Obsidian/.archive/4 Qualifying Exam/3 Notes/How is robust control validation done?.md

Short answer, it isn't.

There's a lot of papers that do 'robust control synthesis and verification' but the verification is actually formal methods

1. farzanRobustControlSynthesis2020
   1. Uses reach tubes to claim robustness.
   2. Verification is several trials with one single plant perturbation.
   3. Different initial conditions for sure, but the plant is only changed once, and is actually a structured perturbation (20% less stiffness)

I hate robot people. They use the word robust too softly.

2. matusuRegionsRobustRelative2023
   1. This plant has a structured perturbation
   2. The stuff they do with PI is kind of cool I guess
   3. No formal methods
   4. They reinvent the factor of safety. Silly