MIT16_410F10_lec12b

MIT16_410F10_lec12b - Model-based Monitoring Brian C...

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 3/6/00 Model-based Monitoring Brian C. Williams 16.412J/6.834J March 6 th , 2006 Brian C. Williams, copyright 2000 Due to the unknown mode, there tends to be an exponential number of diagnoses. U Good Candidates with UNKNOWN failure modes G Good Candidates with KNOWN failure modes U Fault Models don’t help. F1 Fn G U 1 Image credit: NASA. 3/6/00 2 Due to the unknown mode, there tends to be an exponential number of diagnoses. U Candidates with UNKNOWN failure modes Good G Candidates with KNOWN failure modes Good F1 Fn G U Fault Models don’t help. Most of the density space may be approximated by enumerating the few most likely diagnoses But “unknown” diagnoses represent a small fraction of the probability density space. U Sequential Model-based Diagnosis Input: • Set of component mode variables M, with finite domains. • Set of observables X, with finite domains. • Device model Φ over M and X, in propositional logic. • Prior distribution P(M i ) of mode assignments for each component i. • Observation sequence X 1,n = x 1,n provided dynamically. Output: • P(M) Prior Probability of Failure • P(M | X 1,n = x 1,n ) Posterior Given Observation updated after each observation is received. Assume: • Independence of component mode prior distribution. • Conditional independence of observations given candidate (Naïve Bayes). • Uniform distribution of observables, given candidate. 3/6/00 3 Mode Estimation Example Inverter(i): • G(i): Out(i) = not(In(i)) • S1(i): Out(i) = 1 • S0(i): Out(i) = 0 • U(i): X Y A B C 0 0 Nominal, Fault and Unknown Modes • Isolates surprises • Explains Sherlock...
View Full Document

{[ snackBarMessage ]}

Page1 / 12

MIT16_410F10_lec12b - Model-based Monitoring Brian C...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online