Control Bootcamp: Linear Quadratic Gaussian (LQG)

This lecture combines the optimal full-state feedback (e.g., LQR) with the optimal full-state estimator (e.g., LQE or Kalman Filter) to obtain the sensor-based linear quadratic Gaussian (LQG) controller.
These lectures follow Chapters 1 & 3 from:
Machine learning control, by Duriez, Brunton, & Noack
www.amazon.com/Machine-Learni...
Chapters available at: faculty.washington.edu/sbrunto...
This video was produced at the University of Washington

[@math_engine_pro5082]

Could you tell me, what would be the state space model of the feedback system (final closed loop) with LQG SERVOCONTROL (LQG + Integral Action)? Please, I am doing a project with LQG servocontrol or some book where I can find it

[@math_engine_pro5082]

Podria decirme ,¿Cual seria el modelo en espacio de estados del sistema realimentado(lazo cerrado final) con SERVOCONTROL LQG(LQG + Accion Integral) ?porfavor estoy haciendo un proyecto con servocontrol LQG o algun libro en el que lo pueda encontrar

[@pokerihardlyknowher]

Anyone else confused by this magical backwards glass writing?

[@peanutboy41]

I was at too first but I think the video is mirrored

[@jeroendebest1711]

Nice explanations, as are all other video's. Maybe good to refer to a very famous paper by J.C. Doyle called Guaranteed Margins for LQG regulators. Abstract - There are none.

[@jeroendebest1711]

Sorry, two videos later you exactly mention this particular paper. Cool!

[@pilliozoltan6918]

There's a BKr in off diagonals, why does it called diagonal?

[@simonb.979]

i was actually wondering about that too. So, clearly it is coupled and the eigenvalues of the entire system have changed, right?

[@parkerseth14]

I think he meant to say triangular. The eigenvalues of a triangular matrix are exactly its diagonal entries, just like a diagonal matrix!