June 07, 2008
May Series #2 with analysis
I've been working on some MATLAB scripts to come up with some analysis of this and hopefully other races using the GPS data.
Here's the track of the race.
For the first leg, I plotted the relative angle to the mark, that is the angle between the direction of travel and the mark:
I'm trying to find a way to reduce the noise in the data, it's created due to the way I process the dirction of travel. I take the angle from the previous data point, and this tends to jump around a lot.
It's still apparent from the plot that as I got closer to the destination the rate at which the angle to the mark opens up increases. I also seem to let that angle get bigger before I tack which I'm not sure is such a good thing.
The next plot I generated is a plot of Velocity made good to the mark:
Again, there's a lot of noise in the data, due to the angle changing a lot between data points. The red line is the mean value. It shows a little better than 5 ft/sec average VMG, which is just under 3 kts.
I tried taking out the variability in the boat speed due to gusts, etc and plotted the VMG as a percentage of current boat speed. This should in effect normalize the VMG and give insight into how well I was pointing the boat:
Although I didn't get the mean value, it appears it's above 50%. Again, it's easy to see how fast VMG changes when I was close to the mark.
Another thing I was interested in was what the 45 degree laylines were relative to the Rhumbline:
From this it's apparent I chose the right side of the course. It appears a few times that I was sailing on the headed tack on a couple of the starboard legs, but I was sailing toward the rhumbline at the time, and I tend to watch the angle to the mark more when I'm more than halfway up the course.