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:Pre CDC talk - Richard Pates\, David Hayden and Pa
nos Brezas from Control Group\, Engineering Depart
ment
DTSTART;TZID=Europe/London:20121207T140000
DTEND;TZID=Europe/London:20121207T153000
UID:TALK41998AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/41998
DESCRIPTION:1. Stability Certificates for Networks of Heteroge
neous Linear Systems by Richard Pates\n\nAbstract:
We derive robust\, decentralized and scalable sta
bility conditions for networks of heterogeneous li
near systems. The results are based on a graphical
Nyquist type test which depends only on the frequ
ency responses of an individual system and its nea
rest neighbours. If every system in the network sa
tisfies such a condition\, the network is guarante
ed to be stable. The results are suitable for a br
oad class of networks\, allowing arbitrary interco
nnection asymmetries and heterogeneous multi-input
-multi-output systems.\n\n\n2. Robust Network Reco
nstruction in Polynomial Time by David Hayden\n\nA
bstract: We present an efficient algorithm for rob
ust network reconstruction of linear time-invarian
t systems in the presence of noise\, estimation er
rors and unmodelled nonlinearities. The method her
e builds on previous work on robust reconstruction
to provide a practical implementation with polyno
mial computational complexity. The algorithm greed
ily obtains a set of structurally-related candidat
e solutions spanning every level of sparsity. One
of these structures is guaranteed to be the correc
t solution if a magnitude bound on the noise is sa
tisfied. Model selection tools may then be used to
select a single solution from this set. Extensive
simulations quantify the expected performance for
different levels of noise and show that significa
ntly more noise can be tolerated in comparison to
the original method.\n\n3. LQ optimal and risk-sen
sitive control for vehicle suspensions by Panos Br
ezas\n\nAbstract: The talk is concerned with time-
domain optimal control of active suspensions. The
optimal control problem formulation has been gener
alised by incorporating both road disturbances (ri
de quality) and a representation of driver inputs
(handling quality) into the optimal control formul
ation. A regular optimal control problem as well a
s a risk-sensitive exponential optimal control per
formance index is considered. Emphasis has been gi
ven to practical considerations including the issu
e of state estimation in the presence of load dist
urbances (driver inputs).
LOCATION:Cambridge University Engineering Department\, LR6
CONTACT:Dr Jason Z JIANG
END:VEVENT
END:VCALENDAR