1) I understand the basic control theory of how QC's maintain balance. It seems that each ESC board is driven by the controller board. Assuming the motors are labeled N, E, W, and S, N & S would be rotating one direction and E & W are counter-rotating. If you wish to tip the 'copter, pivoting on the E & W axis, you'd decrease the N motor RPM and increase the S motor RPM. If both the N & S motor are turning at 50% of max RPMs (whatever that may be, doesn't matter), you'd decrease N to 45% and increase S to 55%. Are separate channels necessary? Is there a reason that the N can't be the complement of the S proportion? Or is there a scenario where you'd keep N, E and W running at 50% and increase S to 55%? I've not found enough info on the flight characteristics (or math simple enough) to know if such a scenario exists. I'm not trying to optimize a channel out of existence, just wondering about how motors are managed in general.
2) All the algorithms I've run across so far are PID type controls. Has any experimentation been done with fuzzy logic?
3) Thermal Intelligence sensors are used on the Dammar controller. Are these effective a low altitudes, like below the tree line? I haven't found a lot of information on how these work, other than indicating they sense the horizon. If you have a tree line on one side, it seems your horizon would be skewed.
4) Given the 4 rotors on a QC (and having never flown one), it appears that it's not possible to rotate about a point as you could with a single rotor heli with a tail boom. Do any designs include an additional tail-boom type rotor, or some other mechanism to allow turning on a point? Seems like that would be a useful feature for a USPV or pre-programmed flight plans.
Assuming I actually follow through and build one, there's a couple features (I think) I'd like. One is a pusher-mode, where you'd basically have altitude, position and heading hold, and pushing the stick forwards sends it forward (there's probably a name for this flight type, I don't remember if I've seen it), releasing the stick stops it.
I'd also like to be able to do a pre-programmed flight plan. Maybe lay it out in Google Earth or some other tool, down load it and go. The human will worry about pre-programming for collision avoidance :)
I'm also quite interested in the POV flying. It may take a strong stomach for that, but it sure looks like fun. And that's where a positional hold mode would be really nice.
Thanks for reading this,
--jc
Replies
I call that response type "translational rate command / position hold". I plan to implement it on my quad. I'm not sure how well it will work using just accelerometers and gyros, but I'll give it a shot.
- Roy