Publications,
Reports and Other Links
Michael G. Safonov
Unfalsified Control Theory: Quantitative theory of learning and adaptation
· Let the data speak...
· Use evolving real-time data to unfalsify (validate) controllers against hard performance criteria:
o Choose criteria expressible directly in terms of observed data (sensor outputs, actuator inputs).
o Avoid criteria based on “noise models” or other prior beliefs.
· Whenever the currently active controller is falsified by evolving I/O data, it is automatically replaced by an unfalsified controller.
· Selected papers & software demos on Unfalsified Control.
SLIDE SHOW on Data-Driven
Unfalsified Adaptive Control
|