BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:CUED Control Group Seminars
SUMMARY:Statistical Change Detection for Prognosis and Dia
gnosis - Professor Mogens Blanke\, Technical Unive
rsity of Denmark
DTSTART;TZID=Europe/London:20141104T140000
DTEND;TZID=Europe/London:20141104T150000
UID:TALK55945AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/55945
DESCRIPTION:Statistical change detection plays a key role in p
rognosis and diagnosis of faults. Design of change
detection algorithms is well known in theory\, bu
t practice often violates the idealised perquisite
s that theory requires. This talk provides tutoria
l insight in existing methods for change detection
in the presence of Gaussian or Non-Gaussian noise
with focus on reliable diagnosis / prognosis wher
e accurate knowledge of detection and false alarm
probabilities is required. As these features are d
etermined by properties of the test statistic of t
he particular problem\, the talk discusses the con
sequences of correlation in real life\, and compar
es theoretical results with those obtained in actu
al applications. A methodology for change detectio
n is suggested that adapts to real-life conditions
. Using estimation of distribution parameters for
the actual test statistic\, threshold for hypothes
is testing and test sequence length are shown to b
e parameters in an optimization problem that can g
uarantee prescribed properties. Detection with mul
tiple detectors\, based on different indicators of
change\, is then discussed\, and it is shown how
a Copula description of joint probability can be e
mployed to assess properties of combined detectors
. Selected industrial applications from unmanned a
ircraft and drilling illustrate the methods.
LOCATION: Cambridge University Engineering Department\, LR1
2
CONTACT:Tim Hughes
END:VEVENT
END:VCALENDAR