vas
Well-Known Member
hm, summer season over, p1ssing down last few days, boat on it's mooring till next June.
Teaching delayed for a week so starting on Monday, had a few hours over the last week to check and finetune the simplified version of the code.
Two itterations, first shot both video from the dock and the capture of the variables plotted dancing about.
Second just variable dancing (much more meaningful in debuging)
Basically minimised (zeroed!) delay, if rollrate goes off zero value, fins react. Still trying to remember why I wasn't doing that in the first place and was deliberately introducing delays and roll.
Problem is although my code is almost 50-50 code vs comments, it still takes me a good few hours in at least two consecutive days to remember what I've done and why.
Anyway, latest test was on the 7th of Oct (all tests done with the liner enters port in the evening) and was much better than previous ones in consistency and performance.
Analysing the graphs I came up with yet another idea on triggering fin movement as I noticed that on many decent rolls, fin initial movement gets boat rollrate almost zeroed, fins stop moving waiting for next roll, but the remaining of the wake re-acts the rolling and fins still wait for rollrate to zero and then rise before acting... Not quite easy to describe tbh on the last video I'm linking below it's clear at circa 20sec when I first activate the fins, a half flap stabilises (not perfectly) but then fins stop and boat starts rolling again. Fins have more travel they could do and minimise that as well, but are not triggered as they've acted as instructed in the same heel earlier. Hope a dozen of lines of code could sort that we shall see in next test
If you're really bored you can have a look at these three videos one and three are the geeky ones, second is the boaty one ?
and last attempt:
my next attempt at videoing the thing will be to sync graphs with outside video in one video, that will make following what's going on much easier!
cheers
V.
Teaching delayed for a week so starting on Monday, had a few hours over the last week to check and finetune the simplified version of the code.
Two itterations, first shot both video from the dock and the capture of the variables plotted dancing about.
Second just variable dancing (much more meaningful in debuging)
Basically minimised (zeroed!) delay, if rollrate goes off zero value, fins react. Still trying to remember why I wasn't doing that in the first place and was deliberately introducing delays and roll.
Problem is although my code is almost 50-50 code vs comments, it still takes me a good few hours in at least two consecutive days to remember what I've done and why.
Anyway, latest test was on the 7th of Oct (all tests done with the liner enters port in the evening) and was much better than previous ones in consistency and performance.
Analysing the graphs I came up with yet another idea on triggering fin movement as I noticed that on many decent rolls, fin initial movement gets boat rollrate almost zeroed, fins stop moving waiting for next roll, but the remaining of the wake re-acts the rolling and fins still wait for rollrate to zero and then rise before acting... Not quite easy to describe tbh on the last video I'm linking below it's clear at circa 20sec when I first activate the fins, a half flap stabilises (not perfectly) but then fins stop and boat starts rolling again. Fins have more travel they could do and minimise that as well, but are not triggered as they've acted as instructed in the same heel earlier. Hope a dozen of lines of code could sort that we shall see in next test
If you're really bored you can have a look at these three videos one and three are the geeky ones, second is the boaty one ?
and last attempt:
my next attempt at videoing the thing will be to sync graphs with outside video in one video, that will make following what's going on much easier!
cheers
V.
