to time, in other words a delta between the intended course to steer and the actual? That would give you a really good idea how well the autopilot is working. You could do the same with actual track versus expected track if steering off a waypoint.

[seandepagnier]

It is already possible to plot rate of change of heading which I did not show the plot of. This plot makes it pretty easy to see how straight you are steering.

I was thinking about taking the fourier transform of heading so that you could see how the wave tendencies and autosteering time constants match up. For stuff like jib-sheet steering, it would be cool to get quick feedback so you can adjust the sheets and tension the bungee appropriately. Same for pendulum wind vanes.. I want to vary the vane size as well as the counterweight as well as helm offset and gain (position along tiller the lines pull) and this way I can get feedback.

Of course, steering the straightest course is not everything, the distance made good over ground may well be greater if the autosteering uses the waves to propulsion advantage weaving back and forth between them.

[idpnd]

Originally Posted by gjorgensen

Wow, that is really cool!

+1

[sailias]

This is really cool. If you have access to a Raymarine X-5 system the computer has these algorithms for tacking already build in but you need wind speed data inputs. So it automatically calculates the best time to tack based on wind direction, angle, and speed over ground. Perhaps there is some help.

Congratulations. This is very cool,.

[dpons]

Really neat.

I would suggest polar plots on heading, to avoid the not very nice crossover lines between 359 and 1. Or even more easily, don't plot between these two headings.

I attach an example.

[seandepagnier]

I'll add an option for polar plots for heading.

I do not have access to a Raymarine X-5, is it open source? Eventually I will be implementing electronic sheet control which should help pointing ability and overall speed (as wind speed changes)

As far as wind speed/direction, magnetic heading, speed through water, gyroscopes and accelerometers (wave action) etc... (all the data gps does not cover) I have all the hardware to do this, but opencpn needs a better interface for reading this data, and publishing it to plugins. Preferably supporting as many systems as possible.

If anyone knows, please explain how some sailing algorithms could work (maybe that is my next plugin) I could optimize gps speed over ground by slight variations in heading and sheets. But everything changes based on currents and windspeed (or if you reef), but it should still be useful, and pointing angle will change based on it. Maybe we can build support into opencpn to automatically calculate routes that avoid water too shallow for your draft, take advantage of tides/currents, and automatically calculate tack angles and when to do so.

[idpnd]

CapCode-software suite for sailors | Download CapCode-software suite for sailors software for free at SourceForge.net <- capcode has polars in (free software) java

[sailias]

Wrt to Raymarine Im pretty sure its an emedded linux. Will check. Basically how they do it is as follows from the stuff I read in the manual. Assume you are at point A and want to get to point B. Assume as straight vector. Now with the input of windspeed and direction it gives you the best course to head to Point B. As you accellerate you gain speed but also move away from Point B to a degree. The algorithm determines the best time along the course/track to tack to maintain optimum speed and shortest distance to Point B. Tack and then the calculation repeats again on this second tack.

Basically I think its just vectors but above my math skills at present.

[seandepagnier]

I fixed some problems, but it probably wont work in windows, I am posting the patch if anyone wants to test it.

[seandepagnier]

Get the patch here: (not above)

FS#135 : history plugin for opencpn

[sailias]

apologies for the nubie question but applying the patch with patch < filename.patch barfed for the first chunch and then proceeded for the rest. Im assuming this is a git thing....so any instruction on how to apply this patch would be most welcomed.

Best regards and thanks

/ch