Name: Adaptive control
ISBN: 9780201558661

_dpsJan 2018

There's a whole body of theory on Adaptive Control (what the control systems community has called what we would now called "learning based control").

This is one of the classic texts: https://www.amazon.com/Adaptive-Control-Karl-Johan-Astrom/dp...

In short, many techniques that look a lot like modern machine learning have been in use for decades in control theory for tasks exactly like you describe (e.g. automatic tuning of PIDs). If you want to dig into this a bit deeper you might look up the "MIT Rule", which is a gradient style approach for dynamically tuning (and re-tuning) adaptive control systems, and is fairly common for PID tuning.

thread

cashsterlingApr 2023

I'd say RL is an advanced robotics topic. There are some books out there that cover introductory or general RL aspects... after that you'll have to dive into papers and specific code bases.

Might not be a bad idea to start with a Python-based, practical RL book and move from practical demonstrations towards theory.

Although not an RL book, I really like Data-Driven Science and Engineering by Brunton and Kutz:

- https://www.databookuw.com/

- https://www.amazon.com/Data-Driven-Science-Engineering-Learn...

Steve Brunton also has an awesome youtube channel on dynamic systems, control, and machine learning.

Robotic books with some RL concepts (don't have either of these tbh):

- https://introduction-to-autonomous-robots.github.io/

- https://mitpress.mit.edu/9780262046169/learning-for-adaptive...

General texts/resources on adaptive control, optimal control, and RL:

- https://arxiv.org/abs/1912.03513

- http://www.cs.cmu.edu/~cga/dynopt/

- http://www.mit.edu/~dimitrib/RLbook.html