PID Video

[horner]

Notes from prior PR:
#2609

[calcmogul]

This 15-minute video's explanation of feedforward is good, but the middle third brings up frequency domain stuff that isn't relevant to FRC students (they use modern controls for basically everything, whether they know it or not), and the last third presents a Simulink example that isn't reproducible by FRC students, cuz MATLAB costs money. frc-docs already has some JS tutorials for feedforward and feedback.

[calcmogul]

The correct references would be chapter 2 "PID controllers" and section 7.8 "Feedforward"; in a.b.c.d, a is chapter, b is section, c is subsection, and d is subsubsection.

By the way, this page is already a verbatim copy of chapter 2.

[gerth2]

super late to the party but since I was in here looking for controls-related PR's anyway....

What I do think the video brings up is the notion that an FRC-focused pedagogy is the best way to move forward for this competition. While I personally found the video useful, the jump to other tools and laplace domain is going to be a jarring experience for the person who just learned PID fifteen minutes ago.

Core elements I think are useful to a pedagogy in future doc changes:

1. Time-domain focus - Not assuming other domains will be tractable by the average high schooler.
2. Specific examples tied to motors and motor-driven things - this is all an FRC student knows (and mostly all they'll ever need to know about)
3. Consistent application vocabulary (docs are solid on this currently, but caution is needed with external resources that refer to the same thing with different words)
4. Scaffolding by desired outcomes - 99% of students start down the controls path because they're asked to achieve some behavior on a robot. Good flow will involve trying to get them to build up to that behavior, recognize their success, and look to the future to see what else is possible. Critically though... not by skipping steps, and not by discarding ideas simply because they're "simple" or "non-optimal".