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:Stochastic Model Predictive Control: Tractability
and constraint satisfaction - Professor John Lyger
os (Head of the Automatic Control Laboratory\, ETH
Zurich)
DTSTART;TZID=Europe/London:20091201T160000
DTEND;TZID=Europe/London:20091201T170000
UID:TALK21278AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/21278
DESCRIPTION:Exploiting advances in optimization\, especially c
onvex and multi-parametric optimization\, model pr
edictive control (MPC) for deterministic systems h
as matured into a powerful methodology with a wide
range of applications. Recent activity in robust
optimization has also enabled the formulation and
solution of robust MPC problems for systems subjec
t to various kinds of worst case uncertainty. For
systems subject to stochastic uncertainty\, howeve
r\, the formulation and solution of MPC problems s
till poses fundamental conceptual challenges. Opti
mization over open loop controls\, for example\, t
ends to lead to excessively conservative solutions
\, so optimization over an appropriate class of fe
edback policies is often necessary. As in the case
of robust MPC\, the selection of policies one con
siders is crucial and represents a trade-off betwe
en the tractability of the optimization problem an
d the optimality of the solution. Moreover\, in th
e presence of stochastic disturbances hard state a
nd input constraints need to be re-interpreted as
chance constraints\, or integrated chance constrai
nts\, which may be violated with a certain toleran
ce. This interpretation\, however\, makes it diffi
cult to enforce hard input constraints dictated by
the capabilities of the system and the actuators\
, especially if one considers desirable classes of
feedback policies such as affine policies. And wh
at guarantees can one provide in the infinite hori
zon case\, given that the system evolution is obta
ined by solving an infinite sequence of finite hor
izon problems each of which may violate its constr
aints with a finite probability? This talk will ou
tline these challenges and propose solutions for s
ome. The resulting stochastic MPC methods will be
illustrated on benchmark problems and compared wit
h alternatives.
LOCATION:Cambridge University Engineering Department\, Lect
ure Room 5
CONTACT:Dr Guy-Bart Stan
END:VEVENT
END:VCALENDAR