Has any team ever implemented a feed forward control system design with their robot for tele-op? Is anyone trying to do it this year?
If it has never been done before, and say team xyz does it successfully, I’d be willing to be it is guaranteed innovation in control award, even at championship.
Thoughts?
What do you mean specifically by feed forward?
I know many teams already are using feed forward in addition to their normal control loops.
Never heard of anyone winning an award for it though.
Do you mean a feed forward loop without feedback? That just doesn’t seem like it would work by itself, seems like you need feedback to verify your feed forward terms.
33 actually did this last year (someone please correct me if wrong).
They generated trapezoidal accell curves for their drive, and used that with some feedfoward gains to get a smooth drive forward motion w/o sensors (not huge precision on distance obviously).
Yeah I guess that’s true. But for that robot it worked. Try doing only feed forward with 254’s auto… :eek:
I could be wrong here as well, but I think their auto heavily relied on FF, and the control loops handled deviation from path.
From my reading of their code, you’re not wrong…
And there’s no such thing as an assured award. Ever. If students can’t explain it effectively or it’s just flat out not effective it shouldn’t (and probably won’t) get an award. Doubly so at CMP.
In regards to who wins what awards at Championships, I don’t know that embedded control systems (such as feed-forward) are the only subject where the Innovation in Control is aimed. It’s pretty broad in scope. I’ll also echo Andrew’s sentiments about students explaining and demo’ing the technical side on a live robot.
We won Innovation in Control on Galileo for an on-goggle HUD that displayed real-time telemetry to the driver without distracting her. It was similar in functionality and design challenges to the F-35 helmet. Rockwell Collins delivered that helmet to the F-35 program mere weeks after Champs, so I’m guessing that’s why they were so impressed by the HUD. We got to meet a few of the executives at Champs to demo it to them.
I read earlier this season of a team who won the award in their division for their open-sourced command-based Java code architecture. Can’t seem to find the CD post at the moment.
We use feedforward extensively in our control loops. Our 2014 auto mode generated kinematically-feasible trajectories (position, speed, acceleration) for each side of the wheelbase, and then used feedforward gains to follow the path. There was a closed loop component on following error and heading error, but these only had to cancel out disturbances and imperfections as Adam said. Even without the feedback, the robot drove pretty close to the desired path.
This concept generalizes nicely to almost any closed loop system (ex. we have also used feedforward in our flywheel controller), and helps to bridge the gap between hand tuning PID and doing more sophisticated (but still conceptually simple) model-driven types of control. Feedforward gains are often easy to measure/tune in practice, make tuning the PID terms much simpler than if you are relying on PID to do the whole thing, and can be “smoother” than a purely feedback based approach.
I’ve submitted a proposal to do a technical session on “Motion Planning & Control for FRC” to the conference portion of the World Championship this year. If it is selected, you might want to attend if you are interested in this sort of thing 
Just took a read, but I don’t think you won any 2015 awards with this (seeing as the 2015 competition season isn’t here yet :rolleyes:).
I just finished a feed forward control loop in the lab I worked at.
I never learned the frc control system and what people have done. I briefly looked at 254s code back in the day and recognized that it was utilizing ff techniques.
So basically I made this thread without doing research because I was in class when this thought came to mind and that’s when I posted it.
My goal over the course of the next n years I’m with frc is to teach teams machine learning techniques and have them implement it on a robot, whether at competition or not. Feed forward is one I want to see implemented on a massive scale.
Thanks for your responses