FRC-docs simulator tutorials - what do you want to see next?

[Oblarg]

As you may have seen. FRC-docs just added a sequence of simulation-based PID tuning tutorials courtesy of @gerthworm. These should help your programmers get a feel for the mechanics of PID control without access to a robot (or without risking damaging the robot with poorly-tuned control gains).

The simulators included at the moment are very basic, and only allow the user to tune the various control loop constants while keeping all the other simulation properties hardcoded. This is great for teaching PID tuning, but there’s a lot of potential to teach other things with these sorts of simulators, too!

I’m considering writing a number of content pages that use similarly-structured simulations to teach other important control theory concepts. However, I need guidance on what sort of content the community feels is most-important or most-useful, since writing one of these pages is a fairly substantial endeavor and I don’t want to do it for content that people don’t want or need.

With that in mind, here’s a poll with a couple of pages I’ve been considering making. Please rank the choices in order of interest/need for your team, and leave any comments on the general idea in the thread below.

What simulation-based FRC docs tutorial page would you like to see next?

• Sensor and system noise, signal delay, and stability
• Motor power and mechanism mass
• Friction and friction compensation
• Motion profiling and acceleration feedforward
• LQR and “optimal” tuning
• “Full” simulator page with tuneable selectors for all model parameters (motor/gearing choice, mechanism geometry/mass, friction, system noise, friction, and signal delay) - will not have any accompanying tutorial (too many free parameters to guide interaction meaningfully).

[AllenGregoryIV]

A ProfiledPIDController + elevator tutorial would be my first vote.

[Weldingrod1]

I’d like to see a simulator where you can vary flywheel mass/intertia and motor power and look at its shot to shot consistency, in addition to its spin up time.

A nice knob to turn would be the ratio between the flywheel and the actual shooting wheel. If you add this then an overall system weight number could be a useful reference. Fast flywheels = more energy storage per pound. That said, this is probably getting too far from “useful tutorial” and off toward “thesis”

A cool display parameter would be startup energy consumed (Joules or maybe a translation to Amp Hours). I was pretty shocked when I calculated this parameter! Running the shooter the whole match was less energy consumed than spooling up a couple of times!