2 takeoff solutions
Anyways, the T4I produced some more videos.
Recalibrating Marcy 1 continues. There are lots & lots of parameters in machine vision which have to be reconfigured if she is to fly with the lights on. Manely, the algorithm for daylight requires tracking multiple points of light while the algorithm for darkness tracked a single oval.
Also, the 1st calibration had a lot of errors.
There are 2 routes from this step. Reuse the original code which controlled attitude during the takeoff or use a hard coded lookup table of battery voltages to instantly jump off the pad without attitude control.
The lookup table would conceivably never fail, would be fast & simple, but it's never been done before. The problem is it takes off before vision has enough time to fully initialize. It relies on enough stability & camera field of view for the vision to initialize after it's taken off.
The attitude control was sort of like the uncontrolled lookup table. It could only detect attitude on the ground, so it leveled & searched for the right throttle for the given battery voltage on the ground. Then it had to resort to a similar uncontrolled throttle ramp to get off the stand before it tilted. It was complicated, slow & the throttle ramp tended to either overshoot or undershoot.
It's heading towards an attitude controlled system on the ground with a lookup table for the takeoff. The minds that existed in 2011 used well learned knowledge in all the takeoff issues to arrive at the attitude controlled takeoff.
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|For Sale||HELICAM SOLUTIONS MARK IV RC HELICOPTER CAMERA MOUNT cost $2,700||Go4it||FPV Equipment (FS/W)||3||Jul 24, 2012 01:34 AM|
|Video||707 aborted takeoff camera angle 2||demondriver||Electric Plane Talk||0||Nov 14, 2011 10:22 PM|
|For Sale||NIB Simprop Solution 2.0 - 2.835M semi-scale||dckemper||Aircraft - Sailplanes (FS/W)||0||Feb 15, 2006 10:28 PM|
|Sliding left on takeoff?........SOLUTION!!....||SargeNZ||Micro Helis||7||Mar 27, 2005 10:44 AM|