Aerobrakes — Lesson 6: Algorithm Implementation — PID

---

With bang-bang implemented and in the metrics table, the team moves to a controller that can respond proportionally rather than just fully-on or fully-off.

Note

Lessons 5, 6, and 7 are one combined three-session block covering Bang-Bang, PID, and Responsive control. They share the same underlying activity structure and metrics table; each session's page here highlights that week's specific algorithm.

Objective

implement, tune, and evaluate the PID controller against the OpenRocket benchmark from Lesson 2.

Session Activities — PID

Implement the PID controller in the plugin, applying the proportional/integral/derivative intuition from Lesson 3's thermostat-versus-cruise-control discussion. Tune the three gains against clean, noise-free simulated data first — realistic sensor noise doesn't get introduced until Lesson 8.

Metrics

Continue logging into the same shared comparison table started in Lesson 5: altitude error versus target, overshoot magnitude, actuator cycle count, and settling time. This is where PID's tuning trade-offs — usually tighter final accuracy at the cost of more actuator cycling than bang-bang — should start becoming visible in the numbers.

Maps to milestone P1-M4.