What little spare time I've had for testing has been lost to rain. I installed my shiny purple OSHpark baseboard last night.
I found and fixed a bug in the estimation code. I was multiplying a heading error tern by a fraction to reduce its influence over time. The difference was normalized to a range of -180 to 180 only after multiplying. It should have been normalized before the multiplication.
I may stand a chance if...
- the last critical bug is fixed (I know there are plenty of other bugs),
- I can get a working GPS,
- the encoders are fixed once and for all,
- I can find a good-enough Kalman Filter tuning,
- I can get accurate measurements for waypoints Saturday